What Have We Learned?: Macroeconomic Policy after the Crisis

MacroeconomicPolicyaftertheCrisis

editedbyGeorgeAkerlof,OlivierBlanchard,DavidRomer,andJosephStiglitz

Introduction:RethinkingMacroPolicyII-GettingGranular

OlivierBlanchard,GiovanniDell'Ariccia,andPaoloMauroPartI:MonetaryPolicy

1ManyTargets,ManyInstruments:WhereDoWeStand?

JanetL.Yellen

2MonetaryPolicy,theOnlyGameinTown?

LorenzoBiniSmaghi

3MonetaryPolicyduringtheCrisis:FromtheDepthstotheHeights

MervynA.King

4MonetaryPolicyTargetsaftertheCrisis

MichaelWoodford

PartII:MacroprudentialPolicy

5MacroprudentialPolicyinProspect

AndrewHaldane

6MacroprudentialPolicyandtheFinancialCycle:SomeStylizedFactsandPolicySuggestions

ClaudioBorio

7MacroprudentialPolicyinAction:Israel

StanleyFischer

8Korea'sExperienceswithMacroprudentialPolicy

ChoongsooKim

PartIII:FinancialRegulation

9EverythingtheIMFWantedtoKnowaboutFinancialRegulationandWasn'tAfraidtoAsk

SheilaBair

10RegulatingLargeFinancialInstitutions

JeremyC.Stein

11TheContoursofBankingandtheFutureofItsRegulation

JeanTirole

12BankingReforminBritainandEurope

JohnVickers

13Leverage,FinancialStability,andDeflation

AdairTurner

PartIV:FiscalPolicy

14DefiningtheReemergingRoleofFiscalPolicy

JaniceEberly

15FiscalPolicyintheShadowofDebt:SurplusKeynesianismStillWorks

AndersBorg

16FiscalPoliciesinRecessions

RobertoPerotti

RobertoPerotti

17FiscalPolicy

NourielRoubini

PartV:ExchangeRateArrangements18HowtoChooseanExchangeRateArrangement

AgustinCarstens

19RethinkingExchangeRateRegimesaftertheCrisis

JayC.Shambaugh

20ExchangeRateArrangements:SpainandtheUnitedKingdom

MartinWolf

21ExchangeRateArrangements:TheFlexibleandFixedExchangeRateDebateRevisited

GangYi

PartVI:CapitalAccountManagement22CapitalAccountManagement:TowardaNewConsensus?

DuvvuriSubbarao

23CapitalFlowsandCapitalAccountManagement

JoseDeGregorio

24ManagingCapitalInflowsinBrazil

MarcioHolland

25CapitalAccountManagement

HeleneRey

HeleneRey

PartVII:Conclusions

26TheCatintheTreeandFurtherObservations:RethinkingMacroeconomicPolicyII

GeorgeA.Akerlof

27RethinkingMacroeconomicPolicy

OlivierBlanchard

28PreventingtheNextCatastrophe:WhereDoWeStand?

DavidRomer

29TheLessonsoftheNorthAtlanticCrisisforEconomicTheoryandPolicy

JosephE.Stiglitz

Contributors

Index

OlivierBlanchard,GiovanniDell'Ariccia,andPaoloMauro

The 2008-2009 global economic and financial crisis and its aftermath keepforcing policymakers to rethinkmacroeconomic policy. Firstwas the Lehmancrisis,which showed howmuch policymakers had underestimated the dangersposedby the financial systemanddemonstrated the limitsofmonetarypolicy.Then itwas theeuroareacrisis,whichforced themto rethink theworkingsofcurrencyunionsandfiscalpolicy.Andthroughout,theyhavehadtoimprovise,fromtheuseofunconventionalmonetarypolicies,totheprovisionoftheinitialfiscal stimulus, to thechoiceof thespeedof fiscalconsolidation, to theuseofmacroprudentialinstruments.

Wetookafirstlookattheissuesafewyearsago,bothinapaper(Blanchard,Dell'Ariccia,andMauro2010)andatanIMFconferencein2011(Blanchardetal. 2012). There was a clear sense among both researchers and policymakersparticipatingintheconferencethatwehadentereda"bravenewworld"andthatwehadmorequestionsthananswers.Twoyearslater,thecontoursofmonetary,fiscal,andmacroprudentialpoliciesremainunclear.Butpolicieshavebeentriedand progress has beenmade, both theoretical and empirical. This introductionupdatesthestatusofthedebate.Itwaspreparedforasecondconferencethatwashosted by the IMFon the same topic in spring 2013 and as a springboard forfurtherdiscussion.

Afewobservationsonthescopeof theanalysis:ourcommentsfocusonthedesign of macroeconomic policy after the global economy emerges from thecrisis rather thanon currentpolicy choices, such as thedesignof exit policiesfromquantitativeeasingortheprosandconsofmoney-financedfiscalstimulus.Thetwosetsofissuesareobviouslyrelated,butourobjectiveistoanalyzesomegeneral principles that could be used to guide macroeconomic policy in thefutureratherthantosuggestspecificmeasurestobetakentoday.Wealsotakea

relativelynarrowviewofmacroeconomicpolicy,leavingoutanydiscussionofstructural reforms and financial regulation. Although the border betweenfinancialregulationandmacroprudentialpoliciesisfuzzy,weconcentrateonthecyclical componentof financial regulation rather thanon theoverall designofthefinancialarchitecture.

Thisintroductionisorganizedinthreemainsections:monetarypolicy,fiscalpolicy, and-whatmay be emerging as the third leg ofmacroeconomic policy-macroprudentialpolicies.

1.MonetaryPolicy

Themonetarypolicythemethatemergedfromthefirstconferenceonrethinkingmacropolicy,heldinMarch2011,wasthatcentralbankshadtomovefromanapproachbasedlargelyononetargetandoneinstrument(theinflationrateandthe policy rate, respectively) to an approach with more targets and moreinstruments.Twoyearslaterthechoiceofboththesetoftargetsandthesetofinstrumentsremainscontroversial.

A.ShouldCentralBanksExplicitlyTargetActivity?

Althoughthefocusofmonetarypolicydiscussionshasbeen,rightly,ontheroleof the financial system and its implications for policy, macroeconomicdevelopmentsduringthecrisisandafterhaveledtonewquestionsaboutanoldissue, the relation between inflation and output, with direct implications formonetarypolicy.

One of the arguments for the focus on inflation by central banks was the"divine coincidence" aphorism: the notion that, by keeping inflation stable,monetary policy would keep economic activity as close as possible (givenfrictions in the economy) to its potential. So, the argument went, even ifpolicymakers cared about keeping output at potential, they could best achievethis by focusing on inflation and keeping it stable. Although no central bankbelievedthatdivinecoincidenceheldexactly, it lookedlikeasufficientlygoodapproximation to justify a primary focus on inflation and to pursue inflation

targeting.

Since thecrisisbegan,however, the relationbetween inflationandoutput inadvanced economies has been substantially different fromwhat was observedbefore the crisis.With the large cumulative decline in output relative to trendandthesharpincreaseinunemployment,mosteconomistswouldhaveexpecteda fall in inflation, perhaps even the appearance of deflation. Yet in mostadvanced economies (including some experiencing severe contractions inactivity),inflationhasremainedclosetotherangeobservedbeforethecrisis.

Asamatteroflogic,therearetwointerpretationsofwhatishappening.Eitherpotentialoutputhasdeclinednearlyasmuchasactualoutput,sothattheoutputgap (the difference between potential and actual output) is in fact small, thusputting little pressure on inflation, or the output gap is still substantial but therelationbetweeninflationandtheoutputgaphaschangedinimportantways.

With regard to the first interpretation, it is possible that the crisis itself ledpotentialoutputtofall,orthatoutputbeforethecrisiswashigherthanpotentialoutput-for instance, if it was supported by unsustainable sectoral (housing)bubbles-sothattheactualoutputgapissmall.Thiscouldexplainwhyinflationhasremainedstable.Empirically,however, ithasbeendifficult toexplainwhythenaturalrateofunemploymentshouldbemuchhigherthanbeforethecrisis,orwhythecrisisshouldhave led toa largedecline inunderlyingproductivity.And although there is a fair amount of uncertainty around potential outputmeasures (especially in thewake of large shocks such as financial crises), bynearly all estimates, most advanced economies still suffer from a substantialoutputgap.

Thisleadstothesecondinterpretation.Indeed,convincingevidencesuggeststhattherelationbetweentheoutputgapandinflationhaschanged.Recentwork(e.g., theIMF's2013WorldEconomicOutlookreport)attributes thechangetothefollowingtwofactors.

The first factor is more stable inflation expectations, reflecting in part theincreasingcredibilityofmonetarypolicyduringthelasttwoorthreedecades.By

itself, this is awelcome development, and it explainswhy a large output gapnow leads to lower (but stable) inflation rather than to steadily decreasinginflation.

The second factor is aweaker relation (both inmagnitude and in statisticalsignificance)between theoutputgapand inflation foragivenexpected rateofinflation.This ismoreworrisomebecause it implies that fairly stable inflationmaybeconsistentwithlarge,undesirablevariationsintheoutputgap.

Looking forward, the main question for monetary policy is whether thisweakerrelationisaresultofthecrisisitself,andthuswillstrengthenagainwhenthe crisis comes to an end, or whether it reflects a longer-term trend. Thetentative evidence is that part of it may indeed reflect specific circumstancesrelatedtothecrisis-inparticular,thefactthatdownwardnominalwagerigiditiesbecomemorebindingwheninflationisverylow.Butpartoftheweakerrelationseemstoreflectasyetunidentifiedlonger-termtrends.(Theseactuallyseemtohave been present before the crisis; see the IMF's 2013 World EconomicOutlook.)Shouldtherelationremainweak,andthedivinecoincidencebecomeareally bad approximation, central banks would have to target activity moreexplicitlythantheyaredoingtoday.

B.ShouldCentralBanksTargetFinancialStability?

Thecrisishasmadeitclearthatinflationandoutputstabilityarenotenoughtoguaranteesustainedmacroeconomicstability.BeneaththecalmmacroeconomicsurfaceoftheGreatModeration(aperiodofreducedmacroeconomicvolatilityexperienced in theUnited States beginning in the 1980s), sectoral imbalancesandfinancialrisksweregrowing,andultimatelyledtothecrisis.Theseverityoftheensuingcrisisand the limitedeffectivenessofpolicyactionhaschallengedtheprecrisis"benignneglect"approachtobubbles.Andithasreignitedtheissueofwhethermonetarypolicy should include financial stability (proxiedby, say,measuresofleverage,creditaggregates,orassetprices)amongitstargets.

The policy rate is clearly not the ideal tool for dealing with the kind ofimbalances that led to the crisis. Its reach is too broad to be cost-effective.

Instead, a consensus is emerging that more-targeted macroprudential toolsshouldbeusedforthattask.

There are, however, important caveats. Macroprudential tools are new, andlittle is known about how effective they can be. They are exposed tocircumvention and subject to thorny political economy constraints. (more onthese tools below)Given these limitations, the issue ofwhether central banksshould use the policy rate to lean against bubbles hasmade a comeback (see,e.g.,Svensson2009;Mishkin2010;Bernanke2011;King2012).

Shouldcentralbankschoosetoleanagainstbubbles,anoldissueevidentbothinthe2008-2009crisisandinmanypreviousfinancialcrises-isthatbubblesarerarely identifiable with certainty in real time. This uncertainty suggests thatcentralbanksmaywanttoreacttolargeenoughmovementsinsomeassetpriceswithout having to decide whether such movements reflect changes infundamentalsorbubbles.Inotherwords,givenwhatwehavelearnedaboutthecosts of inaction, higher type I errors (assuming that it is a bubble and actingaccordingly, when in fact the increase reflects changes in fundamentals) inexchangeforlowertypeIIerrors(assumingtheincreasereflectsfundamentals,wheninfactitisabubble)maywellbejustified.However,shouldthatroadbetaken,settingappropriatethresholdswillnotbeeasy.Onepossibilitywouldbetofocusoncertaintypesofasset-pricebooms,forinstancethosefundedthroughbankcredit,whichhaveprovenparticularlydangerous.

C.ShouldCentralBanksCareabouttheExchangeRate?

The crisis has shown once again that international capital flows can be veryvolatile. This volatility has not generally been a major problem in advancedeconomies (although the flow reversals in the euro area and the drying out ofdollar liquidity in theEuropean banking systemduring the early stages of thecrisisareareminderthatvulnerabilitiesexistthereaswell).However,shallowerfinancialmarkets,greateropennessandrelianceonforeign-denominatedassets,and less diversified real economies make emerging markets significantlyvulnerabletoswingsincapitalflows.

The volatility of capital flows can have adverse effects on macroeconomicstability, both directly (through effects on the current account and aggregatedemand) and indirectly (through effects on domestic balance sheets and thusfinancial stability).When the exchange rate strengthens on the back of stronginflows,thetradedgoodssectorlosescompetitiveness,potentiallyleadingtoanallocationofcapitalandlaborthatmaybecostlytoundoifcapitalflowsandtheexchange rate swing back. Capital inflows can also lead to balance sheetstructuresthatarevulnerabletoreversalstotheextentthattheinflowspromotecreditbooms(andhenceleverage)andincreasetheuseofforeign-denominatedliabilities. (There is ample evidence, for instance, that the credit booms andwidespreadrelianceonforeigncurrencyborrowinginEasternEuropeinthefirstdecadeof the2000swasassociatedwithstrongcapital inflows[Dell'Aricciaetal.2012]).

The problemswith capital flow volatility have led to a reassessment of thepotentialroleforcapitalcontrols(whichtheIMFcalls"capitalflowmanagementtools").But, justas in thecaseofmacroprudential toolsandfinancialstability,capital controls may not work well enough, raising the issue of whethermonetarypolicyshouldhaveanadditionalobjective(Ostry,Ghosh,andChamon2012).

Couldcentralbankshavetwotargets,theinflationrateandtheexchangerate,and two instruments, the policy rate and foreign exchange intervention?(Inflationtargetingcentralbankshavearguedthat theycareabout theexchangerate to the extent that it affects inflation, but it is worth asking whether thisshouldbe theonlyeffectof theexchangerate theyought toconsider.)Addingexchangeratestothemixraisesissuesofbothfeasibilityanddesirability.

The answer to the feasibility question is probably no for economies withhighlyintegratedfinancialmarkets(andalmostcertainlynoforsmall,veryopen,advanced economies-say, New Zealand). Under those conditions, sterilizedintervention isunlikely tobeeffectivebecausecapital flowsreact immediatelyto interest rate differentials.But the answer is probably yes (and the evidencepointsinthisdirection)foreconomieswithgreaterfinancialfrictionsandmorehighlysegmentedmarkets.Underthosecircumstances,onecouldthusconsider

anextendedinflationtargetingframework,withthepolicyrateaimedatinflation,andforeignexchangeinterventionaimedattheexchangerate.

Butwhataboutdesirability?Theconsensusthathasemergedregardingtheuseand the limitations of capital controls is directly relevant. The issues andconclusionsareverymuchthesame.Interventionistypicallynotdesirablewhenitisaimedatresistingatrendappreciationdrivenbysteadycapitalflowsratherthan by temporary swings (that is, when the movement in the exchange ratereflectsachangeinunderlyingfundamentalsratherthan,forexample,temporaryswingsbetweenriskoffandriskon).Anditmayraiseissuesfromamultilateralperspective(formore,seeOstry,Ghosh,andKorinek2012).

D.HowShouldCentralBanksDealwiththeZeroBound?

Whatmaybemost striking about the crisis is theway inwhich central bankshave experimented with unconventional policies, from quantitative easing, totargeted easing, to new forms of liquidity provision. Will these instrumentsbecomepartofthestandardtoolkit,oraretheyspecifictothecrisis?Toanswerthisquestion,oneneedstodistinguishbetweentwocharacteristicsofthecrisis.

Thefirstistheliquiditytrap,whichconstrainstheuseofthepolicyrate.Thesecond is the segmentation of some financialmarkets or financial institutions.Althoughbothcharacteristicshaveplayedacentral role indeterminingpolicy,they are conceptually separate. One can think of sufficiently adverse butnonfinancial shocks such that central bankswould like to decrease the policyrate further but find themselves constrained by the zero bound. And one canthink instead of financial shocks that trigger segmentation in some financialmarketswhile the policy rate is still positive.We consider the implications ofeachinturn.

Thecrisishasshownthateconomiescanhitthezerolowerboundonnominalinterest rates and lose their ability to use their primary instrument, the policyrate,withhigherprobabilitythanwasearlierbelieved.Thisraisestwoquestions.The first question is what steps can be taken to minimize the probability offallingintoliquiditytrapsinthefuture.Wewillnotelaborateonthediscussion

givenbyBlanchard,Dell'Ariccia,andMauro(2010)regardingtheoptimallevelof inflation in this context, although the argument in that paper and thecounterarguments brought up in the ensuing debate still deserve a nonideological discussion both in academia and in policy forums (see, e.g., Ball2013).

The secondquestion iswhat to do in the liquidity trap.When the crisis hit,most central banks reacted by cutting interest rates aggressively. In severalcases,interestratesrapidlyhitthezerolowerbound.Centralbanksthenmovedto adopt unconventional policies, which have taken many forms, with analphabetsoupofacronyms.Itisusefultodistinguishbetweentargetedeasing(amoreaccuratenamethancrediteasing)measures, that is,purchasesofspecificfinancial assetswithoutachange in themoneysupply, andquantitativeeasingmeasures, which are not sterilized and thus lead to an increase in the moneysupply.

Availableempiricalevidencesuggeststhatsometargetedeasingpolicieshavehadasubstantialimpactonthepricesoftheassetsacquiredbythecentralbank.Much of the impact, however, seems to have come from the unusualsegmentation of financial markets associated with this crisis, as seen, forexample, in the case of themortgage-backed securitiesmarkets in theUnitedStatesin2008and2009(seeGagnonetal.2011).Althoughassetswithdifferentrisk characteristics are always imperfect substitutes and thus relative demandalwaysmatters,theabilityofthecentralbanktoaffectrelativereturnsislikelytobemuchmorelimitedinnormaltimesthanitwasduringthecrisis.

Quantitative easing can be thought of as the combination of targeted easing(thepurchaseofsomeassets,suchaslong-termTreasurybonds,financedbythesaleofshort-termassets)andaconventionalmonetaryexpansion(thepurchaseof short-termassetswith centralbankmoney).Thequestion iswhether, at thezerobound,themonetaryexpansioncomponenthasaneffectperse.TheissueisparticularlyclearinJapan,wherethecentralbankhasannounceditsintentiontodoublethemonetarybase.Ifithasaneffect,ithastobethroughexpectationsofeither low future nominal rates or higher future inflation. (In the Alice inWonderland,upside-downworldof theliquidity trap,higherexpectedinflation

is welcome because it is the only way to obtain a decrease in expected realrates.)Empirical evidence ismixed.Theevidence is abit stronger for anothermeasure with a similar intent, namely, "forward guidance." Announcementsconsistentwithforwardguidance(suchastheintentionorcommitmenttokeepshort-term rates low for a specific period, or for as long as some economicconditions prevail) appear to have had a significant and economically sizableimpact on long-term rates both in Canada and in the United States. Similarannouncements, however, appear to have been less effective for Sweden'sRiksbank(Woodford2012).Withregardtofuturemonetarypolicy,awayfromthezerobound,forwardguidancemaywellbeheretostay.

Thecrisishasalsoledtonewdiscussionsofanumberofoldideas,includingashift toprice-level targetingornominalGDPtargeting.Support for these rulesmaybepartlyopportunistic:acommonfeatureof level-basedapproaches (i.e.,rulesthattargetthepricelevelratherthantheinflationrate,ornominalincomeratherthannominalincomegrowth)isthat,atthisjuncture,theywouldallowforhigher inflation rateswithout undermining central bank credibility in the longrun.Apotential lossof credibilityhasbeen amajor concern for central banksthroughoutthecrisis,asevidencedbythereaffirmationbycentralbanksoftheircommitment to remain vigilant against inflation with every round ofunconventional policies. But these level-dependent rules have severalshortcomings.Animportantoneisthattemporarypriceshocksarenottreatedasbygonesandhavetobeabsorbedthroughinflation,orworse,deflation.

E.ToWhomShouldCentralBanksProvideLiquidity?

Whensomeinvestorsarehighlyspecialized(havestrong"preferredhabitats,"touseanoldexpression)and,forsomereason,reducetheirdemand,outsidersmaynothavethespecializedknowledgeneededtoassesswhetherthelackofdemandcomesfromhigherriskorfromthefactthattheusualbuyersareunabletobuy.Outsidersmay thendecide to stayout.When this happens,market pricesmaycollapse, or some borrowers may lose funding. Illiquidity may then lead toinsolvency. Multiple equilibria may also arise, with the expectation ofinsolvencyleadingtohighinterestratesandbecomingself-fulfilling.

Fromitsearlystages,thecrisisshowedthattheclassicalmultipleequilibriumframework, which provided a rationale for providing banks with depositinsurance and access to a lender of last resort, now also applied towholesalefundingandnonbank intermediaries.Thesituation inEurope later showed thatthe same framework could also extend to sovereigns, even in advancedeconomies. Indeed, sovereigns are even more exposed than financialintermediariestoliquidityproblemsbecausetheirassetsconsistmostlyoffuturetaxrevenues,whicharehardtocollateralize.Theexpectationthatotherinvestorsmaynotrolloverdebtinthefuturemightleadcurrentinvestorstonotwanttorollover,leadingtoaliquiditycrisis.

Centralbanksendedupprovidingliquiditynotonlytobanksbutalsotonon-deposit-taking institutions, and (directly and indirectly) to sovereigns. From atheoreticalstandpoint,thelogicislargelythesame.Nevertheless,theextensiontononbanksraisesanumberofissues.

First,justaswithbanks,theissueofdistinguishingilliquidityfrominsolvencyarises. But for nonbanks this issue happens in the context of potentiallyunregulated entities about which central banks possess limited information.Second, again as for banks, is the issue of moral hazard. The promise (orexpectation) of liquidity provision will induce the accumulation of even lessliquid portfolios beforehand, thereby increasing the risk of a liquidity crisis(Farhi and Tirole 2012). The problem is exacerbated in the case of indirectsupport(throughmarketpurchasesofsovereignbonds,e.g.)because,unlikewithdirect support to banks, it is difficult (or impossible) to administer anypunishment.Haircuts(fordiscountwindowaccess)andconditionality(fordirectpurchases) can partly allay but not eliminate these concerns.And haircuts runcounter to the notion of providing the "unlimited liquidity, no matter whathappens"necessarytoeliminatetheriskofarun.Duringasystemiccrisis,thesearesecondordershortcomingsrelativetotheneedtostabilizetheeconomy.Butthecaseforinterventionappearshardertomakeduringtranquiltimes.

II.FiscalPolicy

Early in thecrisis,withmonetarypolicy facing the liquidity trapand financial

intermediation still in limbo, governments turned to fiscal stimulus to sustaindemand and avoid what they felt could become another Great Depression.However,whentheacutedangerappearedtohavesubsided,governmentsfoundthemselveswithmuchhigherlevelsofpublicdebt(notsomuchbecauseofthefiscal stimulus but because of the large decline in revenues caused by therecession).Since then, thefocusof fiscalpolicydiscussionshasbeenonfiscalconsolidation.

Attheearlierconference,weconvergedontwomainconclusions.First,whatappearedtobesafelevelsofpublicdebtbeforethecrisiswereinfactnotsosafe.Second,a strongcaseemerged for revisiting theprecrisisconsensus that fiscalpolicyhadalimitedcyclicalroletoplay.

Thequestionsaremuchthesametoday,withafewtwists.Inlightofthehighdebtlevels,asignificantpolicyissuethatwillremainwithusbeyondthecrisisisthatoftheproperspeedoffiscalconsolidation.Theanswerdependsontwomainfactors.First,howharmfulordangerousarecurrentdebt levels?Thecrisishasadded one more issue to the usual list of the adverse effects of high debt:multipleequilibriainwhichviciouscyclesofhighinterestrates,lowgrowth,anda rising probability of default may lead to a fiscal crisis. Second, and to theextentthatfiscalconsolidationisnecessary,whatareitseffectsongrowthintheshort run,given the stateof theeconomyand thepathandcompositionof thefiscaladjustment?

Wetakeupeachoftheseissuesinturn.

A.WhatAretheDangersofHighPublicDebt?

At thestartof thecrisis, themediandebt-to-GDPratio inadvancedeconomieswasabout60percent.Thisratiowasinlinewiththelevelconsideredprudentforadvancedeconomies,asreflected,forexample,intheEuropeanUnion'sStabilityandGrowthPact.(Somewhatironically,theprudentlevelforemergingmarketswasconsideredtobelower,about40percent.Theactualmedianratiowaslessthan40percent,whichhasgiventhesecountriesmoreroomforcountercyclical

fiscalpolicythaninpreviouscrises.)

Bytheendof2012,themediandebt-to-GDPratioinadvancedeconomieswasclose to 100 percent and was still increasing. For the most part, the increasestemmedfromthesharpfall inrevenuescausedbythecrisis itself.Toalesserextent, it was attributable to the fiscal stimulus undertaken early in the crisis.Andforsomecountries,itwasduetotherealizationofcontingentliabilities(seeIMF2012a,box2). In Irelandand Iceland, forexample, theneed to rescueanoversizedbankingsystemledtounexpectedincreasesintheirdebtratiosof25and 43 percentage points, respectively. In Portugal, to take a lesswell-knownexample, as the crisis progressed, state-owned enterprises incurred losses and,under Eurostat rules, had to be included within the general government, thedeficit and debts of which increased as a result.Moreover, guarantees startedbeing called on public-private partnerships (which were more sizable than inother countries), thereby adding to the general government's burden. Betweenthoseissuesandfinancialsectorinterventions,theoverallresultwasanincreaseinthePortuguesedebtratioofabout15percentagepoints.

The lessons are clear. Macroeconomic shocks and the budget deficits theyinducecanbesizable-largerthanwasconsideredpossiblebeforethecrisis.Andthe ratio of official debt to GDP can hide significant contingent liabilities,unknown not only to investors but sometimes also to the government itself(Irwin2012).Thissuggeststheneedforbothamorecomprehensiveapproachtomeasuresofpublicdebtandlowervaluesforwhatconstitutes"prudent"officialdebt-to-GDPratios.Unfortunately,giventheextent towhichactualratioshaveincreased,itwilltakealongtimetoattainthoseprudentratiosagain.

Thecostsofhighpublicdebt,fromhigherequilibriumrealinterestratestothedistortionsassociatedwiththetaxesneededtoservicethedebt,havelongbeenrecognized.Thecrisisbroughttolightanotherpotentialcost:theriskofmultipleequilibriaassociatedwithhighlevelsofdebt.Ifinvestors,worriedaboutahigherriskofdefault,requirehigherriskpremiumsandthushigherinterestrates,theymakeitmoredifficultforgovernmentstoservicethedebt,therebyincreasingtheriskofdefaultandpotentiallymakingtheirworriesself-fulfilling.

Inprinciple, suchmultipleequilibriacanexist evenat low levelsofdebt.Averyhighinterestratecanmakeevenalowlevelofdebtunsustainableandthusbeself-fulfilling.Butmultipleequilibriaaremorelikelywhendebt ishigh,forthen even a small increase in the interest rate canmove the government fromsolvencytoinsolvency.Theyarealsomorelikelywhenthematurityofthedebtisshortandrolloverneedsaregreater:ifmostofthedebthastoberolledoversoon, it ismore likely that current investorswillworry about future rollovers,leadingthemtobereluctanttorollovertoday.

Also in principle, central banks can eliminate the bad equilibrium byproviding-or simply by committing to provide-liquidity to the government ifneeded. However, providing this liquidity is not straightforward. Theinterventionmayneed to bevery large.And in light of the usual difficulty ofdistinguishingbetween illiquidityand insolvencyand the fact that the state, asdistinctfrombanks,cannotprovidecollateral,theriskstothecentralbankmaybeconsiderable.

The experience of the crisis suggests that the issue ofmultiple equilibria isrelevant.TheevolutionofSpanishandItaliansovereignbondyieldscanbeseeninthislight,withtheEuropeanCentralBank's(ECB's)commitmenttointervenein their sovereign bondmarkets having reduced the risk of a bad equilibrium.Someothereuroareamembers,suchasBelgium,havebenefitedfromlowratesdespite still high levels of debt and political challenges; how much of thedifferencebetween,say,BelgiumandItalycanbeexplainedbyfundamentalsorbymultiple equilibria is anopenquestion.The relativelybenignperceptionofboth theUnited States and Japanmay be seen as an example in the oppositedirection.Despitehighlevelsofdebt,particularlyinJapan,bothcountrieshavebeenperceivedsofaras"safehavens"andhavebenefitedfromverylowrates,containingtheirdebt-serviceburdens.However,theissueisthestrengthoftheirsafehavenstatusandwhetherthesituationmightchangequickly,leadingtobadequilibriumoutcomesinthesecountriestoo.

B.HowtoDealwiththeRiskofFiscalDominance?

In light of the magnitude of the required fiscal consolidation in so many

advancedeconomies, the issueofwhether to reduce the real valueof thedebtthroughdebtrestructuringorinflationisunlikelytogoaway.

Weshall limitourselves to twobrief remarksondebt restructuring.First, atleastinthecurrentinternationalfinancialarchitecture,debtrestructuringremainsacostlyandcumbersomeprocess.(Howtoimprovethissituationwillcontinuetobeanimportanttopicforresearchandpolicyanalysis.)Second,incontrasttotheemergingmarketexperiencesofthepast,asizableshareofthedebtinmostadvanced economies is held by domestic residents (more than 90 percent inJapan),often financial intermediaries, orby residentsofneighboringorhighlyconnectedcountries(includingthroughthefinancialsystem).Thus,thescopefordebtrestructuringisverylimited.Andinanycaseitwouldcallforextremecareto minimize potentially disruptive redistribution of wealth between domesticbondholdersandtaxpayers,andstrongadverseeffectsonthefinancialsystem.

Against that background, governments facing the need for difficult fiscaladjustment might well put pressure on central banks to help limit borrowingcosts,whichraises the issueoffiscaldominance.Inprinciple,monetarypolicycanhelpreducethepublicdebtburdeninanumberofways.Centralbankscanslowdowntheexitfromquantitativeeasingpoliciesandkeepsovereignbondsontheirbookslonger.Theycanalsodelaytheincreaseinnominalinterestrateswarranted bymacroeconomic conditions and let inflation increase, leading, onbothcounts, to low real interest rates for amoreprolongedperiod thanwouldotherwisebeoptimal.

Indeed, historically, debt has often been reduced through rapid inflation;extremeexamplesincludethewell-knownepisodesofhyperinflationthatwipedout debt in the aftermath ofmajorwars (e.g.,Germany, Japan). Less extremecaseshaverecentlyattractedrenewedattention,notablytheUnitedStatesinthesecondhalf of the1940s,when inflation resulted in significantlynegative realinterestratesand,overtime,lowerdebtratios(seeReinhartandSbrancia2011,whosuggestthatareturntofinancialrepressionisapotentialconcern).

How much difference could such monetary policies make? The answerdepends largelyonhow longcentralbankscanmaintain lowor evennegative

real interestrates.Undertheassumptionthatnominal interestratesreflectone-for-oneincreasesininflation,sothattherealinterestrateremainsconstant(afullandimmediateFishereffectapplyingtoallnewlyissuedorrolled-overdebt),thedecreasedependsontheabilitytoerodethevalueofoutstanding(long-maturity)nominaldebt,andisrathersmall.IMFstaffsimulationssuggestthat,fortheG7economies,ifinflationweretoincreasefromthecurrentaverageprojectedpaceof less than2percent to,say,6percent, thenetdebt ratiowoulddecline,afterfiveyears,byabout10percentofGDPonaverage(Akitoby,Komatsuzaki,andBinder,forthcoming).Theeffectwouldbelargerifcentralbankscouldmaintainlowerreal interest rates forsome time. (It issometimesargued that thiswouldrequire financial repression, i.e., the ability to forcebanks toholdgovernmentbonds.This seems incorrect: as the current evidence shows, central banks canmaintain negative real interest rates for some time if they want to. But thesenegative rates may lead to overheating and inflation. They may also induceinvestorstoshifttoforeignassets,leadingtodepreciationandfurtherinflation.However,ifcentralbanksaccepttheseinflationconsequences,theycanmaintainlowerrealinterestratesforsometime,evenabsentfinancialrepression.)

In short, if regular fiscal consolidation, through higher revenues or lowerspending,provedinfeasible,lowornegativerealinterestratescould,inprincipleandwithin limits, helpmaintain debt sustainability.However, this pathwouldhavesizablecosts:increasesininflationandreductionsinrealinterestratesare,ineffect,asmoother,lessvisibleversionofdebtrestructuring,withsomeoftheburdenofadjustmentshiftedfromtaxpayerstobondholders,andwouldthusfacesimilarlysignificantdistributional,social,andpoliticalissues.

In lightof theseconsiderations, it isessential thatmonetarypolicydecisionscontinue to be under the sole purview of the central bank, unencumbered bypolitical interference.Thecentralbank, in turn,shouldbase itsdecisionon theway the debt situation and fiscal adjustment (or lack thereof) would impactinflation, output, and financial stability. Indeed, central bank purchases ofgovernment bonds during the crisis have occurred against the background oflarge output gaps and often as part of an effort to avoid deflation or a self-fulfillingdebtcrisis.Moregenerally, thecentralbankshouldbemindfuloftherisk that such policy could be viewed as slipping into fiscal dominance,

particularly given the difficulties of assessing the effects on output of variouspossible strategies to keep public debt in check. The risk of fiscal dominanceseemsrelativelylimitedintheeuroarea,wherenosinglegovernmentcanforcetheECBtochangeitsmonetarypolicy.Itismorerelevantelsewhere,andmayremainanissueforyearstocome.

C.AtWhatRateShouldPublicDebtBeReduced?

InlightoftheneedtodecreasetheratioofpublicdebttoGDP,thefiscalpolicydebate has focused on the optimal speed and the modalities of fiscalconsolidation.Manyof the issuesconsolidationraisesarerelevantnotonlyfornow,butmoregenerallyforfiscalpolicyinthefuture.

Identifyingthedynamiceffectsoffiscalpolicyonoutputisdifficult.Itsuffersfromidentificationproblems,andtheeffectsarelikelytodifferdependingonthestateoftheeconomy,thecompositionofthefiscaladjustment,thetemporaryorpermanentnatureofthemeasures,andtheresponseofmonetarypolicy.

Largely as a result of these difficulties, empirical estimates of fiscalmultipliers ranged widely before the crisis (e.g., see Spilimbergo, Symansky,and Schindler 2009). Early in the crisis, some researchers and policymakersargued thatpositive confidenceeffects coulddominate the adversemechanicaleffects of cuts in spending or increases in revenues and lead to "expansionaryfiscal consolidations." Others argued that, in a situation of impaired financialintermediationandthustighterborrowingconstraintsforfirmsandhouseholds,together with the fact that monetary policy was facing the liquidity trap,multiplierswereinsteadlikelytobelargerthaninmorenormaltimes.

Thewide rangeof fiscal policy responses to the crisis and its aftermathhasstimulatednew research (see, e.g., the articles inAmericanEconomic Journal:EconomicPolicy 4, no. 2, 2012).Although still a subject of some debate, theevidence shows that the multipliers have been larger than in normal times,especially at the start of the crisis (Blanchard and Leigh 2013), with littleevidenceofconfidenceeffects(Perotti2011).Beyondthisconclusion,however,manyquestionsremainunanswered-inparticular,thedifferentialeffects,ifany,

ofconsolidationsbasedonspendingcutsratherthanonrevenueincreases.

Underlyingthedebateaboutmultipliershasbeenthequestionoftheoptimalspeedoffiscalconsolidation(withsomeintheUnitedStatesactuallyarguingforfurther fiscal stimulus). In reality, formany countries severely affected by thecrisis,thespeedofconsolidationhasnotbeenamatteroffreechoice;rather,ithas largely been imposed on them bymarket pressures. Indeed, cross-countryvariation in the speed of adjustment has been explained in good part bydifferencesinsovereignbondyields.

For countries that have some fiscal room, conceptually, the issue is how totradeofffirstmomentsforsecondmoments,thatis,howtotradeofftheadverseshort-runeffectsongrowthoffasterconsolidationagainst thedecreaseinriskscomingfromlowerdebtlevelsovertime.(Theargumentthatfiscalstimuluscanmorethanpayforitself,andthusdecreasedebtlevels,seemstobeasweakastheearlierargument that fiscalconsolidationcould increaseoutput in theshortrun). However, because of the relevance of multiple equilibria, and our poorunderstandingofthebehaviorofinvestorsinthiscontext,theserisksaredifficultto assess with any degree of precision. Thus, while fiscal consolidation isneeded,thespeedatwhichitshouldtakeplacewillcontinuetobethesubjectofstrongdisagreement.

Within this context, a few broad principles should still apply, as werearticulated invarious IMFpublications (Cottarelli andVinals 2009;Blanchardand Cottarelli 2010; IMF 2010;Mauro 2011; IMFWorld EconomicOutlook,variousissues;IMFFiscalMonitor,variousissues).Inlightofthedistancetobecovered before debt is down to prudent levels and of the need to reassureinvestorsandthepublicatlargeaboutthesustainabilityofpublicfinances,fiscalconsolidation should be embedded in a credible mediumterm plan. The planshould include the early introductionof some reforms-suchas increases in theretirementagethathavetheadvantageoftacklingthemajorpressuresfromage-relatedexpenditureswhilenotreducingaggregatedemandinthenearterm.

The need to control debt has also attracted renewed interest in fiscal rules.Manycountries,especiallyintheeuroarea,haveintroducedmediumtermfiscal

adjustment plans and have strengthened their commitment to fiscal rules. Forexample,Germany,Italy,andSpainhaverecentlyamendedtheirconstitutionstoenshrineacommitment to reducing thestructuraldeficit tozeroornearlyzerobyspecificdates,allwithinafewyears.Moregenerally,manynewfiscalruleshave been adopted and existing ones strengthened in response to the crisis, inboth advanced economies and emerging market economies (Schaechter et al.2012).Theevidenceonmediumtermfiscaladjustmentplansshowsthatawiderangeofshocks-especiallythosetoeconomicgrowth-havethepotentialtoderailimplementation (Mauro 2011; Mauro and Villafuerte 2013). This potentialhighlights theimportanceofexplicitly includingmechanismstodealwithsuchshocks, thus permitting some flexibility while credibly preserving themediumtermconsolidationobjectives.Examplesofhelpfulmechanismsincludemultiyear spending limits; the exclusion of items that are cyclical (e,g,,unemployment benefits), nondiscretionary (e.g., interest payments), or fiscallyneutral (e.g.,EU-fundedprojects);or theuseofcyclicallyadjusted targets thatlettheautomaticstabilizersoperateinresponsetocyclicalfluctuations.

D.CanWeDoBetterThanAutomaticStabilizers?

Other things equal, if the concern is output growth in the short run, weakerprivate demand (whether domestic or foreign) should call for slower fiscalconsolidation. This argument has led several countries to shift from nominalfiscaltargetstostructuraltargets,soastoletautomaticstabilizersfunction.

Thisleadstoaquestionraisedinourearlierpaper.Althoughlettingautomaticstabilizersworkisbetterthannotdoingso,stabilizersareunlikelytodelivertheoptimalcyclicalfiscalpolicyresponse.First, theusualargument that theeffectofautomaticstabilizersondebtcancelsoutovertimeappliesonlytotheextentthatmovementsinoutputaretemporary.Thismaynotbethecase.Asdiscussedin section III, it is not clear, for example, howmuch of the recent decline inoutput(relativetotrend)istemporaryorpermanent.Second,theoverallstrengthofautomaticstabilizersvariesfromcountrytocountryanddependsonsocietalchoices-on the size of the government, as well as on tax and expenditurestructures-that weremade on the basis of objectives other than cyclical fiscalpolicy.Thus,thestrengthoftheautomaticstabilizerscouldbeinsufficient,orit

couldbeexcessive.

Thus, our earlier paper asked, why not design better stabilizers (Blanchard,dell'Ariccia, and Mauro 2010)? For instance, for countries in which existingautomaticstabilizerswereconsideredtooweak,proposalsforautomaticchangesin tax or expenditure policies are appealing. Examples include cyclicalinvestment taxcredits,orprelegislatedtaxcuts thatwouldbecomeeffectiveif,say, job creation fell belowa certain threshold for a fewconsecutivequarters.Perhapsbecause thepolicy focushasbeenonconsolidation rather thanon theactive use of fiscal policy, there has been, as far aswe know, little analyticalexploration (an exception is McKay and Reis 2012) and essentially nooperationaluptakeofsuchmechanisms.

III.MacroprudentialInstruments

Oneoftheunambiguouslessonsfromthecrisisisthatdangerousimbalancescanbuild beneath a seemingly tranquil macroeconomic surface. Inflation can bestable,outputcanappeartobeatpotential,butthingsmaystillnotbequiteright.Sectoralboomsmayleadtoanunsustainablecompositionofoutput-forexample,toomuch housing investment.Or financial risksmay build up because of theway real activity is funded (e.g., excessively leveraged financial institutions,excess household indebtedness, excess maturity mismatches in the bankingsystem, recourse to offbalance-sheet products entailing large tail risks).Critically, the effects of these imbalances can be highly nonlinear. Long andgradualbuildupscanbefollowedbyabruptandsharpbusts,withmajorwelfareconsequences.

Beyondadesirablestrengtheningofprudentialsupervisionoverthefinancialsector,whatelsecanbedone topreventsuchproblemsfromreoccurringor tocushion their blow? Monetary and fiscal policies are not the best tools foraddressingtheseimbalances(atleastasafirstlineofdefense).Monetarypolicyhas toobroad a reach todeal cost-effectivelywith sectoral boomsor financialrisks.Fiscalmeasurescanbemoretargeted,buttimelagsandpoliticaleconomyproblems limit their usefulness. These shortcomings have led to increasinginterest in more targeted "macroprudential instruments" (see Borio and Shim

2007foranearlydiscussion).Thepotentialuseoftheseinstrumentswasamajorthemeofourfirstconference,andithasbeenanactivefieldofresearchsincethestart of the crisis (e.g., ECB 2012). Now that some of these tools have beenadoptedinpractice,webetterunderstandtheireffectsandtheirlimitations.Butwe are still a long way from knowing how to use them reliably. Empiricalevidenceontheeffectivenessofthesemeasuresisscant,andthewaytheyworkand interact with other policies is likely to depend on a country's specificfinancialsectorstructureandinstitutions.

Among the conceptual issues that need to be solved are the articulationsbetween macroprudential and microprudential regulations, and betweenmacroprudentialpoliciesandmonetarypolicy.Wetaketheminturn.

A.HowtoCombineMacroprudentialPolicyandMicroprudentialRegulation?

Traditional microprudential regulation is partial equilibrium in nature. As aresult,itdoesnotsufficientlytakeintoaccounttheinteractionsamongfinancialinstitutions and between the financial sector and the real economy. The samebank balance sheet can have very different implications for systemic riskdependingonthebalancesheets(andtheinterconnections)ofotherinstitutionsandthestateoftheeconomyasawhole.Thus,prudentialregulationhastoaddasystemic and macro dimension to its traditional institution-based focus.Regulatory ratios must reflect risk not in isolation but in the context of theinterconnections in the financial sector, and must also reflect the state of theeconomy.

Theseconsiderationssuggestthatmicro-andmacroprudentialfunctionsshouldbeunder the same roof.However,political economyconsiderationsmay favorkeeping the two functions under two different agencies. Several aspects ofregulation(e.g.,thedegreeofbankcompetition,policiestofostercreditaccess,orthosedeterminingforeignbankparticipation)maybepoliticallytoodifficultto delegate to an independent agency. On the contrary, the macroprudentialfunction ismore akin tomonetary policy (with some caveats outlined below):unpopulartaskssuchasleaningagainstthewindduringacreditboomarelikelybest performed by an independent agency. If that is the case, an alternative

design could have the macroprudential authority in charge of the cyclicalmanagement of certain prudential measures, leaving the rest to themicroprudential regulator. (This is the approach followed in the UnitedKingdom,wheretheFinancialPolicyCommitteeoftheBankofEnglandwillbeabletovarythecapitalratiostobeappliedbymicroprudentialregulators.)

B.WhatMacroprudentialToolsDoWeHave,andHowDoTheyWork?

Onecanthinkofmacroprudentialtoolsasfallingroughlyintothreecategories:(1) tools seeking to influence lenders' behavior, such as cyclical capitalrequirements, leverage ratios, or dynamic provisioning; (2) tools focusing onborrowers'behavior,suchasceilingsonloan-to-valueratios(LTVs)orondebt-to-incomeratios(DTIs);and(3)capitalflowmanagementtools.

CyclicalCapitalRatiosandDynamicProvisioningThelogicofcyclicalcapitalratiorequirementsissimple:theyforcebankstoholdmorecapitalingoodtimes(especiallyduringbooms)soas tobuildbuffersagainst losses inbad times. Inprinciple, cyclical requirements can smooth a boom or limit credit growthbeforehand, aswell as limit the adverse effects of a bust afterward. Dynamicprovisioning can do the same, by forcing banks to build an extra buffer ofprovisionsingoodtimestohelpcopewithlossesifandwhenbadtimescome.

Inpractice,however, implementation isnot soeasy.First is the issueof theregulatory perimeter. Requirements imposed on banks may be circumventedthroughrecoursetononbankintermediaries,foreignbanks,andoffbalance-sheetactivities. Regulators might find themselves incrementally extending theregulatoryperimeterasmarketparticipantsdeviseevermoreinnovativewaystocircumventit.Secondisthepracticalquestionofwhatmeasuresthecyclicalityof requirements should be based on: the economic cycle, credit growth (assuggested underBasel III), asset-price dynamics (typically real estate)? Third,procyclicality isnoteffective ifbanksholdcapitalwell inexcessofregulatoryminimums(asoftenhappensduringbooms).Finally, timeconsistencyis likelytobeanissue:regulatorsmayfinditpoliticallydifficulttoallowbankstoreduceriskweightsduringabust(whenborrowersbecomelesscreditworthyandbankbalance sheets aremore fragile). In the past, regulators have achieved this, to

someextent,throughinformalforbearance.Amoretransparentapproachmaybemore difficult to sell to the public (recall the outcry against excessivelyleveragedbanksinthewakeofthecrisis).Thiscallsforarules-basedapproachandan independentpolicymaker. (However,given theproblems justdescribedand the political economy issues discussed in a later paragraph, rules-basedapproachespresenttheirowndifficulties.)

Dothesetoolswork?Evidenceismixed(seeSaurina2009;Croweetal.2011;Dell'Aricciaetal.2012).Tightercapitalrequirementsanddynamicprovisioninghave typically not stopped credit and real estate booms. But in a number ofcases,theyappeartohavecurbedthegrowthofparticulargroupsofloans(suchas foreign exchange-denominated loans), suggesting that these episodeswouldhavebeenevenmorepronouncedhadactionnotbeentaken.Inaddition,insomecases,thesemeasuresprovidedforlargerbuffersagainstbanklossesandhelpedtocontainthefiscalcostsofthecrisis(Saurina2009).

Loan-to-Value and Debt-to-Income Ratios Limits on LTV andDTI ratios areaimedatpreventingthebuildupofvulnerabilitiesontheborrower'sside.Afterabust, they can potentially reduce bankruptcies and foreclosures, leading tosmallermacroeconomicbusts.

Again, implementation is challenging. First, these measures are difficult toapply beyond the household sector. Second, attempts to circumvent themmayentail significantcosts. Inparticular, theymay result in liability structures thatcomplicatedebtresolutionduringbusts(e.g.,LTVlimitsmayleadtowidespreaduse of second lien mortgages, which become a major obstacle to debtrestructuringifabustoccurs).Circumventionmayinvolveashiftingofrisksnotonlyacrossmortgageloanproductsbutalsotooutsidetheregulatoryperimeterthrough expansion of credit by nonbanks, less-regulated financial institutions,and foreign banks (whichmay result in increased currencymismatches as theproportionofforeignexchange-denominatedloansrises).Undesiredsideeffectscan also occur to the extent that housing wealth is used as collateral incommercialloans(e.g.,bysmall-businessowners).

However, the limited existing empirical evidence suggests that these are

promisingmeasures.For instance,duringepisodesofquickly rising real estateprices,LTVandDTIlimitsappeartoreducetheincidenceofcreditboomsandtodecreasetheprobabilityoffinancialdistressandbelowpargrowthfollowingtheboom(seeCroweetal.2011;Dell'Aricciaetal.2012).

Capital Controls Capital controls (which the IMF refers to as "capital flowmanagement tools") are aimed at risks coming from volatile capital flows.Although they have a long history, their use has been controversial. In recentyearstheIMFhasarguedthat,ifmacropoliciesareappropriate,andiftheflowsarehavinganadverseimpactonfinancialormacroeconomicstability,theuseofthese tools can be appropriate, typically in combination with othermacroprudentialtools(Ostryetal.2010;IMF2012b).Theargumentsaresimilartothosedevelopedintheearlierdiscussionoftherationaleforforeignexchangeintervention. Capital controls and foreign exchange intervention are bothcomplementsandsubstitutes:complementsbecausecapitalcontrolsdecreasetheelasticityofflowswithrespecttorelativeratesofreturn,therebymakingforeignexchange interventionmore powerful; substitutes because both can be used toaffecttheexchangerate.Anadvantageofcapitalcontrolscomparedwithforeignexchangeinterventionisthattheycanbetargetedatspecificflows,but,preciselybecausecontrolsaretargeted,theyarealsomoreexposedtocircumvention(e.g.,whenflowsareopportunisticallyrelabeledtothatend).

Becausecapitalcontrolshavebeenusedmanytimesinthepast,evidenceontheir effects ismore abundant, but still surprisingly inconclusive (Ostry et al.2010).Anoftenstatedconclusionisthatcontrolsaffectthecompositionofflowsbutnottheirlevel;this,however,seemsunlikely,giventhespecializationofthedifferent types of investors. If capital controls decrease short-term flows, it isunlikelytheywillbereplacedbylong-termflows,oneforone.Firstreadingsofthe experience of Brazil, which has used taxes on capital inflows during thecurrentcrisis,varyingboth the tax rateand theperimeterof the taxover time,are mixed: despite some circumvention, they appear to have slowed downportfolio inflows and limited exchange rate appreciation (for two views, seeJinjarak,Noy,andZheng2012;ChamonandGarcia2013).

C.HowtoCombineMonetaryandMacroprudentialPolicies?

If macroprudential tools are to play an important role in the future, a centralissueisthewayinwhichmacroprudentialandmonetarypoliciesinteract:ontheone hand, low policy rates affect behavior in financial markets, leading topotentiallyexcessiverisktaking.Macroprudentialtools,ontheotherhand,affectaggregatedemandthroughtheireffectsonthecostofcredit.

In theory, if both policiesworked perfectly-that is, if they could be used toachievefullmacroeconomicandfinancialstability-thenmacroeconomicstabilitycould be allocated to the monetary authority and financial stability to themacroprudential authority. If a change in themonetarypolicy stance led to anexcessive increase or decrease in risk taking, macroprudential tools could beadjusted accordingly. Similarly, monetary policy could offset any decline inaggregatedemandassociatedwithatighteninginmacroprudentialconditions.

Inpractice, however, both toolswork far less thanperfectly.Therefore, onepolicy cannot be blind to the limitations of the other. To the extent thatmacroprudential tools work poorly, monetary policy must take into accountfinancial stability, as discussed in the section on monetary policy. Similarly,whenmonetarypolicy isunavailable todealwithan individualcountry'scycle(asunderacurrencyunionoranexchangeratepeg),macroprudentialtoolshavetocontributetothemanagementofaggregatedemand(foradiscussion,seeIMF2012c).

Inprinciple,coordinationbetweenthetwoauthoritiescansolvethisproblem;however,itislikelythateachpolicymakercaresprimarilyabouthisorherownobjective. If this is the case, separate agencies with different powers andmandates(acentralbank,muchlikethosewehavenow,inchargeofmonetarypolicy and tasked with price and output stability, and a financial authority incharge of macroprudential policy and tasked with macrofinancial stability)independently setting monetary and macroprudential policy will typically notendupcoordinatingon the first-best solution.Forexample, ina recession, thecentralbankmaycutthepolicyrateaggressivelytostimulatedemand.Worriedabout the effects of a relaxed monetary stance on risk taking, the financialauthoritymay reactby tighteningmacroprudential regulation.Anticipating thisresponseanditscontractionaryeffectondemand,thecentralbankmaycutrates

evenmoreaggressively.Andsoon.Theoutcome isapolicymixwith interestratesthataretoolowandmacroprudentialmeasuresthataretootightrelativetowhatacoordinatedsolutionwoulddeliver.

The obvious solution, on paper, to this problem is consolidation: puteverything under one roof, which is probably the preferable design. Indeed,beyond thearguments justgiven,putting thecentralbank inchargeofmicro -andmacroprudentialtoolsgivesitinformationusefultotheconductofmonetarypolicy (see, e.g., Coeure 2013; see Jacome, Nier, and Imam 2012 for adiscussionof institutional arrangements inLatinAmerica).Yet, just as for theconsolidation of micro - and macroprudential policies, there are also costsassociatedwiththisarrangement.

First,totheextentthatmacroprudentialtoolsworkimperfectly,acentralbankwith adualmandatewill have aharder timeconvincing thepublic that itwillfight inflation (and thus anchor expectations) if and when inflation fightingconflictswith theotherobjective. (Thiswasoneof theargumentsusedearlierformovingprudentialsupervisionoutofcentralbanksandgivingittofinancialstabilityauthorities.)

Second, and perhaps more critical, consolidation raises political economyissues. Central bank independence (achieved through the outsourcing ofoperationaltargetstononelectedtechnocrats)wasfacilitatedbyaclearobjective(inflation)andrelativelysimpleoperationaltools(openmarketoperationsandapolicy rate). The measurable nature of the objective allowed for easyaccountability, which in turn made operational independence politicallyacceptable. The objectives of macroprudential policy are murkier and moredifficult to measure, for several reasons. First, there are multidimensionalintermediate targets: credit growth, leverage, asset-price growth, and so on.Second is the issue of understanding the relationship of the macroprudentialobjectives to the financial stabilityobjective.Third,defining financial stabilityandidentifyingitsdesirablelevelisdifficult:apolicyratehikecanbedefendedafterthefactbyshowingthatinflationisclosetothetargetandarguablywouldhave exceeded it if tightening had not occurred, whereas a tightening ofmacroprudential measures that prevents a financial crisis could be attacked

afterwardasunnecessary.Fourth, theveryfact that themacroprudential tool istargeted implies that itsusemay raise strong, focusedpoliticalopposition.Forexample,younghouseholdsmaystronglyobject toadecreaseinthemaximumLTV.Becauseof thesefeatures, theindependenceofmacroprudentialpolicyisonweakerground.Andopponentsof the ideaofacentralizedauthorityworrythat political interference with macroprudential policy will undermine theindependenceofmonetarypolicy.(Again,theUKmaybeshowingtheway,byhaving aMonetary Policy Committee and a Financial Policy committee, bothwithintheBankofEngland).

IV.Conclusions

Togoback to the issue raisedat thestartof thediscussion,despitesignificantresearchprogressandpolicyexperimentationinthelasttwoyears,thecontoursof futuremacroeconomic policy remain vague. The relative roles ofmonetarypolicy, fiscalpolicy,andmacroprudentialpolicyarestillevolving.Wecanseetwo alternative structures developing: at a less ambitious extreme, a return toflexible inflationtargetingcouldbeforeseen,with littleuseoffiscalpolicyformacroeconomic stability purposes, and limited use of macroprudentialinstruments as they prove difficult or politically costly to use. At a moreambitious extreme, central banks could be envisaged to have a broadmacroeconomic and financial stability mandate, using many monetary andmacroprudential instruments, together with a more active use of fiscal policytools. Where we end up is likely to be the result of experimentation, withlearningpainsbutwiththeexpectationofmoresuccessfuloutcomes.

Note

This paper was written as background for the conference "RethinkingMacroeconomic Policy II," sponsored by the International Monetary Fund,Washington, D.C., April 16-17, 2013. We thank George Akerlof, MarkusBrunnermeier,OlivierCoibion,JorgDecressin,AvinashDixit,ChrisErceg,JoshFelman,andJonathanOstryforusefulcommentsandsuggestions.

References

Akitoby, Bernardin, Takuji Komatsuzaki, and Ariel Binder. Forthcoming."Inflation and Public Debt Reduction in the G7 Economies." InternationalMonetaryFund,Washington,DC.

Ball, Laurence. 2013. "The Case for Four Percent Inflation." Manuscript,DepartmentofEconomics,JohnsHopkinsUniversity,Baltimore.

Bernanke, Ben. 2011. "The Effects of the Great Recession on Central BankDoctrineandPractice."KeynoteaddressattheFederalReserveBankofBoston's56th Economic Conference, "Long-Term Effects of the Great Recession,"Boston,October18-19.

Blanchard,Olivier,andCarloCottarelli,2010."TenCommandmentsforFiscalAdjustment in Advanced Economies." Blogpost, IMF Direct blog, June 24.http:// blog-imfdirect.imf.org/2010/ 06/24/ ten-commandments-for-fiscal-adjustment-inadvanced-economies.

Blanchard,Olivier,GiovanniDell'Ariccia,andPaoloMauro.2010."RethinkingMacroeconomicPolicy."JournalofMoney,CreditandBanking42:199-215.

Blanchard,Olivier,andDanielLeigh.2013."GrowthForecastErrorsandFiscalMultipliers." IMF Working Paper 13/1, International Monetary Fund,Washington,DC.

Blanchard,Olivier,DavidRomer,MichaelSpence,andJosephStiglitz.2012.IntheWakeoftheCrisis.Cambridge,MA:MITPress.

Borio,Claudio,andIlhyockShim.2007."WhatCan(Macro-)prudentialPolicyDo to Support Monetary Policy?" BIS Working Paper 242, Bank forInternationalSettlements,Basel.

Chamon, Marcos, and Marcio Garcia. 2013. "Capital Controls in Brazil:Effective?"DiscussionPaper606,DepartmentofEconomics,PUC-Rio,RiodeJaneiro.

Coeure, Benoit, 2013. "Monetary Policy and Banking Supervision." Speechdelivered at the Institute for Monetary and Financial Stability, GoetheUniversity,Frankfurt,February7.

Cottarelli,Carlo, and JoseVinals. 2009. "AStrategy forRenormalizingFiscalandMonetaryPoliciesinAdvancedEconomies."IMFStaffPositionNote09/22,InternationalMonetaryFund,Washington,DC.

Crowe, Chris, Giovanni Dell'Ariccia, Deniz Igan, and Pau Rabanal. 2011."PoliciesforMacrofinancialStability:OptionstoDealwithRealEstateBooms."IMFStaffDiscussionNote11/2,InternationalMonetaryFund,Washington,DC.

Dell'Ariccia,Giovanni,DenizIgan,LucLaeven,andHuiTong,withBasBakkerand JeromeVandenbussche.2012. "Policies forMacrofinancialStability:Howto Deal with Credit Booms." IMF Staff Discussion Note 12/6, InternationalMonetaryFund,Washington,DC.

European Central Bank (ECB). 2012. "Report on the First Two Years of theMacroPrudential Research Network." Frankfurt, European Central Bank,October. http://www.ecb.europa.eu/ pub/pdf/other/macroprudentialresearchnetworkreport20121Oen.pdf.

Farhi, Emmanuel, and Jean Tirole. 2012. "CollectiveMoral Hazard,MaturityMismatchandSystemicBailouts."AmericanEconomicReview102(1):60-93.Gagnon, Joseph,MatthewRaskin, JulieRemache,andBrianSack.2011. "TheFinancialMarketEffectsoftheFederalReserve'sLarge-ScaleAssetPurchases."InternationalJournalofCentralBanking7(1):3-43.

InternationalMonetary Fund (IMF). 2008. "TheChangingHousingCycle andthe Implications for Monetary Policy," chap. 3 in World Economic Outlook.InternationalMonetaryFund,Washington,DC.

InternationalMonetaryFund (IMF). 2010. "Strategies forFiscalConsolidationinthePost-CrisisWorld."FiscalAffairsDepartmentPaper10/04,InternationalMonetaryFund,Washington,DC.

InternationalMonetaryFund(IMF).2012a.FiscalTransparency:Accountabilityand Risk. Fiscal Affairs Department in collaboration with the StatisticsDepartment,InternationalMonetaryFund,Washington,DC.

InternationalMonetaryFund(IMF).2012b.TheLiberalizationandManagementof Capital Flows: An Institutional View. International Monetary Fund,

Washington,DC.

International Monetary Fund (IMF). 2012c. The Interaction of Monetary andMacroprudentialPolicies.Washington,DC:InternationalMonetaryFund.

International Monetary Fund (IMF). 2013. "The Dog That Didn't Bark: HasInflationBeenMuzzledorWasItJustSleeping?,"Chap.3inWorldEconomicOutlook.InternationalMonetaryFund,Washington,DC.

Irwin, Timothy. 2012. "Accounting Devices and Fiscal Illusions." IMF StaffDiscussionNote12/02,InternationalMonetaryFund,Washington,DC.

Jacome,LuisI.,ErlendW.Nier,andPatrickImam.2012."BuildingBlocksforEffective Macroprudential Policies in Latin America: InstitutionalConsiderations." IMF Working Paper 12/183, International Monetary Fund,Washington,DC.

Jinjarak, Yothin, Ilan Noy, and Huanhuan Zheng. 2012. "Capital Controls inBrazil:StemmingaTidewithaSignal?"WorkingPaper,SchoolofEconomicsandFinance,Victoria,UniversityofWellington,NZ.

King,Mervyn, 2012. "TwentyYears of InflationTargeting." StampMemorialLecture,LondonSchoolofEconomics,London,October9.

Mauro, Paolo, ed. 2011. Chipping Away at Public Debt: When Do FiscalAdjustment Plans Fail? When Do TheyWork? Hoboken, NJ: JohnWiley &Sons.

Mauro, Paolo, and Mauricio Villafuerte. 2013. "Past Fiscal Adjustments:LessonsfromFailuresandSuccesses."IMFEconomicReview61(2):379-404.

McKay,Alisdair,andRicardoReis,2012."TheRoleofAutomaticStabilizersinthe US Business Cycle." Manuscript, Department of Economics, BostonUniversity,Boston, andDepartmentofEconomics,ColumbiaUniversity,NewYork.

Mishkin,Frederic.2010. "MonetaryPolicyStrategy:Lessons from theCrisis."Paper presented at the ECB Central Banking Conference, "Monetary PolicyRevisited:LessonsfromtheCrisis,"Frankfurt,November18-19.

Ostry, Jonathan David, Atish R.Ghosh, and Marcos Chamon. 2012. "TwoTargets,Two Instruments:Monetary andExchangeRatePolicies inEmergingMarketEconomies." IMFStaffDiscussionNote12/01, InternationalMonetaryFund,Washington,DC.

Ostry, Jonathan, Atish R.Ghosh, Karl Habermeier, Luc Laeven, MarcosChainon, Mahvash S.Qureshi, and Annamaria Kokenyne. 2010. "ManagingCapital Inflows: What Tools to Use?" IMF Staff Discussion Note 11/06,InternationalMonetaryFund,Washington,DC.

Ostry, Jonathan,AtishGhosh,andAntonKorinek.2012."MultilateralAspectsof Managing the Capital Account." IMF Staff Discussion Note 12/10,InternationalMonetaryFund,Washington,DC.

Perotti, Roberto. 2011. "The `Austerity Myth': Gain Without Pain?" NBERWorking Paper 17571, National Bureau of Economic Research, Cambridge,MA.

Reinhart, Carmen M., and M.Belen Sbrancia. 2011. "The Liquidation ofGovernmentDebt."NBERWorkingPaper16893,NationalBureauofEconomicResearch,Cambridge,MA.

Saurina,Jesus.2009."DynamicProvisioning:TheExperienceofSpain."CrisisResponseNote7,WorldBank,Washington,DC.

Schaechter, Andrea, Tidiane Kinda, Nina Budina, and Anke Weber. 2012."FiscalRulesinResponsetotheCrisis:Towardthe`Next-Generation'Rules-ANew Dataset." IMF Working Paper 12/187, International Monetary Fund,Washington,DC.

Spilimbergo,Antonio,StevenA.Symansky,andMartinSchindler.2009."FiscalMultipliers." IMF Staff Position Note 2009/11. International Monetary Fund,Washington,DC.

Svensson,Lars,2009."FlexibleInflationTargeting:LessonsfromtheFinancialCrisis."SpeechdeliveredatDeNederlandscheBank,Amsterdam,September21.

Woodford, Michael. 2012. "Methods of Policy Accommodation at theInterestRate LowerBound." Paper presented at the JacksonHole Symposium,

"TheChangingPolicyLandscape,"August31-September1.

JanetL.Yellen

Thankyou to the InternationalMonetaryFund for allowingme to takepart inwhatIexpectwillbeaverylivelydiscussion.'

Only five or six years ago, therewouldn't have been a panel on the "manyinstruments"and"manytargets"ofmonetarypolicy.Beforethefinancialcrisis,the focus was on one policy instrument: the short-term policy interest rate.Although central banksdidnot uniformly relyon a single policy target,manyadopted an "inflation targeting" framework that, as the name implies, gives acertain preeminence to that one objective.Of course, the Federal Reserve haslongbeenabitofanoutlierinthisregard,withitsexplicitdualmandateofpricestability andmaximumemployment. Still, the discussionmight not have gonemuchbeyond"oneinstrumentandtwotargets"ifnotforthefinancialcrisisandits aftermath, which have presented central banks with great challenges andtransformedhowwelookatthistopic.

Letmestartwithafewgeneralobservationstogettheballrolling.Intermsofthe targets,or,moregenerally, theobjectivesofpolicy, I seecontinuity in theabiding importance of a framework of flexible inflation targeting. By oneauthoritative account, about twenty-seven countries now operate full-fledgedinflation-targetingregimes.2TheUnitedStatesisnotonthislist,buttheFederalReservehasembracedmostofthekeyfeaturesofflexibleinflationtargeting:acommitment topromote lowand stable inflationovera longer-termhorizon, apredictable monetary policy, and clear and transparent communication. TheFederalOpenMarketCommittee (FOMC) struggled for years to formulate aninflation goal that would not seem to give preference to price stability overmaximumemployment.InJanuary2012,thecommitteeadopteda"StatementonLonger-RunGoals andMonetaryPolicyStrategy,"which includes a 2 percentlonger-runinflationgoalalongwithnumericalestimatesofwhatthecommittee

viewsasthelonger-runnormalrateofunemployment.Thestatementalsomakesclear that the FOMC will take a "balanced approach" in seeking to mitigatedeviations of inflation from 2 percent and employment from estimates of itsmaximum sustainable level. I see this language as entirely consistent withmoderndescriptionsofflexibleinflationtargeting.

For thepast fouryears,amajorchallengefor theFederalReserveandmanyother central banks has been how to address persistently high unemploymentwhen the policy rate is at or near the effective lower bound. This troublingsituationhasnaturallyandappropriatelygivenrisetoextensivediscussionaboutalternativepolicyframeworks.Ihavebeenverykeen,however,toretainwhatIsee as the key ingredient of a flexible inflation-targeting framework: clearcommunicationaboutgoalsandhowcentralbanksintendtoachievethem.

With respect to the Federal Reserve's goals, price stability and maximumemployment are not only mandated by the Congress but are also easilyunderstandableandwidelyembraced.Well-anchoredinflationexpectationshaveproventobeanimmenseassetinconductingmonetarypolicy.Theyhavehelpedkeep inflation low and stable while monetary policy has been used to helppromoteahealthyeconomy.After theonsetof thefinancialcrisis, thesestableexpectationsalsohelped theUnitedStatesavoidexcessivedisinflationorevendeflation.

Of course, many central banks have, in the wake of the crisis, found itchallenging toprovideappropriatemonetarystimulusafter theirpolicy interestratehit theeffective lowerbound.This is thepointwhere"many instruments"entersthediscussion.ThemaintoolsfortheFOMChavebeenforwardguidanceonthefuturepathofthefederalfundsrateandlarge-scaleassetpurchases.

Theobjectiveofforwardguidanceistoaffectexpectationsabouthowlongthehighly accommodative stance of the policy interest ratewill bemaintained asconditionsimprove.Byloweringprivate-sectorexpectationsofthefuturepathofshort-term rates, this guidance can reduce longer-term interest rates and alsoraise asset prices, in turn, stimulating aggregate demand.Absent such forwardguidance, the public might expect the federal funds rate to follow a path

suggestedbypastFOMCbehaviorin"normaltimes"-forexample,thebehaviorcapturedbyJohnTaylor'sfamousTaylorrule.Iampersuaded,however,bytheargumentslaidoutbyourpanelistMichaelWoodfordandotherssuggestingthatthepolicyrateshould,underpresentconditions,beheld"lowerforlonger"thanconventionalpolicyrulesimply.

IseetheseideasreflectedintheFOMC'srecentpolicy.SinceSeptember2012,the FOMChas stated that a highly accommodative stance ofmonetary policywill remain appropriate for a considerable time after the economic recoverystrengthens.SinceDecember2012,thecommitteehassaiditintendstoholdthefederalfundsratenearzeroatleastuntilunemploymenthasdeclinedbelow6~/apercent,providedthatinflationbetweenoneandtwoyearsaheadisprojectedtobenomorethanahalfpercentagepointabovetheCommittee's2percentlonger-rungoal,andlonger-terminflationexpectationscontinuetobewellanchored.Ibelievethattheclarityofthiscommitmenttoaccommodationwillitselfsupportspendingandemploymentandhelptostrengthentherecovery.

Asset purchases have complemented our forward guidance, and the manydimensions of different purchase programs arguably constitute "manyinstruments."Indesigningapurchaseprogram,onemustconsiderwhichassetsto buy: justTreasury securities or agencymortgage-backed securities aswell?Which maturities? The Federal Reserve, the Bank of England, and, morerecently,theBankofJapanhaveemphasizedlongerdurationsecurities.Atwhatpaceshouldthesecuritiesbepurchased?Andhowlongshouldtheybeheldoncepurchasescease?Eachofthesefactorsmayaffectthedegreeofaccommodationdelivered.Twoinnovations in theFOMC'scurrentassetpurchaseprogram,forexample, are that it is open-ended rather than fixed in size like past programsandthattheoverallsizeoftheprogramisexplicitlylinkedtoseeingasubstantialimprovementintheoutlookforthelabormarket.

Inthesebriefremarks,Iwon't thoroughlyreviewthebenefitsorcostsofourhighly accommodative policies, emphasizing only that I believe they have, onnet, provided meaningful support to the recovery. But I do want to spend amomentononepotentialcost-financial stabilitybecause this topic returnsus tothe theme of "many targets" for central banks. As Chairman Bernanke has

observed, in the years before the crisis, financial stability became a "juniorpartner" in themonetarypolicyprocess, incontrastwith its traditionally largerrole.3 The greater focus on financial stability is probably the largest shift incentralbankobjectiveswroughtbythecrisis.

Somehaveaskedwhethertheextraordinaryaccommodationbeingprovidedinresponsetothefinancialcrisismayitselftendtogeneratenewfinancialstabilityrisks.Thisisaveryimportantquestion.Toputitincontext,let'srememberthatthe Federal Reserve's policies are intended to promote a return to prudentrisktaking, reflecting a normalization of credit markets that is essential to ahealthyeconomy.Obviously, risktakingcango toofar.Lowinterest ratesmayinduce investors to take on toomuch leverage and reach too aggressively foryield.Idon'tseepervasiveevidenceofrapidcreditgrowth,amarkedbuildupinleverage,orsignificantassetbubblesthatwouldthreatenfinancialstability,buttherearesignsthatsomepartiesarereachingforyield,andtheFederalReservecontinuestocarefullymonitorthissituation.

However, I thinkmost centralbankersviewmonetarypolicyas ablunt toolfor addressing financial stability concerns and many probably share my ownstrong preference to rely on micro - and macroprudential supervision andregulation as themain lineofdefense.TheFederalReservehasbeenworkingwith a number of federal agencies and international bodies since the crisis toimplement a broad range of reforms to enhance our monitoring, mitigatesystemic risk, and generally improve the resilience of the financial system.Significantworkwillbeneededtoimplementthesereforms,andvulnerabilitiesstill remain. Thus, we are prepared to use any of our many instruments asappropriatetoaddressanystabilityconcerns.

LetmeconcludebynotingthatIhavetouchedononlysomeoftheimportantdimensionsofmonetarypolicytargetsandinstrumentsthathaveariseninrecentyears.IlookforwardtoadiscussionthatIexpectwillexploretheseissuesandperhapsraiseothers.

Notes

1. The views I express here are my own and not necessarily those of mycolleaguesintheFederalReserveSystem.

2.SeeGillHammond,StateoftheArtofInflationTargeting,CentreforCentralBanking Studies,CCBSHandbookNo. 29 (London:Bank of England, 2012),www.bankofengland.co.uk/education/ Documents/ccbs/handbooks/pdf/ccbshb29.pdf.

3. http://www.federalreserve.gov/ newsevents/lectures/ the-aftermath-of-the-crisis.htm.

LorenzoBiniSmaghi

Inaddressingmonetarypolicy'stargetsandinstruments,Iwouldliketofocusonhowmuchof theprecrisis inflationtargetingframeworksweshouldkeepgoingforward.Therearetwodimensionstotheissue.Thefirstrelatestotheabilityofthe inflationtargeting framework to ensure price stability, in particular duringtimesofmarketexuberance,whichmay lead to thebuildupofbubbles,whoseburst jeopardizes financial stability and thus price stability. In light ofexperience, a pure inflationtargeting framework, one that ignores financialimbalances,maynotallowapropercalibrationofmonetarypolicy.Theseconddimension relates to the postcrisis developments, where the inflationtargetingframework has not provided an appropriatemonetary policy stance capable ofaddressing the challenges that advanced economies are currently facing. Thetaskofcleaningupthemessafterthebubbleburstismuchmorecomplexthanwhattheprecrisisconventionalwisdomexpected.

Thereisanincreasingliteratureonthefirstissue,inparticularthataddressinghow to make inflationtargeting frameworks more flexible to incorporate thefinancialsectorandhowtobroadentherangeoftoolsavailabletocentralbanksto address financial stability issues so as to prevent credit bubbles. I willtherefore concentrateon the second issue,whichquestions theuseof inflationtargetingastheappropriateframeworkforgettingoutofacrisislikethe2008-2012one.

Indeed, monetary policy does not currently seem to be as effective asexpected.Thiscertainlypointstotheneedtoimproveeconomicmodelsthatareused for policy analysis. The question is whether central banks should worktowardimprovingtheexistingmodelsorwhethertheyneedtototallychangetheapproach.

Independent of any improvement that might be made in understanding

monetary policy, we should not depart from a few fundamental principles,relatedinparticulartoassigningpolicyinstrumentstotargets.Twoprinciplesareworthrecallinginallcircumstances.Thefirstisthateachinstrumentshouldbeassignedtoaspecifictarget.Thesecondprincipleisthattheassignmentshouldbebasedonefficiency;thatis,eachinstrumentshouldbeassignedtothetargetitcanachievemosteffectively.

Thesetwoprinciplessuggestthereisnoreasontodepartinafundamentalwayfrom inflation targeting as the basic analytical framework formonetary policyduringandafter a crisis.Thatwedonotunderstandwhycertain relations thatwere assumed to hold in the past no longer hold is not a good reason tocompletelychangethepolicyframework.

The main frustration with the inflationtargeting framework comes from thefactthatmonetarypolicy,evenwhenpushedtotheextremeofkeepingverylowinterest rates for a prolonged period of time and implementing nonstandardmeasures, which have enormously increased the size of the central bank'sbalancesheet,doesnotseemtobeeffectiveinraisinggrowth,whereasinflationremainsbroadlyinlinewithitstarget.Indeed,targetinginflationisnotagoalinitself but aims at creating the conditions for sustainable growth. If keepinginflationontargetisnotleadingtostrongergrowth,thenwhatisitsworth?

Therearetwopossiblereactionstothisfrustration.Thefirstistoconsiderthatifmonetary policy has not been very effective so far in supporting growth, itshouldbecomeevenmoreexpansionary.Ifthemoneymultiplier,sotospeak,islowerthanexpected,weneedmoremonetaryexpansion,inallofitsforms.Theoppositereactionis toconsider thatperhapsthewayinwhichmonetarypolicyhasbeenconductedsofarisnotveryeffectiveinaddressingtheproblemsfacedbyadvancedeconomies.Tryingtopushfurtheronthisfrontmayactuallymakemonetarypolicyevenlesseffectiveandincreasethecollateraldamageovertime.

Thetwohypothesesshouldbetested.Theproblemisthatwedonotseemtohave the right analyticalmodel to conduct such an exercise. Indeed, the resultwilldependon themodel that isused tocompare thecostsandbenefitsof thetwo alternatives. A standard neo-Keynesian model, for instance, without a

sophisticatedfinancialsector,wouldprobablyvalidate thefirsthypothesis.Butwe know that models that ignore the financial sector do not provide a gooddescriptionofadvancedeconomies.Theiruncriticaluseinthepastmayactuallyhavebeenresponsibleforthepoliciesthatledtothecrisis.

Thekeychallengeisthustoimprovemodelswithaviewtoacquiringabetterunderstanding of monetary policies' impact on agents' behavior, in particularaftertheglobalfinancialcrisisof2008-2009anditsaftermath.HereIwouldliketodrawananalogywiththestateofmonetarytheorybeforerationalexpectationsbecameanacceptedpartoftheeconomicsliterature.StandardKeynesiantheoryof the 1960s assumed that agents would react in a naive way to monetaryauthorities'attempttoinduceinflationsoastoreducetherealvalueofdebtsandwages.That theory, however, could not explainwhymonetary policywas notcapable of systematically increasing employment. The reason was that thetheoreticalmodel usedby the policymakerswasmisspecified.Theoldmodelsassumed agents to be naive, forming their expectations in a backwardlookingfashion,whereasinfacttheywerenot.Agentsaremuchmoreforward-lookingthanwasthoughtatthetimeandknowthatwhentheeconomyslowsdown,thecentral bank will try to stimulate the economy by creating surprise inflation.Theywill thusrationally try toprotect their incomeandwealth. Ifagentshaverational expectations, monetary policy becomes less effective. Rationalexpectations are an extreme assumption, of course, but one that is useful inexplainingthelimitsoftraditionalmodelsofmonetarypolicy.

Backtothepresent:wemayaskourselveswhethermonetaryauthorities-andeconomists-are not now making a similar mistake in considering economicagents naive when in fact they are not.We know in particular-and economicagentsknow-thattherearethreewaystogetoutfromadebtoverhang.Thefirstistosaveyourwayout,whichmeanslowgrowthforaprotractedperiodoftime.Thesecondisdefaultordebtrestructuring.Thethirdisinflation.

Economicagentswanttounderstandwhatobjectivethecentralbankpursueswhenitembarksonexceptionalnonstandardmeasures.Theireffectivenessmaybeverydifferentiftheaimistorepairthetransmissionmechanismofmonetarypolicywhenthelatterisimpairedbyinefficienciesorbottlenecksinthefinancial

system, as may be the case, for instance, in the euro area because of thesovereign debt crisis or an undercapitalized banking system, or when theobjective is to increase the amount of liquidity in the system with a view toinduceinvestorstodiversifyintomoreriskyassetssoastostimulateaggregatedemand,asmaybethecasetodayinJapanandtheUnitedStates.

In other words, it's different if monetary policy tries to address liquidityproblems in specific markets instead of stimulating aggregate demand orrepairingsolvencyproblems.

In the first case, monetary policy uses instruments to try to unclog thetransmissionmechanism,enhancingthecorrelationbetweenthepolicyrateandthe interest rate to end users. An example is the Securities Market Programimplementedby theEuropeanCentralBank,whoseeffectson themoneybasearesterilized,orthepurchaseofcoveredbonds,whichrepresentsanattempttotry to revitalize a specific segment of the banking system. The impact oninflationandwealthdistributionislimited.

In thesecondcase, theobjectiveofmonetarypolicy isbroader,as it tries toenhance its overall effectiveness by influencing the allocation of resourcesacross agents andwithin financialmarkets, inparticularbetweencreditors anddebtors. Quantitative easing and forward guidance are aimed at this second,broader objective. This type of policy ismore likely to generate reactions byeconomicagents, inparticularcreditors,whichmay in turnhave repercussionsfortheeffectivenessofthepolicyitself.

Inflationreducestherealvalueofthedebt,andthusredistributeswealthfromcreditors to debtors. The best way to redistribute wealth ex post is to createunexpectedinflationwhilemaintaininglowinterestratesforaprotractedperiodoftime.Thisisthestockadjustmenteffect.Byreducingtherealreturnonsafeassets, thisstrategyalsoredistributeswealththroughafloweffect,as investorsshifttheirpreferencestowardriskierinvestmentsuntilinflationpicksupandthereturnsontheseassetsturnouttobelowerthananticipated.

The instruments that are currently used by central banks, such as very low

interest rates for a prolonged period of time, forward guidance, and massivebalance sheet expansion, are ways to try to tax creditors and subsidizeborrowers, either directlyby inferring capital losses andgainsor indirectlybysocializinglossesaccumulatedoncentralbanks'balancesheets.Itisanotherwayoftryingtodowhatcentralbanksdidinthe1960sbyreducingtherealvalueofwagesor,toputitinMiltonFriedman'sterms,tofoolsomeofthepeople(i.e.,creditors) at least some of the time. This is why such a policy is also named"financialrepression."Financialrepressionisactuallytheoptimalwaytogetoutofadebtcrisis,muchassurpriseinflationusedtobeconsideredtheoptimalwaytoreducerealwagesandstimulateemployment.

Suchapolicyassumes,however,thateconomicagents,inparticularcreditors,arenaiveanddonotreacttopolicymakers'attemptstoreducetherealvalueoftheirassets.Thisiswhatthemodelsusedbythepolicyauthoritiesassume.Andthisiswhatmakesthepolicyoptimal.Butit isfairtoask,assomeeconomistsdid 40 years ago, whether the real world really looks like the models thatpolicymakersuse,andwhetheragentsareindeedsonaiveastopassivelyacceptfinancial repression. The fact that monetary policy is not as effective as wethoughtmaysuggestthatit'snottheeconomicagentswhoarenaive.

Iwouldliketoraiseacoupleofissuesinthisrespect.

The first is that financial repression is not easy to implement in highlysophisticated financial markets, where it is the task of investors to protectthemselves against risk, including especially the risk of being trapped intoexcessively low returns on investments. Our models may be too simple tocapturesuchbehavior.

Second,evenwhenagentsarenotable toprotect themselvesandarepushedtowardinvestinginhighlyriskyassetsbecauseofalackofalternatives,thatdoesnotmean they are naive andwill take the loss passivelywhen itmaterializes.Sophisticatedinvestorsmayknow,forinstance,thatincurrentmarketconditionssome assets are overpriced, in particular in the fixed incomemarket, but theymaystillbewillingtoholdthemandevencontinuetobuythemaslongastheythink that (1) in theshort run theseassetsareneverthelessrelativelyattractive,

and itmay be risky to hold positions that are contrarian to the central bank'spolicy,and(2)theywillbethefirst tosell theseassetswhenthebubbleburstsandwillthereforecontainanyloss.Weknow,however,thatnotallinvestorscanbefirstoutofthedoor,andthussomewillhavetobearlargelosses,buttherushtothedoorcouldbesodisorderlythataburstingbubblemightbeverydamagingtoeconomicandfinancialstability.

Onereasonwhycentralbanksmayfeelconfident inusing theoldmodels isthat inflation has remained low and inflation expectations are still anchored,despitetheveryexpansionarymonetarypoliciesthathavebeenimplementedsofar.

Hereagain,onemightquestionwhetherwehavetherightmodelofinflation.Our simple models may not sufficiently take into account the fact that assetpricesmovemorequickly thandogoodsandserviceprices. In fact, assetpricebubblesmaybuildupandburstevenbeforegoods inflationstarts rising.Afterall,isn'tthatwhathappenedin2005-2007?Thebubbleburstwhileinflationwasstillrelativelymoderate.

We could thus envision a world in which inflation remains low but theaccommodativemonetary policy of the central bank creates assetprice bubblethat may burst even before inflation materializes. The burst itself generatesdeflationarypressuresinthegoodsmarkets,whichthenrequiresthatthecentralbankkeepaveryaccommodativemonetarypolicy.Insuchaworld,realinterestratesthatarekeptlowfortoolonggenerateboutsoffinancialinstability,whichnegatively affects the real economy and justifies keeping rates low for evenlonger.Therelativestability in inflationand in the(low) levelof interest ratescreates the conditions for instability in the financial system and in the realeconomy. In other words, it's not monetary conditions that adapt to real andfinancialconditionsbut thereverse.Financialconditionsand therealeconomybecome more unstable as a result of low and stable inflation, fostered byaccommodativemonetaryconditions.

Howfar is thisdescriptionfromreality?Doesn't it lookverysimilar towhatwehaveobservedoverthepastdecade,withrelativelylowrealratesofinterest

andmorevolatilerealgrowth?Themodelsusedbycentralbankstoimplementmonetary policy do not consider these effects. Maybe we just need anotherfinancialcrisisforsuchissuestogetgreaterattention.

Iwould like tomakea finalpoint,whichalso touchesupon the structureofthisconference,inwhichthevariousmacropoliciesaretreatedseparately.Ifyouthink about the current discussion about the role of monetary policy-and, inparticular, the need for monetary policy to focus more on growth-it is partlyobscuredbythefactthatotherpolicyinstrumentsaresubjecttoseverefinancialandpoliticalconstraints.Thatotherinstrumentsareconstraineddoesnotjustifyshifting the responsibility for achieving goals other than price stability tomonetarypolicywithoutabetterunderstandingof thepotentialdistortionsandrisks thatmaybe created in the short tomedium term.Before abandoning thebasicprinciplethatmonetarypolicyshouldprimarilyaimatpricestability,thereshould be a sound and transparent analysis of the intertemporal costs that thismayentailforsocietyasawhole.

For instance, by conducting aggressive asset purchases and maintaininginterest rates at very low levels, the central bank undoubtedly takes away theincentivefromthefiscalauthoritiestoimplementmediumtermfiscaladjustmentplans.Wehaveseenthisintheeuroarea.ItisalsoapparentintheUnitedStates.The policy of persistently low long-term interest rates removed the incentivefromtheObamaadministrationandtheUSCongresstoagreeonamediumtermfiscal adjustment plan, and the lack of such a plan creates uncertainty aboutfuturetaxation,whichmayhinderfirmsfrominvesting.

Thequestionshouldbe,areinvestmentsmoresensitivetothepromiseoflowinterest rates for a prolonged period of time or to the mediumterm fiscaluncertaintythatsuchlowinterestratesinvoluntarilyproduce?

Thelowleveloflong-termratesproducedbythebondpurchaseprogramalsocontributes to the illusion, held even among some prominent economists, thatthereisnoneednorpressureforbudgetaryadjustment.

Another institutional issue relates to the redistributive effects of monetary

policy, which are essential for its effectiveness in a post-bubble situation. Itmightbequestionedwhethersuchasensitivepolitical issuecancontinuetobedealtwithinasemi-disguisedwaybythecentralbankonly,ratherthanthroughthe normal democratic decision-making process,which involves, in particular,theparliament.

The crisis has shown how political authorities in our democracies try topostponepainfuldecisionsandactonlywhentheycomeundermarketpressure,asalastresort,attheedgeofthecliff.Byrelievingmarketpressure,thecentralbank de facto can be led to play a very important political role. It can helpgovernments gain time for implementing their policies, but it can also leadgovernments to waste time and indefinitely postpone urgent decisions. Thecentralbankcanactuallybecorneredintobeing"theonlygameintown."

So, isn't it a bit naive to assume thatwhenmonetarypolicy starts aimingatother goals, which should be in the realm of the political authorities, theinstitutional framework underlying the central bank and its independencewillnot be affected? Isn't it a bit naive to assume that the distortions that areintroducedintheeconomyasaresultofcentralbanksenteringotherfieldsthanmonetary policy can be removed through decisions that can be of a technicalnature,thatis,decisionsthatcanbemadejustbythecentralbank?

Inotherwords,thereisariskthattheexitstrategyfromamonetarypolicythatproducesfiscalordistributionaleffectswillnolongerbejustamonetarypolicyissuebutamajorpoliticalone.Itisthus,inmyview,delusionaltothinkthatthedebateabouttheexitfromthecurrentpolicyisoverwhetherthecentralbankhasthe instruments to exit such a policy or not. The central bank does have theinstruments.Rather,therealquestioniswhetheritwillhavethepoliticalabilitytodecidetheexit.

Tobeblunt, the independenceof thecentralbankisatriskwhenthecentralbankentersthefieldoffiscal,regulatory,anddistributionalpolicies.

Inmyview,itisverydangeroustoaddresssuchquestionsas"Shouldcentralbanks more explicitly target activity? Should central banks target financial

stability,andifso,how?"withoutconsideringwhattheroleoftheotherpoliciesshould be, in particular fiscal, structural, and supervisory policies. Anyunconditionalanswertothesequestions-whichwouldmakethecentralbanktheonly player left in town-would lead to a timeinconsistent monetary policybecauseitwouldcreateincentivesamongtheotherpolicymakerstoactinawaythatwouldleadmonetarypolicytobecomeineffective,andwouldatsomepointhavetoberenegedon.Thiswouldinevitablyleadtoalossofcredibilityofthecentralbank.Insofarasalltheotherpolicymakershavealreadylostpartoftheircredibility, thepossibility that thecentralbankmay lose itaswellshouldbeaconcernforsocietyasawhole,notonlyforthecentralbank.Thisissueshouldbeat theheartof the surveillancemandateof international institutions suchastheInternationalMonetaryFund.

Iwillconcludewithafewwordsonwhethercentralbanksshouldcareabouttheexchangerate.Itseemstomethatcentralbanksactuallycaretoomuchabouttheexchangerate,althoughtheywillnotadmitit.Inthecurrentenvironmenttheexchange rate remains themostpowerful channelof transmissionofmonetarypolicy.However,thisistrueonlytotheextentthatothercentralbanksignoreit.Becauseothercentralbankswillnotignoreit,andcannotignoreitinthecurrentenvironment,monetarypolicydesignedseparatelybyeachnationalcentralbankwill inevitablybesuboptimaland lead toexcessmoneycreation,and thus fuelthenextassetbubble.

This isa typicalexampleofcoordination failure.Again, this isan issue thatcannotbeignoredbyinternationalmacroeconomicsurveillance.

MervynA.King

The past five years have been an extraordinary period for central banks. Thebreadthandscaleofouroperationshaveexpandedinwaysthatwerepreviouslyunimaginable aswe responded to a crisis in the banking sector and thewidereconomy.Inmonetarypolicy,wehavemovedintounchartedterritory.Buthasournotionofwhatcentralbanksshoulddo,andhow,changed?Nowisagoodtimetoreflectonwherewestand.

Iwanttofocusthisdiscussionontwoareas.First,Iwanttodistillwhathavewelearnedabouttheobjectivesofmonetarypolicy.Second,Iwanttoreflectontheimplicationsoftheproliferationofinstrumentsthathavebeenusedtomeetthoseobjectives.

Objectives

Formore than30years, theobjectiveofcentralbankswasclear: itwas to setmonetary policy to achieve long-run price stability.But the events of the pastfiveyearshave raisedquestionsabouthowcentralbanksmanage the tradeoffsbetweenpricestability,output stability,and financial stability inorder tomeetouroverallmacroeconomicobjectives.

Throughout the era of inflation targeting, the importance of the tradeoffbetween output and inflation stabilization in the short term has been wellunderstood.Monetarypolicywasseenasaimingata target for inflation in thelongrun,whichwas tobeachievedbybringing inflationback to targetoverasuitabletimehorizonsoastoavoidexcessivevolatilityofrealvariables,suchasoutputandemployment.Optimalmonetarypolicywasseenasachoiceofhowbesttonavigatetheshort-runtradeoffwhileensuringthatthelong-runobjectivewasmet.

Figure3.1

UKInflationandOutputGapVariances.

Source:OfficeforNationalStatisticsandBankofEnglandcalculations.

Failure to deliver price stability is costly, as the experience of the UnitedKingdomamplydemonstrates.Figure3.1showsthevarianceofinflationandthevarianceoftheoutputgapintheUKusingquarterlydatafortwoperiods.'Herewemay contrast the performance in the 35 years up to 1992with the first 15years of inflation targeting. Better policy can take some credit for thisimprovement,astheanchoringofinflationexpectationsledtoahugereductionininflationvolatility.

Itwastemptingtothinkthatwehadmovedontothe"Taylorfrontier,"whichmaps feasible combinations of the smallest variances of the output gap andinflation. The Great Stability appeared to be a permanent break from earlierperiods-periods when monetary policy exhibited more unpredictable behaviorandleft theeconomyaboveand to therightof theTaylorfrontier (figure3.2).Andinoneimportantsense,itwas:thedarkdaysofdouble-digitinflationwereconsignedtothepast.

ButtheGreatStabilitywasnotrepresentativeofanewnormal.Thevarianceoftheoutputgap-thoughnotofinflation-hasbeenmuchhigheroverthepastfiveyearsasthefinancialcrisisgeneratedadeeprecession(figure3.3).Sowhathavewelearned?

The banking and broader economic crisis has demonstrated thatmacroeconomic policy can face an additional tradeoff between ensuring thesoundnessofthefinancialsysteminthemediumtermandkeepingoutputinlinewith potential output and inflation on target in the near term. Such a tradeoffarises because financial vulnerabilities canbuild evenwhile output is growingsteadilyandinflationislowandstable.

Figure3.2

TheUK'sPrecrisisTradeoff?

Figure3.3

TheGreatStabilityandCrisisPeriods.

Let me give three examples of the sort of underlying mechanisms in play.First,persistentmisperceptionsoffuturespendingpowermaygenerateamixofdemand thatproves tobeunsustainable. Ibelieve thiswasan important factorunderlying the crisis. Although output in deficit countries, such as the UK,appearedtobegrowingatasustainablerate,thatgaveamisleadingimpressionofthesustainabilityoftheGreatStability.Infact,thelevelofdomesticdemandwastoohighandthelevelofnetexportscorrespondinglytooweak.Second,asHymanMinsky described, periods of stability encourage exuberance in creditmarkets,leadingeventuallytoinstability.2Andthird,lowshort-termpolicyratesmayencourageinvestorstotakeonmoreriskthantheywouldotherwiseacceptasthey"searchforyield."

It is arguable that monetary policy paid insufficient heed to the potentialimpactofsuchfinancialvulnerabilities.Financialshocksarecostlybecausetheireffectscanbetoorapidforpolicyeasilytooffsetandbecausetheyhitpotentialsupply,therebycreatingatradeoffbetweenoutputandinflation.Inotherwords,the Taylor frontier is less favorable (further from the origin)when account istakenoffinancialshocksthanwemighthavebelieved.Takingtheentireperiod

of inflation targeting, including the recent past, might give a more accurateindication of where this "Minsky-Taylor" frontier lies than using data for theGreatStabilityperiodalone(figure3.4).

Figure3.4

TheTrueTradeofffortheUKover20Years?

What implicationsdoes thishave formonetarypolicy?Possiblynone, ifwecanrelynowonmacroprudentialtoolstoensurefinancialsectorresilience.Butsetthosetoolstoonesideforamoment.MonetarypolicycouldbeusedtoreachapointmorelikePinfigure3.5,withlessvariationintheoutputgapandmorevariationininflationthanwehaveactuallyexperiencedoverthepast20years.Putanotherway,higherinterestratesintherun-uptothefinancialcrisismighthave reduced the impact of the subsequent bust-at the cost of below-targetinflationandbelowtrendoutputbeforethecrisishit.3

Figure3.5

APreferableOutcome?

Inpractice,newmacroprudentialtoolsandbettermicroprudentialsupervisionwill improve the possibilities available to monetary policymakers. Havingadditional instruments in effect brings about a favorable shift in theMinsky-Taylor frontier (or surface), which defines the possibilities open topolicymakers.4

Nevertheless, consistent with the new remits given to the Monetary PolicyCommittee by theUK government lastmonth, the experience of recent yearssuggests there may be circumstances in which it is justified to aim off theinflationtargetforawhileinordertomoderatetheriskoffinancialcrises.

Instruments

For institutions generally regarded as conservative or even hidebound, centralbanks have been remarkably innovative in their creation of new instrumentsduring the crisis. Lowering official interest rates virtually to zero wasextraordinary in itself. But the wholesale redesign of frameworks to supplyliquiditytothebankingsystem,theexpansionofthemonetarybasebymultiples

ofitsprecrisislevelthroughthepurchaseofassetsheldbythenonbankprivatesector, and the involvement of central banks in risky credit easing operationshaveall raisedseriousquestionsabout theroleofcentralbanks-andhaveevenchallengedtheideaofcentralbankindependence.

Table3.1showsjusthowlargethescaleofcentralbankexpansionhasbeen.Themonetarybasehasrisenbyunprecedentedproportionsaswetriedtopreventacollapseofbroadmoneyandcredit,ashappened in theUnitedStatesduringthe Great Depression and is happening in Greece today. And central bankbalancesheetshaverisenacrosstheindustrializedworld,asshownintable3.2.Thatexpansionhas reflectedbothmoneycreation throughassetpurchasesandlendingagainstcollateral.Allmajorcentralbankshavecreatednewwaystolendagainst collateral.' Far from their image as conservative creatures, are centralbanksat riskof throwing their traditionalcaution to thewindand ignoring thelimitsonmonetarypolicy?

Table3.1

MonetaryBase(January2007=100)

Table3.2

SizeofCentralBankBalanceSheets

Two limitsare relevant incurrentcircumstances.First,nomatterhowmuchliquidity is thrown at the banking system, lending and the economy will notrecover if the banking system is inadequately capitalized and suffering fromexcessive leverage. That is why the Bank of England's Financial PolicyCommittee has placed weight on the need for the weaker UK banks to raisecapital.ItisnotsurprisingthatthemorestronglycapitalizedbanksintheUKareexpandinglendingandthepoorlycapitalizedbanksarecontractinglending.

Second,therearelimitsontheabilityofdomesticmonetarypolicytoexpandreal demand in the faceof the need for changes in the real equilibriumof theeconomy.IdonotbelievethatthepresentproblemsintheUKstemonlyfromalarge negative shock to aggregate demand. In common with many othercountries, theUK's problems also reflect the underlying need to rebalance theeconomy,requiringareallocationofresourcesbothwithinandbetweennations.Itisnotsimplyaquestionofboostingaggregatedemandbutofhelpingtobringaboutashifttoanewequilibrium.That,inturn,impliestheneedbothforlargechanges in relative prices, especially between tradable and nontradable goodsandservices,andashiftintherelativelevelsofdomesticdemandathomeandoverseas. There are, therefore, limits on what any one country's domesticmonetarypoliciescanachievewithoutthesupportofothers.

Despite these limits, circumstances have demanded that central banks takeextraordinary measures. Such measures can risk moving into territory morenormallyassociatedwithfiscalpolicyand,indoingso,putatriskcentralbanks'hard-wonindependence.Thereare,itseemstome,threethreatstocentralbank

independence.

First,thereistheriskofappearingtopromisetoomuchorallowingtoomuchto be expected of central banks.With constraints on other policy instruments,centralbanksareseenas"theonlygameintown."Butfailuretomakeclearthelimits to monetary policy risks disillusionment with central banks and theinevitablepoliticalpressureonthemthatwouldfollow.

Second,atthezerolowerboundthereisnocleardistinctionbetweenmonetaryandfiscalpolicy.Butitisstillimportanttoensurethatcentralbanksdonottakeon to their balance sheets risks to the taxpayer that are properly matters thatshould be decided by elected politicians. To ensure price stability in the longrun, it isvital tomaintain theoperational independenceofacentralbank.Anydecisionsthatputtaxpayers'moneyatriskmustbemadebyfinanceministries,and central banks must protect their balance sheets by imposing appropriatehaircutsoncollateralandavoidingthepurchaseofriskyprivate-sectorassets.

Third-andimportantevenwhenwemoveawayfromthezerolowerbound-theexpansion of central banks' responsibilities to include macroprudential policyand,inthecaseoftheBankofEngland,responsibilityforregulatingthebankingsystem has made independence much harder to define. The deployment ofresponsibilities outside monetary policy cannot be divorced from thegovernmentinthesamewayasispossibleformonetarypolicy.Forexample,intheareaoffinancialstabilityandbankingsupervision,therewillbetimeswhenpublicfundsmaybeputatriskwhenrescuingorresolvingafailinginstitution-and that decision is properly one for the finance ministry. It is far fromstraightforward for a central bank governor to be completely independent intermsofmonetarypolicy,somewhatindependentintermsoffinancialstability,andnotatallindependentintermsofoperationsthatrisktaxpayers'money.

The financial crisis has challenged our understanding of the objectives ofmonetary policy and exposed its limits. And, through the proliferation ofinstrumentsandtheresultingincreaseinresponsibilities,ithascomplicatedthequestion of central bank independence. How should we respond to this morecomplexenvironment?Wemustkeepsightof threeimportantprinciples.First,

although they should be realistic about what can be achieved, it is right thatelectedpoliticiansandparliamentsdecideontheobjectivesofpolicy.Second,aswe learned in the 1970s, if the central bank is to achieve price stability-itsfundamentalrole-thenitmustbesufficientlyindependent.Andthird,toprotectthat independence, its limits should be very clearly circumscribed, and centralbanks should be exceptionally careful with decisions that put public funds atrisk.

The challenge remains, as it was 20 years ago, to make "constraineddiscretion"workinpractice.Butithasgotharder.

Notes

IwouldliketothankCharlieBean,AlexBrazier,SpencerDale,AndyHaldane,and lain deWeymarn for helpful comments and suggestions in preparing thischapter,andinparticularTimTaylor,whomIregardasacoauthor,althoughheisabsolvedofanyerrorsinthecurrentdraft.

1. The inflation measure used is the quarterly rate of increase of the GDPdeflator, defined as the ratio of nominalGDP to realGDP.The output gap isestimatedasthedifferencebetweenthelogofrealGDPandthetrendpathoflogGDP,derivedusingtheHodrick-Prescottfilter.

2.HymanP.Minsky,"TheFinancial-InstabilityHypothesis:CapitalistProcessesand the Behavior of the Economy," in Financial Crises: Theory, History, andPolicy, ed. Charles Kindleberger and Jean-Pierre Laffargue (Cambridge:CambridgeUniversityPress,1982).

3.Amoredetailedappraisalofthisargumentiscontainedin"TwentyYearsofInflationTargeting,"a2012speechbyMervynA.King,availableathttp://www.bankofengland.co.uk/publicationsPagesspeeches/2012/606.aspx.

4. A social planner would use both instruments together to pick thewelfaremaximizing point on the frontier (or surface), and in the UK, theMonetary Policy Committee is expected to work closely with the FinancialPolicy Committee, which has the statutory power to deploy macroprudentialinstruments.

5. The relevant passage in the Monetary Policy Committee's remit states:"CircumstancesmayalsoariseinwhichattemptstokeepinflationattheinflationtargetcouldexacerbatethedevelopmentofimbalancesthattheFinancialPolicyCommittee may judge to represent a potential risk to financial stability. TheFinancialPolicyCommittee'smacroprudentialtoolsarethefirstlineofdefenceagainstsuchrisks,butinthesecircumstancestheCommitteemaywishtoallowinflationtodeviatefromthetargettemporarily,consistentwithitsneedtohaveregardtothepolicyactionsoftheFinancialPolicyCommittee."Thefullremitisavailable at http://www.bankofengland. co.uk/monetarypolicy/ Documents/pdf/chancellorletter130320r.pdf.

6. Facilities introduced by the Bank of England have included the DiscountWindow Facility, the Special Liquidity Scheme, Extended Collateral TermRepos, Extended Collateral Long Term Repos, and the Funding for LendingScheme. The ECB initiated its Long Term Repo Operations. And FederalReserve facilities include the TermAuction Facility, the TermAssert BackedSecurities Loan Facility, the Primary Dealer Credit Facility, the CommercialPaperFundingFacility,theTermSecuritiesLendingFacility,theABCPMoneyMarketFundLiquidityFacility, theMoneyMarket InvestingFundingFacility,andtheTermDiscountWindowProgram.

MichaelWoodford

During the global financial crisis and its aftermath, central banks haveundertakenunprecedentedactionsofmanykinds.Thisraisesanaturalquestion:has the crisis revealed that the previous consensus framework for monetarypolicy was inadequate and should now be fundamentally reconsidered? It issurely true thatcentralbankswerenot toowellpreparedfor thecrisisandthatnew policies had to be created, to a large extent on the fly. And it wouldobviouslybedesirabletotrytolearnfromthisexperience,tobebetterpreparedforanappropriateresponsenexttimeandperhapseventoreducetheprobabilityoftherebeinganexttime.

This does not mean that all previous conventional wisdom must now bediscarded. In particular, it has not been shown that central banks erred incommitting themselves to explicit, quantitative inflation targets. Inflationtargeting-andthe"implicit"inflationtargetingthatwaspracticedbysomeothercentralbanks-hasresulted inahighdegreeofstability inmedium-runinflationexpectationsduringthecrisisanditsaftermath,andthishaslikelyimprovedthestabilityoftherealeconomyaswell.Iftheprolongedhighunemploymentofthepast several years had led to a deflationary spiral, our situation would surelyhavebeenfarworse.ThathasbeenabenefitoftheFed'sandothercentralbanks'credibility in inflationstabilization in theyearsprior to thecrisis,and it isnotsomethingweshouldwanttocasuallydiscard.

Nevertheless, it is important to stress that inflation targeting need notmeanand shouldnotmean the caricature sometimesgivenof it, according towhichinflation control should be the sole objective of policy at all times becauseinflation stabilization by itself will be sufficient to guarantee macroeconomicstability. Recent events have obviously cast considerable doubt on this overlysimplisticview.However,it'simportanttorememberthatthiswasnottheviewadvocated by most proponents of inflation targeting even before the crisis.

MervynKingfamouslycalledthatviewthe"inflationnutter"positioninoneofhis classic early discussions of the theory of inflation targeting' and arguedinsteadforamoreflexibleformofinflationtargeting.OtherleadingproponentsofinflationtargetingsuchasBenBernankeandLarsSvenssonalsoconsistentlyargued for a flexible conception.2 They believed it was important to conductmonetarypolicyinsuchawayastomaintainmedium-runinflationexpectationsrelativelyconstantatthepreannouncedtargetratebutthatitwaspermissibletoallowtemporarydeparturesoftheinflationratefromthismedium-runtargetforthesakeofotherstabilizationobjectives.Anearterminflationratenearthetargetwasneithernecessarynorsufficientforgoodpolicy.

But this doctrine, while sensible as far as it goes, does leave an importantquestionunanswered:Whatdoes itmean to conductpolicy in the short run insuchawayastoensurethatmedium-runinflationexpectationsremainanchored,eventhoughoneisnotalwaysactingtokeepinflationascloseaspossibletothemedium-runtarget?Inflationtargetingcentralbankstalkalotabouthowtheytryto assess whether inflation expectations are still anchored and whether theirinternalmodels still forecastan inflation ratenear the target someyears in thefuture,buttheyarefrequentlylessclearaboutwhatitisaboutthewayinwhichtheyintendtomakepolicydecisionsthatwouldmakethatacorrectexpectation.

Vaguenessonthispointdidn'tcreategreatdifficultiesinthe15yearsorsoofrelativemacroeconomic stability prior to the global financial crisis. Butwhenlarger disturbances occur, the incompleteness of the flexible inflationtargetingdoctrine becomes more of a problem. Inflationtargeting central banks haverecentlybeenconductingpolicy inways thatdon'tseemtobedirectlydictatedbytheirinflationtargetingframework,butthatobservationraisesquestionsaboutwhethertheframeworkremainsineffect.

In my view, flexible inflation targeting doesn't need to be repudiated as apolicy framework, but it does need to be completed. Inflationtargeting centralbanksneedtocommitthemselvesnotonlytoamedium-runinflationtargetbutalso to criteria formakingnearer-termpolicy decisions thatwill, amongotherdesiderata,implythattheinflationrateshouldbenearthetargetifoneaveragesoverasufficientnumberofyears.'

As an example of such a criterion, a central bank might commit itself tomaking short-run decisions that keep nominal GDP as close as possible to aparticulartargetpath,eveninthenearerterm.ThetargetpathfornominalGDPcouldbechosensothatkeepingnominalGDPonthatpathshouldensure,overthemedium run, an average inflation rate equal to the inflation target.At thesame time, itwould imply that inflationwould not be the sole determinant ofshort-run policy decisions. For example, a loosening of policy might beappropriate evenwhen inflation is not runningbelow target if insufficient realgrowthhasresultedinalevelofnominalGDPbelowthetargetpath.4

Anotherrespectinwhichtheinflationtargetingdoctrinepriortothecrisishasproventobeincompleteisinitsfailuretospecifyhowpolicyshouldbecarriedout if aggregate demand remains insufficient to achieve the central bank'sstabilization targets, even when the zero lower bound on short-term nominalinterestratesisreached,asithasbeeninmanycountriesoverthelastfewyears.

Oneapproachusedbyseveralcentralbankshasbeen"forwardguidance,"orindicationsbythecentralbankthatinterestrateswillremainlowinthefuture,asa substitute for further immediate interest rate cuts. Such announcements doseemtohavebeenabletoinfluencemarketexpectationsaboutfutureshort-termratesandhenceabletoinfluencelonger-terminterestratesandotherassetprices.But important questions remain about the form that such forward guidanceshould take and how the existence of such statements should constrainsubsequentpolicydecisions.

Onequestioniswhetherforwardguidanceshouldtaketheformofastatementabout future policy intentions, or if it suffices for the central bank to offer aforecastofitslikelyfuturedecisions,giventheconditionsthatcanbeanticipatedat present. The idea ofmerely offering a forecast has had a certain appeal tocentral bankers since it doesn't tie the hands of the future policy committee.Unfortunately, there isnoobviousreasonforamereforecast tobeeffectiveinstimulatingdemand.

For a central bank forecast of future interest rates to change marketexpectations,itwouldhavetorevealeithernewinformationaboutlikelyfuture

conditions or new information about the central bank's future policy reactionfunction. But convincing people that interest rates will remain low for longerthantheyhadpreviouslyexpected-eitherbecausetheeconomicrecoverywillbeslowerthanpreviouslyexpectedorbecausedeflationiscoming,andnotbecauseofanychangeinthecentralbank'sreactionfunction-shouldbeexpectedtohaveacontractionaryratherthananexpansionaryeffectoncurrentexpenditure.5

Thisissurelynottheaimofforwardguidanceatthezerolowerbound.Hence,to be effective, the announcement must communicate a different view of thefuturereactionfunction;thatis,oftheconditionsunderwhichpolicywillorwillnotbetightenedinthefuture.

But if this is thegoal,amereforecastoffuture interestrates isnot themosteffectivewaytochangeexpectations.Ifacentralbankintendstoconductpolicylater inaway that isdifferent fromwhatpeople in themarketswouldalreadyexpect, then it should seek to communicate that intention by talking directlyabouthowfuturepolicydecisionswillbemade.

What kindof statements about futurepolicydecisionswouldbedesirable ifthatwereone'saim?Recently,severalcentralbankshavemadestatementsaboutspecific dates until which the policy rate is expected to remain at its lowerbound.Butwhiletradersandfinancialmarketsarecertainlyinterestedinhearingaboutsuchdates, Idon't thinkadate-basedapproachmakessenseasawayofcommunicatingfuturepolicyintentions.Itwouldnotmakesense,afterall,foracentralbanktoactuallybinditselfnottoconsiderraisingratesbeforeaspecificdateasfarastwoyearsinthefutureregardlessofwhatmayoccurintheinterim.Henceitishardfordate-basedforwardguidancetobeunderstoodasagenuinecommunicationofpolicyintentionsratherthanasamereforecast.

A better approach would instead specify economic conditions that must bereachedinorderforittobeappropriatetoraisethepolicyrate.Suchastatementshould allowmarket participants to form judgments about the likely length oftimeforwhichlowratesmightcontinue,butitwillimplythattheactualliftoffdatewoulddependonfutureoutcomes,asindeeditshould.

The recent move of the US Federal Open Market Committee (FOMC) toreplace date-based forward guidance with explicit numerical thresholds foreconomicindicatorsisadesirablestep.Thethresholds,however,havehadtobedetermined on an ad hoc basis and do not obviously follow from previouslyannouncedpolicytargets.NordotheyindicatethepolicythatoneshouldexpecttheFOMCtofollowafterthecurrentanomalousperiod.

Under the version of flexible inflation targeting that I've just proposed, thecriterionforliftofffromthezerolowerboundcouldfollowfromthesametargetcriterionthatguidespolicydecisionsatothertimes.AcentralbankthatseekstouseitspolicyinstrumentstokeepnominalGDPonacertain,steadygrowthpathcould also,when the zero lowerboundmakes it unable toprevent a sustainedshortfall of nominalGDP relative to the target path, commit itself tomaintainunusualpolicyaccommodationuntilnominalGDPcanbebroughtbackto thattargetpath,eventhoughthiswouldmeanseekinghigher thanaveragenominalgrowthduringatransitionalperiod.6

An approach of this kind to forward guidance during a zero lower boundepisodewould have two advantages over the ad hoc approach. First, itwouldprovide an explanation for pursuing unusually aggressive policies in theaftermathofazerolowerboundepisode,evenasamonetarystimulusbegantohaveeffects,anditwoulddothis inawaythatshouldnotcreatedoubtsaboutthecumulativeincreaseinprices thatmightoccurbefore thepolicyhasended,fortheexistenceofatargetpathforthelevelofnominalGDP-atargetthathasnot been raised as a result of the crisis-implies that nominal growth shouldindeed be capped. And the pursuit of such a temporary policy would remainperfectly consistentwith a stated intention topursue a subsequent approach topolicy (namely, keeping nominalGDP near that target path) that should onceagaindeliveranaverageinflationratenearthelong-runinflationtarget.

Thesecondadvantageisthatifsuchapolicywereexpectedtobefollowedassoon as the zero lower bound was reached, this anticipation should have astabilizingeffect,reducingthedistortionsassociatedwiththezerolowerboundoninterestrates.IfadeclineinnominalGDPgrowthowingtoaninabilitytocutthepolicyratebelowthezerolowerboundwereexpectedtoautomaticallyimply

fasternominalGDPgrowth later toundo the shortfall, this anticipation shouldreducethesizeofthatinitialshortfall.

Ibelievethatconfidencethatcentralbankswouldnotallowinflationtodriftpermanentlybelowtheirlong-runinflationtargetshasbeenastabilizingfactorinthe recent crisis. In the same way, I believe that had there been an existingcommitmenttoanominalGDPtargetpath,thiswouldhavebeenanevengreaterstabilizingfactor.

Finally, anotherquestion raisedby the recent crisis iswhether centralbanksshouldhavepaidmoreattentiontothegrowingriskstofinancialstabilitybeforethecrisis.Or,toposethemorepracticalquestionofrelevancetousnow,towhatextent should central banks consider risks to financial stability when makingmonetarypolicydecisionsgoingforward?

Certainly, this issue can't be dismissed as easily as it often was before thecrisis.Apopularargumentthenwasthatitwasdifficulttobesureabubblewasformingbefore itburstand that itwas thereforemorepracticalnot toconsiderthequestionuntilafterthecrash,andthenusemonetarypolicytodealwiththeconsequences of the crash. But surely the events of 2008-2009 and theiraftermathhavedentedourconfidenceregardinghoweasyitisto"mopupafterthecrash"withthetoolsactuallyavailabletocentralbanks.

It therefore makes sense going forward to seek to assess potential risks tofinancial stability before they grow too large, as difficult as that undoubtedlywillbe.Thisdoesnot,however,meanthatmonetarypolicyshouldbetheonlyline of defense. To say that monetary policy might have some capacity torestrainthegrowthofdangerousdegreesofleveragedoesn'timplythatnoothermeasures to restrain such developments should be needed, if only we had asoundmonetarypolicy.

Using monetary policy for this purpose, even under the assumption that itcould be fully effective, would surely have costs in terms of decreasing theextent to which monetary policy could simultaneously achieve its usualstabilization objectives. Hence, it behooves us to seek to improve financial

regulation, and to develop instruments of macroprudential policy as well.Multipleinstrumentsshouldincreasetheextenttowhichmultipleobjectivescansimultaneouslybepursued,andtheyaremuchtobedesiredinthiscase.

Still,inourcurrentsituation,withoutyethavingthesealternativepoliciesthatcan be relied on to fully eliminate the issue of controlling risks to financialstability,howshouldmonetarypolicytakeaccountoftheissue?Simplytrackingthe outlook formeasures of inflation and real activitywill not, in general, besufficient forsoundmonetarypolicydecisions. Itmaywellbe thatundermostcircumstances, risks to financial stabilitywillbesmallenoughunderallof thecurrently contemplated interest rate decisions for interest rate policy to be setpurelyon thebasisofexpectedconsequencesfor inflationandoutput.Butoneshouldatleastrecognizethepossibilityofexceptionstothatsituationandkeepaneyeoutforthem.

ThismeansthattheproposalI'vebeendescribing,thatinterestratepolicybeused tokeepnominalGDPona fixed targetpath, shouldnotbeviewedasanabsolute rule. It might well be reasonable, under some circumstances, tomaintaintighterpolicyinordertorestrainexcessivegrowthofleverage,evenifthis requiresnominalGDPtofallbelowthe targetpath.But thiswouldnot, inmyview,maketheexistenceofanominalGDPtargetpathpointless.

Inparticular,eveninthecaseofatemporarydeparturefromthenominalGDPtargetpathbecauseoffinancialstabilityconcerns,itmakessenseforthecentralbanktoremaincommittedtoeventuallyreachingthattargetpathagain,throughasubsequentperiodofhigherthanaveragenominalgrowthtomakeupfortheperiodofinsufficientnominalgrowth.7Theargumentisthesameasinthecaseof the zero lower bound: an expectation that current undershooting of thenominal GDP target path will subsequently be compensated by a period ofhigher nominal GDP growth should reduce the extent to which a temporarilyhighpolicyratecausesnominalGDPtoundershootinthefirstplace.

To the extent that such anticipation effects occur, they should reduce thetension between the goals of restraining risks to financial stability, on the onehand, and maintaining macroeconomic stability on the other. This is another

advantageofmodifying theunderstandingof flexible inflation targeting in themoreambitiouswayI'vejustsketched.

Notes

1.MervynA.King,"Changes inUKMonetaryPolicy:RulesandDiscretion inPractice,"JournalofMonetaryEconomics39(1997):81-97.

2. See, for example, Ben S.Bernanke and Frederic S.Mishkin, "InflationTargeting: A New Framework for Monetary Policy?," Journal of EconomicPerspectivesSpring (1997):97-116;BenS.Bernankeet al., InflationTargeting(Princeton,NJ:PrincetonUniversityPress,1999);LarsE.O.Svensson,"InflationTargeting:SomeExtensions,"Scandinavian Journal ofEconomics101 (1999):337-361;andidem,"InflationTargetingasaMonetaryPolicyRule,"JournalofMonetaryEconomics43(1999):607-654.

3. For further discussion of the desirability of such an intermediate targetcriterion,seeMichaelWoodford,InterestandPrices:FoundationsofaTheoryofMonetary Policy (Princeton, NJ: Princeton University Press, 2003), chap. 7;idem,"TheCaseforForecastTargetingasaMonetaryPolicyStrategy,"JournalofEconomicPerspectivesFall(2007):3-24;andidem,"ForecastTargetingasaMonetaryPolicyStrategy:PolicyRulesinPractice,"inTheTaylorRuleandtheTransformationofMonetaryPolicy,ed.E.EKoenig,R.Leeson,andJ.B.Taylor(Stanford,CA:HooverInstitutionPress,2012).

4.AnominalGDPtargetpathcanbeviewedasasimplerversionoftheproposalofatargetpathforan"output-gap-adjustedpricelevel,"aproposal thatcanbeshown to represent a theoretical ideal in certain New Keynesian models, asdiscussed in Michael Woodford, "Optimal Monetary Stabilization Policy," inHandbook of Monetary Economics, vol. 3B, ed. B.M.Friedman andM.Woodford (Amsterdam: Elsevier, 2011). On the advantages of a nominalGDP target path as a practical proposal, seeMichaelWoodford, "Methods ofPolicy Accommodation at the Interest-Rate Lower Bound," in The ChangingPolicyLandscape (KansasCity: FederalReserveBank ofKansasCity, 2012);andidem,"InflationTargeting:FixIt,Don'tScrapIt," inIsInflationTargetingDead?CentralBankingaftertheCrisis,ed.L.ReichlinandR.Baldwin(London:CentreforEconomicPolicyResearch,2013).

5.Forfurtherdiscussion,seeWoodford,"MethodsofPolicyAccommodation."

6.ThewaythiswouldworkandtheadvantagesofthisproposaloverthekindofthresholdsintroducedbytheFOMCarediscussedfurtherinMichaelWoodford,"ForwardGuidancebyInflationTargetingCentralBanks,"http://www.columbia.edu/-mw2230/RiksbankIT.pdf.

7.SeeMichaelWoodford,"InflationTargetingandFinancialStability,"SverigesRiksbankEconomicReview1(2012):7-32,forananalyticaldemonstrationthataruleofthiskindrepresentsanoptimalpolicycommitmentinthecontextofasimplenewKeynesianmodelwithatrade-offamongthethreegoalsofinflationstabilization, output gap stabilization, and minimization of the economicdistortionsassociatedwith financialcrises.Theruleshown there tobeoptimalgeneralizes the optimal policy rule for a new Keynesian model that abstractsfrom endogenous risks to financial stability; thus, it actually involves a targetpathforanoutput-gap-adjustedpricelevelratherthanfornominalGDP.Underthe optimal commitment, the deterministic target path for this variable is notshiftedbyvariationsinfinancialriskorbytheoccurrenceoffinancialcrises,buttemporarydeparturesfromthetargetpatharejustifiedinproportiontovariationsin a "marginal crisis risk"variable.Thevariantwith a constant targetpath fornominalGDPisintendedasanapproximationoftheoptimalrulederivedthere(whichis,however,onlyexactlyoptimalunderquitespecificassumptions).

AndrewHaldane

Macroprudentialpolicy is thenewkidon theblock,perhapseven thenextbigthing.Hopes are high.Reflecting that,wehavenewmacroprudential agenciesandpolicies poppingup all over theworld, in both developed anddevelopingeconomies (see, e.g.,Aikman,Haldane, andKapadia 2013).But that begs thequestion:Whatactuallyismacroprudentialpolicy?Howshoulditbeexecuted?Andhoweffectivewillitbe?

Thissessioniswellpositionedtoanswerthesequestions,comingaftertheoneonmonetarypolicy,becausetherearedirectparallelswith,andlessonsthatcanbelearnedfrom,monetarypolicyinthedesignofamacroprudentialframework.

Thestateofknowledgeaboutmacroprudentialregimestodayisroughlywheremonetarypolicywasinthe '40s-andifIambeingcharitable,thatwouldbethe1940sratherthanthe1840s.Itiseasytoforgetthat70yearsagotherewasstillagreat deal of uncertainty about the key tenets of an optimal monetary policyframework:

•Whatwereappropriateobjectives?

•Whatinstrumentsshouldbestbedeployed?

•Whatwastheappropriategovernanceandaccountabilityframework?

Since then, all three of those design features have been, if not resolved, thenmuchbetterarticulatedformonetarypolicyframeworks.

Whenitcomestomacroprudentialpolicyframeworks,allthreefeaturesare,ifnot undefined, then poorly articulated at present. Letme touch briefly on theeachoftheminturn.

First,objectives.Macroprudentialis,asitsnameimplies,abouttheinteractionorinterfacebetweenprudentialpolicyandthemacroeconomy.Buthowexactlydoes that translate into a macroprudential objective? In particular, should wethinkofmacroprudentialobjectivesasbeing:

•Toprotect thefinancialsystemfromswingsandcycles in therealeconomy?Thatwouldgivemacroprudentialpoliciesapuristfinancialstabilityobjective.

Or,moreambitiously,aremacroprudentialpolicies:

•Toprotect therealeconomyfromswingsandcycles in thefinancialsystem?Thatwouldgivemacroprudentialpoliciesamoreovertlymacroeconomicfocus.

Putdifferently,shouldmacroprudentialpolicybeaboutprovidingpower to theelbow of the microprudential supervisor? Or is macroprudential policy alegitimatelydistinctarmofmacroeconomicpolicy,workingalongsidemonetaryandfiscalpolicy?

This question has a direct parallel with the monetary policy debate aboutappropriatemandates, a debate that to someextent still exists today (see, e.g.,Woodford2012).Whatistheappropriatebalancebetweenapuristpricestabilitymandate, on the one hand, and a dual mandate that also weighs output andemploymentobjectiveson theother?Mostcountries'monetarymandates thesedaystendtoweighwider(thanpricestability)objectives.

Existing international experience suggests some important differences in thescope of macroprudential mandates. For example, the recent US stress testsfocusedontheimplicationsofaseveremacroeconomicdownturnforUSbanks'financial resilience. In other words, stress testing provided a macroprudentialoverlay to-or power to the elbow ofmicroprudential supervision (Bernanke2013).Thisisconsistentwithatype1macroprudentialmandate.

Bycontrast,intheUKtheBankofEngland'snewFinancialPolicyCommittee(FPC) has a statutory macroprudential mandate with a clear lexicographicorderingofobjectives.Thisplacesfinancialresilienceastheprimaryobjective,but thenweighsoutputandemploymentstabilizationasasecondaryobjective.

Thereisadualbutorderedmandate-atype2mandate.

Thatmandate has had an important bearing on the actions of the FPC. Forexample, in the middle of 2012 the FPC reduced UK banks' liquidityrequirements. This was an overtly countercyclical attempt to stimulate banklendingandoutputinthewidereconomy.MorerecentlytheFPChasaskedUKbankstoboosttheircapital,withanexpliciteyetoloweringfundingandlendingrates and thereby supporting the economy. In other words, macroprudentialpolicyhasoperatedasanextraarmofmacroeconomicpolicy.

ThemacroprudentialactionsrecentlyundertakenbyBrazil,HongKongSAR,India,Korea,andIsrael,amongothercountries,arealsoconsistentwithatype2macroprudential mandate. In each case the objective appears to have been tomodulate fluctuations in asset markets-for example, the housing market-andtherebythewidereconomy.Sufficeittosay,internationallythejuryisstillouton appropriate objectives for macroprudential policies and their scope andambition.

Second, instruments.Hereagain there is aparallelwith themonetarypolicydebates of a couple of generations ago. Back then, monetary theorists andpractitioners actively debated the relative merits of price-based instruments(such as the setting of short-term interest rates) and quantitybased instruments(suchas thesettingof targets forbasemoneysupplyor selectiveuseofcreditandcapitalflowcontrols).Inthetimesince,thatdebatehassettledfirmlyonthesideofprice-basedinstruments,certainlyduringnormaltimes.

Today,onekeystrandof themacroprudentialdebate iswhether it shouldbeexecuted using price-based instruments (such as by setting of capital andliquidity ratios or by taxing certain financial transactions) or quantitybasedmeasures(suchasbysettingloan-to-value[LTV]ordebtto-income[DTI]limitsfor mortgages, or margin requirements for secured financing transactions)-orindeed,byacombinedtoolkitofboth.

Existing international macroprudential practices differ sharply on thisquestion.A recent surveybyLimandothers (2011)makes clear the extent of

these differences, summarized in table 5.1. In theUK, the Bank of England'sFPChasexpressedapreferenceforusingcapitalandliquiditytools-price-basedinstruments-at least when it issues directions to other regulators (Bank ofEngland2013).Thiswaspartlyonthegroundsthatprice-basedinstrumentswilltendtobelessdistortionaryintheirimpactonbehavior.

Ontheotherhand,anumberofemergingmarketshaveinsteadusedLTVorDTI interventions-quantity-based interventions-as their macroprudentialinstrumentofchoice(table5.1).Thiswasdonepartlyonthegroundsthatthesemeasuresarelikelytohaveamoredirectandimmediateimpacton,forexample,thehousingmarketorcreditorflows.Anumberofcountrieshaveemployedbothprice - and quantitybased macroprudential instruments. Once again, the jury,academicallyandpractically,remainsout.

Table5.1

Cross-CountryUseofMacroprudentialInstruments(2000-2010)

Third, governance and accountability.Here there is an interestingdifferencebetweenmonetaryandmacroprudentialpolicymandates.Monetarypolicyisnot,byitsnatureandinnormaltimes,granular;itis,ifnotblind,thenblindfoldedtoitsdistributionalconsequences.Centralbankscannotsetdifferent interest ratesfor the north versus south of a currency area, for those with and without amortgage,forsmallfirmsversuslarge.

Macroprudentialpoliciescandojustthat-andinpracticesometimeshave.Forexample,LTVorDTIinterventionshavesometimesbeentargetedatparticularregionsorcitiesorloantypes.Inthatrole,theyareovertlydistributionalintheir

impactand,atleastinprinciple,almostinfinitelygranular.Whilethiscanhavebenefitsintargetingselectiveareasofriskforpolicyintervention,thiscomesatapoliticaleconomycost.Becauseoftheirdistributionalimpact,macroprudentialpolicies raise questions about appropriate governance and democraticaccountability.

These tensions are clearly evident in existing governance structures formacroprudentialpolicyaroundtheworld.Table5.2providesasummaryofthegovernancearrangementsformacroprudentialpolicyinaselectionofcountries.Some regimes have placed the central bank in the driver's seat, as in theUK.OthershavetheleadroleplayedinsteadbytheFinanceMinistryorTreasuryaspartofacollegeofregulators,asintheUnitedStates.

Table5.2

MacroprudentialDecision-makingFrameworksbyCountry

Even when in the driver's seat, a central bank assuming macroprudentialresponsibilities is likely to face additional pressures on its independence. Alargersetofpowers,whichhaveadirectimpactonthedistributionandlevelsofGDP, raise important questions about democratic legitimacy. Accountabilitypracticesmayneedtoberatchetedupwardaccordingly.Forcentralbanks, thatmay be a price worth paying for having an extra degree of macroeconomicfreedom,butitisapricenonetheless.

Resolvingallofthesemacroprudentialframeworkquestionswilltaketime:foracademics, time to conduct research on the efficacy and design ofmacroprudential policies; for policymakers, time to execute and adapt thesepoliciesinthelightofexperience.Aswithmonetarypolicyhalfacenturyago,bothacademicsandpolicymakershaveanimportantroletoplayindeveloping

an intellectually coherent yet operationally practicable framework formacroprudentialpolicy.Doubtless,aswithmonetarypolicy,thiswillbeaslow,evolutionary process of trial and error. But the biggest error we could makewouldbetonottry.

References

Aikman, D., A.G.Haldane, and S.Kapadia. 2013. "Operationalising aMacroprudentialRegime:Goals,ToolsandOpenIssues."EstabilidadFinanciera[Financial Stability Journal of the Bank of Spain] 24:10-30.http://www.bde.es/f/webbde/ GAP/Secciones/Publicaciones/InformesBoletinesRevistas/ RevistaEstabilidad Financiera/13/Mayo/Fic/ref2013241.pdf.

Bank of England. 2013. "The Financial Policy Committee's Powers toSupplement Capital Requirements: A Draft Policy Statement."http://www.bankofengland .co.uk/financialstability/Documents/fpc/policystatement130114.pdf.

Bernanke,B.S.2013."StressTestingBanks:WhatHaveWeLearned?"Paperpresented at the Federal Reserve Bank of Atlanta Conference, "MaintainingFinancial Stability: Holding a Tiger by the Tail," April.http://www.federalreserve.gov/newsevents/speech/bernanke20l30408a.pdf.

Lim, C. H., E Columba, A.Costa, P.Kongsamut, A.Otani,M.Saiyid, T.Wezel,and X.Wu. 2011. "Macroprudential Policy: What Instruments and How AreThey Used? Lessons from Country Experiences." IMF Working Paper 238,InternationalMonetaryFund,Washington,DC.

Woodford,M. 2012. "Methods of PolicyAccommodation at the Interest-RateLowerBound."PaperpresentedattheJacksonHoleSymposium,August.http://www.columbia.edu/-mw2230/JHole20l2final.pdf.

ClaudioBorio

This chapter seeks to provide some context for the macroprudential policydebate.TheobjectiveistoexplorewhatIconsiderthemajorsourceofsystemicrisk, namely, the financial cycle and its linkwith systemic financial (banking)crisesand the farbetterknownbusinesscycle. Iwould like tohighlighta fewstylizedfactsandthenturntotheimplicationsformacroprudentialpolicy.

By"financialcycle"Imean,somewhatloosely,theself-reinforcinginteractionbetween risk perceptions and risk tolerance, on the one hand, and financingconstraintsontheotherthat,asexperienceindicates,canleadtoseriousepisodesoffinancialdistressandmacroeconomicdislocations.Thisiswhathasalsocometobeknownastheprocyclicalityofthefinancialsystem.

There are three takeaways from my presentation. First, the financial cycleshouldbeat theverycoreofourunderstandingof themacroeconomy.Tomymind, macroeconomics without the financial cycle is very much like Hamletwithout the prince. Second, the financial cycle has significant implications forthe design and limits of macroprudential policy. And finally, it also hassignificant implications for the design and limits of other policies, notablymonetaryandfiscal.

Iwilladdress twoquestions in turn.Whatare thepropertiesof the financialcycle?Iwillhighlightseven.Whatarethepolicyissuesitraises?Iwillhighlightfour.

IshouldmentionthatwhatIwillbepresentingisbasedonresearchcarriedoutattheBankforInternationalSettlements(BIS)overtheyears.Butmanyofthe

findings are quite consistentwithwork carried out elsewhere, including at theIME'

TheFinancialCycle:SevenProperties

Itisusefultothinkofthefinancialcycleashavingsevenproperties.2

First, itsmost parsimonious description is in terms of the joint behavior ofcreditandpropertyprices.Insomerespects,equityprices,whilesoprominentinfinanceandmacroeconomics,canbeadistraction.Thisinturnisrelatedtothenextproperty.

Second, the financial cycle has amuch lower frequency than the traditionalbusiness cycle.' By "traditional business cycle" I mean how economists andpolicymakersthinkofthebusinesscycleandmeasureit.Thisbusinesscyclehasa duration of up to eight years. By contrast, the financial cycle that is mostrelevantforseriousmacroeconomicdislocationshassincetheearly1980shadadurationof16to20years.Itisamedium-termprocess.Thisiswhatwefoundinasampleofsevenadvancedeconomiesforwhichwehadgooddata(Drehmann,Borio, and Tsatsaronis 2012). The point is illustrated for theUnited States infigure6.1.

Figure6.1

FinancialandBusinessCyclesintheUnitedStates.

Note:Darkgrayandlightgraybarsindicatepeaksandtroughsofthecombinedcycle using the turning-point method. The frequency-based cycle (continuousline)istheaverageofthemedium-termcycleincredit, thecredittoGDPratioand house prices (frequency based filters). The short-termGDP cycle (dashedline)isthecycleidentifiedbythetraditionalshort-termfrequencyfilterusedtomeasurethebusinesscycle.ShadedareasindicateNBERrecessions.

Note:Theamplitudesofcyclesdrawnbythedottedandcontinuouslinesarenotdirectlycomparable.NBER,NationalBureauofEconomicResearch.

Source:Drehmann,Borio,andTsatsaronis(2012).

The figureplots the traditionalbusiness cycle (dotted line) and the financialcycle (continuous line) asmeasured throughband-pass filters aswell aspeaksand troughs (vertical lines).The financial cycle is identifiedby combining thebehavior of credit, property prices, and the ratio of credit to GDP.4 Thedifferenceindurationisobvious.

Equitiescanbeadistraction in thesense that their time-seriespropertiesareclosertothoseofGDPintermsofthedurationofswings.Forexample,thestockmarketcrashesof1987and2000werefollowedbyslowdownsinGDPgrowthoroutrightrecessions.Butthefinancialcycleasmeasuredbycreditandpropertypricescontinuedtoexpand,onlytoturnafewyearslater(early1990sand2007-2008, respectively), bringing the economy down and causing even greaterdamage.Seenfromthislonger-termperspective,theearlycontractionphasesineconomic activity can thus be considered "unfinished recessions" (Drehmann,Borio,andTsatsaronis2012).

Third, peaks in the financial cycle tend to coincide with systemic bankingcrisesorseriousstrains.Thiswastrueforallcrisesafter1985inthesampleofadvancedcountriesweexamined(Drehmann,Borio,andTsatsaronis2012).Andthosecrisesthatoccurredwellawayfromthepeakwere"imported";thatis,theyreflectedlossesoncross-borderexposurestofinancialcycleselsewhere.Irecallhere,forinstance,thelossesthatGermanandSwissbanksincurredinthemostrecent financial crisis from their exposure to the US financial cycle. Notsurprisingly,businesscyclecontractionsthatcoincidewithabustinthefinancial

cyclearemuchdeeper.

Fourth, thanks to the financial cycle, simple leading indicators can identifyrisksofbanking crises fairlywell in real time (ex ante)' andwith agood lead(between two to fouryears,dependingoncalibration).The indicatorswehavefoundmostusefulattheBISarebasedonthepositivedeviationsofthe(privatesector)credit-to-GDPratioandofassetprices,especiallypropertyprices,jointlyexceeding their respective historical trends (e.g., Borio and Drehmann 2009;BorioandLowe2002).6Wemaythinkoftheseindicatorsasreal-timeproxiesforthebuildupoffinancialimbalances:Thedeviationsofassetpricesprovideasenseofthelikelihoodandsizeofthesubsequentreversal;thoseofthecredit-to-GDPratioprovideasenseoftheloss-absorptioncapacityofthesystem.TheseindicatorsflashedredintheUnitedStatesinthemid-2000s(figure6.2).

Figure6.2

Leading Indicators ofBankingCrises:Credit andPropertyPriceGaps,UnitedStates.

Note:The shaded areas refer to the threshold values for the indicators: 2 to 6percentagepointsforthecredit-to-GDPgap(a),and15to25percentfortherealpropertyprice gap (b).The real property price gap is theweighted averageofresidential and commercial property prices with weights corresponding toestimates of their share in overall property wealth. The legend refers to the

residentialpropertypricecomponent.Theestimatesfor2008(aandb)arebasedon partial data (up to the third quarter). OFHEO, Office of Federal HousingEnterpriseOversight.

Source:BorioandDrehmann(2009).

Thereisalsogrowingevidencethatcross-bordercreditoftenoutpacespurelydomestic credit during such financial booms (e.g., Borio, McCauley, andMcGuire2011;Avdjiev,McCauley,andMcGuire2012).'

Fifth, and for much the same reasons, the financial cycle also helps inconstructing estimates of sustainable output that, compared with traditionalpotential output estimates, are much more reliable in real time, as well asstatistically more precise (Borio, Disyatat, and Juselius 2013). None of thecurrent methods, ranging from full-fledged production function approaches tosimple statistical filters, spotted that outputwas above its potential sustainablelevel ahead of the financial crisis. In recent work, we have found thatincorporatinginformationaboutthebehaviorofcreditandpropertypricesallowsustodojustthat.

Figure6.3illustratesthisfortheUnitedStatesbycomparingourestimatesofthe output gaps (the so-called "finance-neutral" gap)with those from the IMFandOECD, based onmore full-fledgedmodel approaches andwith a popularstatisticalfilter(theHodrick-Prescottfilter).

Figure6.3

USOutputGaps:Full-Sample (expost)andReal-Time(exante)Estimates (inpercentagepointsofpotentialoutput).

Notes: a. International Monetary Fund. b. Organisation for Economic Co-operation and Development. c. Hodrick-Prescott filter. d. Finance-neutralestimateofBorioandothers(2013).Thesearelinearestimatesoffinancialgaps;thenonlinear estimates,which shouldbetter capture the forcesatwork,wouldshowanoutputgap that is considerably larger in theboomand smaller in thebust.

Source:Borio,Disyatat,andJuselius(2013).

The traditional estimates made in real time during the economic expansionthat preceded the crisis indicated that the economy was running below, or atmost close to, potential (continuous lines in the corresponding panels). Onlyafter the crisis did the estimates reveal, though tovaryingdegrees, that outputhad been above its potential sustainable level (dotted lines). By contrast, thefinance-neutral measure sees this all along (panel d in figure 6.3, continuousline).Andithardlygetsrevisedastimeunfolds(thecontinuousanddottedlinesarepersistentlyveryclosetoeachotherinpaneld).Onereasonwhyproductionfunction and similar approachesmiss the unsustainable expansion is that they

drawonthenotionthatinflationistheonlysignalofunsustainability.But,asweknow,aheadof thecrisis itwas thebehaviorofcreditandpropertyprices thatsignaled that outputwason anunsustainable path; inflation remained lowandstable.

Sixth,andcritically,theamplitudeandlengthofthefinancialcycleareregimedependent: They are not and cannot be a kind of cosmological constant.Arguably, three key factors support financial cycles: financial liberalization,whichweakens financing constraints;monetary policy frameworks focused onnear-term inflation control, which provide less resistance to the buildup offinancial imbalances as long as inflation remains low and stable; and positivesupply-sidedevelopments (e.g., theglobalizationof real economy),which fuelthe financial boom while at the same time putting downward pressure oninflation.Itisnotacoincidence,therefore,thatfinancialcycleshavedoubledinlength since financial liberalization in the early and mid-1980s and that theyhavebecomeespeciallyvirulentsincetheearly1990s(seefigure6.1).

Finally, busts of financial cycles go hand-in-hand with balance sheetrecessions.'Inthiscase,comparedwithotherrecessions,debtsandcapitalstockoverhangsaremuch larger, thedamage to the financial sector ismuchgreater,and the policy room for maneuver is much more limited as policy buffers-prudential, monetary, and fiscal-get depleted. Evidence also indicates thatbalancesheetrecessionsresultinpermanentoutputlosses(growthmayreturntoitslong-runprecrisisrate,butoutputdoesnotregainitsprecrisistrajectory)andusher in slowand long recoveries.Why? I suspect it reflects the legacyof thepreviousboomandthesubsequentfinancialstrains.

TheFinancialCycle:FourObservationsaboutMacroprudentialPolicy

How should prudential policy address the financial cycle?The financial cyclerequiresthatprudentialpolicyhaveasystemic,ormacroprudential,orientation.This means addressing the procyclicality of the financial system head-on, orwhathas come tobeknownas the timedimensionofmacroprudential policy;this is the dimension that relates to how systemwide or systemic risk evolvesovertime(e.g.,Crockett2000;Borio2011;Caruana2012a).9

The general principle is quite simple to describe but quite difficult toimplement: it is tobuildupbuffersduringfinancialboomssoas todrawthemdown during busts. This has two objectives. It would make the systemmoreresilient andbetter able towithstand thebust.And, ideally, itwould constrainthe financial boom in the first place, thereby reducing at its source theprobability and intensity of the bust. Note that these two objectives are verydifferent;thesecondismuchmoreambitiousthanthefirst.Iwillreturntothispointlater.

Letmenowhighlight four observations aboutmacroprudential policy.Theyare all intended to manage expectations about its effectiveness and to set arealisticbenchmarkaboutwhatitcanandcannotdo-andthisfromsomeonewhohasbeenastrongadvocateoftheapproachformorethanadecadenowandwhocontinuestobeone.Thereasonisthatthefinancialcycleisahugelypowerfulforce.

FirstObservation

Bewareofmacrostress testsasearlywarningdevices in tranquil times(Borio,Drehmann,andTsatsaronis2012).Tothebestofmyknowledge,noneofthemflashed red ahead of the recent crisis.10 Their relentless message was "thesystemissound."

Therearetworeasonsforthis.

Thefirsthastodowithourriskmeasurementtechnology.Ourcurrentmodelsareunabletocaptureconvincinglythefundamentalnonlinearitiesandassociatedfeedback effects that are at the core of the dynamics of financial distress. Inessence,nomatterhowhardyou shake thebox, little fallsout.This shifts theburden to the required size of the shocks, which becomes unreasonably largeand,therefore,isdiscountedbypolicymakers.Thedeeperpointhereisthattheessenceof financial instability is thatnormal-sized shockscause the system tobreakdown.Anunstablesystemisnotonethatwouldbreakdownonlyifhitbyahugeshock, suchasanoutsized recession.Anunstable system is fragile.Asempiricalevidenceindicates,crisesbreakoutclose to thepeakof thefinancial

cycle,wellbeforeGDPhasplunged intoadeep recessionorassetpriceshavecollapsed.

The second reason has to dowith the context, or whatmight be called the"paradoxof financial instability" (Borio2011). Initial conditionsareunusuallystrong just before financial strains emerge. Credit and asset prices have beensurgingahead;leveragemeasuredatmarketpricesisartificiallylow;profitsandassetqualitylookespeciallyhealthy;andriskpremiaandshort-termvolatilitiesareextraordinarilycompressed.Takenat facevalue, thesesignalspoint to lowriskwheninfact theyaresignsofhighrisk taking.Thesystemismostfragilewhen it looks strongest. And this point is reached after years of solid andrelentless expansion, typically alongside widespread financial innovations.Undertheseconditions,thetemptationtobelievethatthistimethingsarereallydifferentisextraordinarilypowerful(ReinhartandRogoff2009).

Bottom line: at worst, macro stress tests can lull policymakers into a falsesenseofsecurity.Thatsaid, ifproperlydesigned, theycanbeaneffective toolfor crisis management and resolution-a tool to promote balance sheet repair.After all, the crisis has already broken out, nonlinearities have revealedthemselves,andhubrishasgivenway toprudence. "Properlydesigned"meansthattheauthoritiesneedtohavethewilltoshakethesystemhard,needtostartthe tests from very realistic asset valuations, and should put in place thenecessaryliquidityandsolvencybackups.

SecondObservation

Bewareofnetworkanalysisasatooltodetectvulnerabilities(Borio,Drehmann,and Tsatsaronis 2012). As a source of vulnerabilities, bilateral links(counterpartyexposures)matterfarlessthancommonexposurestothefinancialcycle.

Networkanalysisviewsthefinancialsystemasawebofconnectionslinkinginstitutions.It thenmodelssystemicriskbytracingtheknock-oneffectsof thedefaultofoneinstitutionontherestalongthoseinterconnections.Thelargertheportionofthesystemthatfails,thelargeristhesystemicrisk.

Themainproblemisthat,asempiricalevidenceconfirms,giventhesizeoftheinterconnections, it is too hard to get large effects. The reason is simple:mechanical exercises abstract from behavior. A financial crisis ismore like atsunamithatsweepsawayallthatgetsinitswaythanitislikeaforceknockingdown one domino after another. The main force driving it is indiscriminatebehavioralresponses.Thisalsoexplainswhythefailureofsmallandseeminglyinnocuousinstitutionscantriggeramajorcrisis.Smallinstitutionsdonotmatterbecauseofwhat theyarebutbecauseofwhat they signal about the rest.Theysignal shared vulnerabilities; they are the canary in the coal mine.When thefinancial cycle turns, the failure of the first institution can shake previouslyseeminglyunshakableconvictionsandtriggeraparadigmshift.

That said, thisdoesnot imply that informationaboutbilateral exposureshaslittle value. Much like macro stress tests, it can be very valuable in crisismanagementasatooltoidentifypressurepointsandunderstandwhereandhowbest to intervene. But for this to be the case, the information has to be quitegranularandveryuptodate(Borio2013).

ThirdObservation

Beware of overestimating the effectiveness of macro prudential policy (Borio2011;Caruana2012a).Therearetwosetsofreasonshereaswell,whichinsomewaysechothosethatexplainthelimitationsofstresstests.

The first set is technical. The tools are more effective in strengthening theresilienceof the financial system (the first objectivementioned above) than inconstrainingfinancialbooms(thesecondandmoreambitiousobjective).Tobesure,someinstrumentsaremoreeffectivethanothers.Forinstance,itstandstoreason, and it seems to be confirmed by empirical evidence, that ceilings onloan-to-valueanddebt-to-incomeratioshavemorebitethancapitalrequirements(e.g.,CGFS2012).Afterall,capitalischeapandplentifulduringbooms.Butfortypicalcalibrationofthetools,itwouldbeimprudenttoexpectastrongimpact.Moreover, and critically, all such tools are vulnerable to regulatory arbitrage.Andthelongertheystayinplace,theeasierarbitragebecomes.

Thesecondsetof reasonshas todowithpoliticaleconomy.Comparedwithmonetarypolicy,it isevenhardertotakeawaythepunchbowlwhenthepartygetsgoing.Thelagsbetweenthebuildupofriskanditsmaterializationareverylong,certainly longer thanthosebetweenexcessdemandandinflation(Irecallherehowlongthefinancialcycleiscomparedwiththebusinesscycle).Forsomeofthetools,thedistributionaleffectsaremoreprominentandconcentrated.Andwhile there is a constituency against inflation, there is hardly any against theinebriating feeling of getting richer. All this puts a premium on soundgovernancearrangementsandonmaintainingarightbalancebetweenrulesanddiscretion.

FourthObservation

Beware of overburdeningmacroprudential policy (e.g., Caruana 2011, 2012b;Borio2012a,2012b).Thisfollowsnaturallyfromthepreviousobservation.Thefinancial cycle is simply too powerful to be tackled exclusively throughmacroprudentialpolicyor indeedprudentialpolicymoregenerally,be itmicroormacro.Macroprudentialpolicyneedstheactivesupportofotherpolicies.

What does this mean in practice? For monetary policy, it means leaningagainst thebuildupof financial imbalanceseven ifnear-term inflation remainsundercontrol(exercisingthe"leanoption")."Monetarypolicysetstheuniversalpriceofleverageinagivencurrency.Incontrasttomacroprudentialtools, it isnotvulnerabletoregulatoryarbitrage:youcanrunbutyoucan'thide.Forfiscalpolicy,itmeansbeingextra-prudent,recognizingthehugelyflatteringeffectoffinancialboomsonthefiscalaccounts.Thisisbecauseoftheoverestimationofpotentialoutputandgrowth(seefigure6.3),therevenue-richnatureoffinancialboomsowing to compositional effects, and the contingent liabilities needed toaddressthesubsequentbust.

As an important aside, a big open question is how macroprudentialframeworks should address sovereign risk. These frameworks were originallydesignedwithprivate-sectorvulnerabilitiesinmind,linkedtothefinancialcycle.Butsuchcyclesleaveintheirwakeseriouslydamagedsovereigns,whichcanalltoo easily sap banks' strength.Moreover, as history indicates, sovereignsmay

cause banking crises quite independently of private-sector excesses.At a timewhen the sovereigns' creditworthiness is increasingly in doubt, much moreattentionshouldbedevotedtothisissue.

Against this broad backdrop, is there a risk that adjustments in policyframeworksarefallingshort?Ostensibly,thiswasthecasebeforethecrisis,butwhataboutsincethen?Myansweristhattheriskshouldnotbeunderestimated.

Progresshasbeenunevenacrosspolicies(Borio2012b).

Prudential policy has adjusted most. A major shift from a micro - to amacroprudentialorientationhastakenplaceinregulationandsupervision.HerewemayrecalltheadoptionofacountercyclicalcapitalbufferinBaselIII(BCBS2010;Drehmann,Borio,andTsatsaronis2011)and,moregenerally, theeffortsunder way to implement full-fledged macroprudential frameworks around theworld (CGFS2012).That said, expectations aboutwhat these frameworks candeliverarerunningtoohigh,andthereisaquestionofwhetherenoughhasbeendone with respect to instruments, their calibration, and governancearrangements.Moreover,morecouldandshouldhavebeendonetorepairbanks'balancesheetsinsomejurisdictions.

Monetary policy has adjusted less. To be sure, there has been some shifttowardadoptingtheleanoption,butthewilltoexerciseithasbeenquitelimited.The temptation to rely exclusively on the newmacroprudential tools has beenvery powerful, to avoid disturbing monetary policy. And it is worth askingwhether the limitationsofmonetarypolicyasameans to tacklefinancialbustshavebeenfullyappreciated.

Fiscalpolicyhasadjustedleast.Thereisasyetlittlerecognitionofthehugelyflatteringeffectsof financialboomson the fiscalaccountsandof thebig risksthatbustsposeforthesustainabilityandeventheeffectivenessoffiscalpolicy.

Bottom line: there is a real risk that policies are not sufficiently mutuallysupportive.And,critically,theyarenotsufficientlysymmetricbetweenfinancialboomsandbusts.Theytightentoolittleduringbooms,withtheseriousdangerthatbuffersgetdepletedduringbusts.Thisposesahugeconstraintontheroom

formaneuver-one thatbecomes tighterover successivecycles.Policyhorizonsaresimply tooshort-notcommensuratewith thedurationof the financialcycle(Borio2012b).

Conclusion

There is a need to bring the financial cycle back into macroeconomics.Macroeconomics without the financial cycle is very like Hamlet without theprince. This raises huge analytical challenges that the profession is justbeginningtotackle.

The financial cycle has major implications for macroprudential policy andbeyond. I have highlighted four observations:Beware ofmacro stress tests asearlywarningdevices.Bewareofnetworkanalysisasatooltoidentifyfinancialvulnerabilities.Bewareofthelimitationsofmacroprudentialpolicy.Andbewareofoverburdeningit.

Hasenoughbeendonetoadjustpolicyframeworks?Notquite.Inthecaseofmacroprudential policy, more and better can be done with respect to thecalibration and activation of the instruments. In the case of monetary policy,morecanbedonewithrespecttotheexerciseoftheleanoption.Andinthecaseof fiscal policy, there is a need to recognize the hugely flattering effect thatfinancialboomshaveonthefiscalaccounts.

Somuchforpreventionandhowtoaddressthefinancialboom;whataboutthequestion of how to address the bust? If anything, here the questions are evenbigger and more controversial, while progress has been more limited (Borio2012a, 2012b). There is a serious risk, in particular, that the effectiveness ofmonetaryandfiscalpolicyisoverestimatedandofanew,moreinsidious,formoftimeinconsistency.Butthatisanotherstory.

Notes

TheviewsexpressedinthischapteraremyownandnotnecessarilythoseoftheBankforInternationalSettlements.

1.The references are almost exclusively toBISwork, especially recentwork,although the institution's support of macroprudential policy goes back a longway (e.g., Clement 2010). That work contains extensive references to theliterature.

2.ThissectiondrawsinparticularonBorio(2012a).

3.Thequalification"traditional"isimportant.ThedataalsoreveallongerswingsinGDP,whichareclosertothoseforthefinancialcycle.SeeDrehmann,Borio,andTsatsaronis(2012).

4. Although the changes in amplitude over time in the business and financialcyclesaremeaningful,becausethefinancialcyclecombinesdifferentseries,itisnot possible to draw inferences about the relative amplitude of the two cyclesfromthegraphinfigure6.1.SeeDrehmann,Borio,andTsatsaronis(2012)foradiscussionofthetechnicalissuesinvolved.

5.Real-timeor exante refers toanestimate that isbasedonlyon informationavailableatthetimetheestimateismade.

6. Not surprisingly, these trends are consistent with the average length of thefinancialcycle(seeDrehmann,Borio,andTsatsaronis2011).

7.Allthiscastsdoubtontheviewthatcurrentaccountimbalanceswereacauseof the financial crisis. For an in-depth discussion of this issue, see Borio andDisyatat(2011).

8.Koo(2003)seemstohavebeenthefirsttousesuchaterm.Heemploysittodescribe a recession driven by nonfinancial firms' seeking to repay theirexcessivedebtburdens,suchasthoseleftbytheburstingofthebubbleinJapanin the early1990s.Specifically, he argues that theobjectiveof financial firmsshiftsfrommaximizingprofitstominimizingdebt.Thetermisusedheremoregenerallytodenotearecessionassociatedwiththefinancialbustthatfollowsanunsustainable financial boom. But the general characteristics are similar, inparticularthedebtoverhang.Thatsaid,wedrawdifferentconclusionsabouttheappropriate policy responses, especially with respect to prudential and fiscalpolicy(seeBorio2012a).

9. There is also a cross-sectional dimension, which relates to how risk isdistributed in the financial systemat a point in time (see, e.g.,Crockett 2000;Boric,2011).10.EventheFinancialStabilityAssessmentProgramforIceland,releasedinAugust2008,concludedthat"stress testssuggest that thesystemisresilient";seeIMF(2008,8).

11. The existence of a "risk-taking" channel of monetary policy, wherebychanges in interest rates (and other monetary policy tools) influence riskperceptions and risk tolerance, strengthens the case for an active role formonetarypolicy.Itisnot,however,anecessaryconditionforit.SeeBorioandZhu(2011).

References

Avdjiev, S., R.McCauley, and P.McGuire. 2012. "Rapid Credit Growth andInternational Credit: Challenges for Asia." BISWorking Paper 377, Bank forInternationalSettlements,Basel,April.http://www.bis.org/publ/work377.pdf.

BaselCommitteeforBankingSupervision(BCBS).2010.GuidanceforNationalAuthorities Operating the Countercyclical Capital Buffer. Basel: Bank forInternationalSettlements,December.http://www.bis.org/publ/bcbs187.htm.

Borio, C. 2011. "Implementing a Macroprudential Framework: BlendingBoldness and Realism." Capitalism and Society 6 (1): art. 1.http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2208643.

Borio, C. 2012a. The Financial Cycle and Macroeconomics: What Have WeLearnt? BIS Working Paper 395, Bank for International Settlements, Basel,December. http://www.bis.org/publ/ work395.htm. Forthcoming in the JournalofBanking&Finance.

Borio, C. 2012b. "On Time, Stocks and Flows: Understanding the GlobalMacroeconomic Challenges." Lecture at the Munich Seminar series, CESIfo-Group and Siiddeutscbe Zeitung, October 15. http://www.bis.org/speeches/sp121109a.htm.ForthcomingintheNIESRReview.

Borio, C. 2013. "The Global Financial Crisis: Setting Priorities for NewStatistics."BISWorkingPaper408,Bank for InternationalSettlements,Basel,

April. http://www.bis.org/publ/ work408.htm. Forthcoming in the Journal ofBankingRegulation.

Borio, C., and P.Disyatat. 2011. "Global Imbalances and the Financial Crisis:LinkorNoLink?"BISWorkingPaper346,BankforInternationalSettlements,Basel,May.http://www.bis.org/publ/work346.htm.

Borio,C., andP.Disyatat, andM.Juselius.2013. "RethinkingPotentialOutput:Embedding Information about the Financial Cycle." BISWorking Paper 404,Bank for International Settlements, Basel, February. http://www.bis.org/publ/work404.htm.

Borio, C., and M.Drehmann. 2009. "Assessing the Risk of Banking Crises:Revisited." BIS Quarterly Review, March, 29-46. http://www.bis.org/publ/qtrpdf/r_gt0903e.pdf.

Borio,C.,M.Drehmann, andK.Tsatsaronis.2012. "Stress-testingMacroStressTesting:Does ItLiveUp toExpectations?"BISWorkingPaper369,Bank forInternational Settlements, Basel, January. http://www.bis.org/publ/work369.htm.ForthcomingintheJournalofFinancialStability.

Borio, C., and P.Lowe. 2002. "Assessing the Risk of Banking Crises." BISQuarterly Review, December, 43-54. http://www.bis.org/publ/qtrpdf/r_gt0212e.pdf.

Borio, C., R.McCauley, and P.McGuire. 2011. "Global Credit and DomesticCredit Booms." BISQuarterly Review, September, 43-57. http://www.bis.org/publ/qtrpdf/r_gtll09f.pdf.

Borio, C., and H.Zhu. 2011. "Capital Regulation, Risk-taking and MonetaryPolicy:AMissingLinkin theTransmissionMechanism?"JournalofFinancialStability, December. Also available as BIS Working Paper 268, Bank forInternational Settlements, Basel, December 2008. http://www.bis.org/publ/work268.htm.

Caruana, J. 2011. "MonetaryPolicy in aWorldwithMacroprudentialPolicy."Speech delivered at the SAARCFINANCE Governors' Symposium, Kerala,India,June11.http://www.bis.org/speeches/sp110610.htm.

Caruana,J.2012a."DealingwithFinancialSystemicRisk:TheContributionofMacroprudential Policies." Panel remarks at the Central Bank of Turkey/G20ConferenceonFinancialSystemicRisk,Istanbul,September27-28.http://www.bis.org/speeches/sp121002.htm.

Caruana,J.2012b."InternationalMonetaryPolicyInteractions:ChallengesandProspects."SpeechdeliveredattheCEMLA-SEACENconference,"TheRoleofCentralBanksinMacroeconomicandFinancialStability:TheChallengesinanUncertainandVolatileWorld,"PuntadelEste,Uruguay,November16.http://www.bis.org/speeches/spl2lll6.htm?ql=l.

Clement, P. 2010. "The Term `Macroprudential': Origins and Evolution." BISQuarterly Review, March, 59-67. http://www.bis.org/ publ/qtrpdf/r_gt1003h.htm.

CommitteeontheGlobalFinancialSystem(CGFS).2012.OperationalisingtheSelection and Application of Macroprudential Instruments, publication 48.Basel: Bank for International Settlements, December. http://www.bis.org/publ/cgfs48.him.

Crockett,A.2000. "Marrying theMicro - andMacroprudentialDimensionsofFinancial Stability." BIS Speeches, Bank for International Settlements, Basel,September21.http://www.bis.org/review/r000922b.pdf.

Drehmann,M.,C.Borio, andK.Tsatsaronis. 2011. "AnchoringCountercyclicalCapitalBuffers:TheRoleofCreditAggregates."InternationalJournalofCentralBanking 7 (4): 189-239.Also available as BISWorking Paper 355, Bank forInternational Settlements, Basel, November. http://www.bis.org/ publ/work355.htm.

Drehmann,M.,C.Borio,andK.Tsatsaronis.2012."CharacterisingtheFinancialCycle:Don'tLoseSightof theMediumTerm!"BISWorkingPaper380,Bankfor International Settlements, Basel, November. http://www.bis.org/publ/work380.him.

International Monetary Fund (IMF). 2008. "Iceland: Financial StabilityAssessment: Update." Monetary and Capital Markets and EuropeanDepartments, International Monetary Fund, Washington, DC, August 19.

http://www.imf.org/external/pubs/ft/scr/2008/cr08368.pdf.

Koo,R.2003.BalanceSheetRecession.Singapore:JohnWiley&Sons.

Reinhart, C., andK.Rogoff. 2009. This Time Is Different: Eight Centuries ofFinancialFolly.Princeton,NJ:PrincetonUniversityPress.

StanleyFischer

Macroprudential policy or supervision relates to the behavior of the financialsystem as awhole,with a focus on systemic interactions, systemic indicators,andsystemicrisk.TheconsequencesoftheLehmanBrothersbankruptcyarethearchetypeoftheresultofsystemicrisk.Macroprudentialsupervisionandpolicyare directed at monitoring these risks and using available tools to reduce therisksandtheconsequencesoftheiroccurrence.

However,theword"macroprudential"isusedinavarietyofsenses,anditisnotclearthatweallmeanthesamethingwhenweusetheterm.InthischapterIdiscussmainlytheIsraelieconomyandwhatwehavedonethatmightbecalledmacroprudential supervision,ormacroprudentialpolicy implementation.At theendofthechapterIbrieflyraisethequestionofwhetheralltheactionsthatarecalledmacroprudentialareindeedthat.

Israel:TheInstitutionalSetting

MonetaryandmacroprudentialpoliciesinIsraelarecoordinatedviatheBankofIsrael-and the bank's ability to coordinate is entirely the result of institutionalfeatures, in the creation of which thinking about macroprudential issues hadpracticallyno role. I servedasgovernorof theBankof Israel fromMay2005untilJune2013,andhopetooffersomeinsightsfromthisperspective.

Inour institutional setup,banksupervision is located in thecentralbank. Itspurposes and structure are defined in a law (BankingOrdinance 1941) that isseparate from the Bank of Israel law. The supervisor of banks reports to thegovernorofthebank-"reportsto"inthesensethatthegovernorcaninstructthesupervisorastowhattodo,andhastosignoffonsignificantmeasures.Thislineofreporting,inwhichthegovernorcaninstructthesupervisorofbanks-subject,of course, to being in conformity with the law relating to bank supervision-appearstobeunusual.

Monetary policy decisions are made by the Monetary Policy Committee(MPC),whichhassixmembers,threeofwhomarenotemployeesoftheBankofIsrael.Intheeventofatiedvote,thegovernorhasadoublevote.TheBankofIsrael law does not give the MPC the authority to instruct the supervisor ofbanks.Even though the newBank of Israel lawwas passed inApril 2010, asunderstanding about the importance of macroprudential issues wasstrengthening, at that time we did not think of the need to coordinatemacroprudentialpolicywithmonetarypolicy.

Wehave aFinancialStabilityUnit in theResearchDepartment,whichdoesgoodwork.Once a quarter there is a specialmeeting of theMPC devoted tofinancial stability. The Financial Stability Unit, in collaboration with theDepartment of Bank Supervision and the Markets Department (which isresponsible for the bank's operations in both the shekel and foreign exchangemarkets),preparesandpresentstotheMPCacomprehensivereportonfinancialstability.

Apartfromthequarterlyreport,membersoftheMPCmayrequestinformationon recent developments in the banking sector or other parts of the financialsystem from the relevant departments. Further, when the bank supervisor isplanningactionsrelevant to,orhas informationabout, financialstability,heorshe will be invited to brief the MPC on those topics. The members of thecommitteeareentitledtoaskallthequestionstheywantandtoofferadvice,buttheydonotmakeanydecisionsabouttheactionsofthesupervisorofbanks.

InpracticethereisinformalcooperationbetweenthebanksupervisorandtheMPC, and there is a great deal of information provision to the MPC by thesupervisor. But there is no formal coordination between the MPC and thesupervisor:Thechannelofcoordinationgoesthroughthegovernor,whoisbothchairmanoftheMPCandthepersontowhomthesupervisorofbanksreports.

One difficulty in this regard is that the Banking Ordinance 1941, whichdetermines the responsibilities of the supervisor of banks, does not relate tomacroprudentialissues.Thesupervisor'sprimaryresponsibilityistomaintainthestabilityofthebankingsystem.Onoccasionthebanksupervisorhasrejecteda

request to implement measures that would be useful from the viewpoint ofmonetarypolicybutthathejudgedtobeinconsistentwithhismandate.

Giventheinstitutionalsetupinwhichbanksupervisionishousedinthecentralbank and in which the banks are the predominant financial institutions,coordinationbetweenmonetarypolicyandmacroprudentialsupervisionoperatesreasonably well. However, there are other regulators of the financial system,notablythecommissionerofthecapitalmarkets,theinsuranceandsavingsandinsurance supervisor within theMinistry of Finance, and the chairman of theIsraelSecuritiesAuthority(ISA),theequivalentoftheSecuritiesandExchangeCommission in theUnitedStates.The ISA is an independent institution,morecloselyassociatedwiththeTreasurythanwithanyotherinstitution.

The actions of the other two financial regulators are also relevant to thestabilityofthefinancialsystem.Thethreesupervisorsworkwelltogether.Theyhaveacommitteethatmeetsregularly,atwhichtimetheyexchangeinformation.Thefactthateachsupervisorisaprofessionalseemstoenablethemtocooperateinidentifyingsystemicrisks,discussingwhatneedstobedone,andexchanginginformation about what is going on in each of the areas for which they'reresponsible.

Gettingjointactionistypicallyharder,buttheinformationisthereandwould,inalllikelihood,beavailabletothosewhowouldneeditintheeventofacrisis.Thenewlaw(passedin2010)allowstheBankofIsraeltoactaslenderoflastresort to nonbank financial institutions, and we are allowed to demand anyinformationweneedfromanyfinancialinstitution,even,ifnecessary,beforetheinstitution needs liquidity or lender-oflast-resort assistance from the Bank ofIsrael.

In addition, we have tried-and we continue to try-to establish a highlevelfinancial stabilitycommittee (FSC)anchored in law.Thatprocess isnotgoingwell. The first key issue is the relative roles of the finance minister and thebank'sgovernor;and thesecond is thatunderstandablynoneof thesupervisorswants to be subject to any decision or action of the highlevel committee thatwouldrequirethemtodoanythingtheywouldnototherwisewanttodo.

Wemaybeenroutetosolvingtheproblemoftherelativerolesofthefinanceminister and the governor, using a framework parallel to that of the Bank ofIsrael law:The law specifies that so long as a financial institution needs onlyliquidityassistance,theBankofIsraelcanhandletheproblemonitsown;butassoonastheinsolvencyofafinancialinstitutionhastobedealtwith,theTreasuryhastobeinvolved.Thinkingalongthoselines,wehopeitwillbepossiblethatinnormal times, when the central bank is providing no more than liquidityassistance, the governor will chair the FSC, and when we are involved withsolvencyissues,thefinanceministerwillbethechair.

With regard to the second issue, the powers of the FSCwith respect to thethree supervisors, it may be possible to rely on the principle of "comply orexplain."Underthisapproach,theFSCcouldsuggestthataparticularsupervisordosomething.Hewilltheneitherdoitor,ifhecannotordoesnotwanttodoit,hecanexplainwhy.

This sounds simple, but the question arises whether such recommendationswillbemadepublic.Thesupervisorsopposetherecommendationsbeingmadepublic;theyarguethatiftherecommendationsaremadepublic,theFSCwillineffect be using the public to force actions on them that theymay notwant totake. So the second issue, that of the powers of the FSC, is not resolved.Nodoubt,inaperfectlyoperatingsystem,wewouldallsitdownandworkitoutandcomeupwithasolution.Buttheseareimportantofficials,andeachhasavalidinterest inmaintaining his independence-and I'm sure if Iwas an independentregulator,Iwouldalsowanttomaintainmyindependence.

Inpractice,therewouldbenowayofkeepingsuchrecommendationsfromthepublic. There is a public information law. Further, this is the sort ofrecommendationthatwouldbelikelytoleak.SoforthemomentweseemtobestuckontheissueofthepowersofapossibleFSC.

What are we doing on macroprudential issues in the Bank of Israel? Wemonitorthestabilityofthebankingandinsurancesectors.Thosearethesectorstowhichwewillhavetosupplyliquidityifandwhenthetimecomes.WehavedataonthebanksfromwithintheBankofIsrael,andwearereceivingdataon

the insurance companies. We use stress tests to try to detect points ofvulnerability in these institutions or in the financial system. In addition, weanalyze the risks in the portfolios of institutional investors who are movingincreasingly into lending, which means they are becoming a shadow bankingsystem.

We monitor developments in financial markets to try to detect potentialproblemsthatmightemergethere.Wehaveavarietyofindicatorsforassessingfinancialstability,amongthemaradarchartfortheentirefinancialsector.

Withregardtothestresstests,weareusingcontingentclaimsanalysis,aboutwhich members of both the Article IV mission and the Financial SectorAssessment Program (FSAP) team educated us. I have some questions aboutthis,becausethecontingentclaimsanalysisassumesthatthestockmarketgotitrightatsomepoint,andusesstockmarketdatatoidentifyandquantifyrisks-andthere could be a big gap between what the stockmarket is signaling and theobjectivesituation;thatis,irrationalexuberance.

What will we do if we see developments in sectors outside our areas ofsupervisionthatthreatenfinancialstability?Wewouldofcoursefirsttrytoworkwith the relevant supervisors.But if that does not succeed,wewould have tomobilize the supervisors of the supervisors and other senior members of thegovernment to try toget theappropriateactions implemented. In that regard, InotethatalthoughtherewasnoformalsystemofcoordinationamongtheBankofIsrael,theTreasury,andtheOfficeofthePrimeMinisterduringtheperiodofintensecrisisin2009and2010,cooperationwasinpracticecloseandeffective.

TheProblemofHousePrices

Now that I have described the institutional structure for macroprudentialsupervision,letmegiveoneexampleofitsusetodealwithaproblem,whichisourhousingmarket.

Earlier in this conference, Allan Meltzer asked why the extraordinarymeasuresimplementedbytheFedandothercentralbanksseemtohavehadsolittleeffectonrealactivity,orinotherwords,whythetransmissionmechanism

ofmonetarypolicyseemstohavebrokendown.Ibelievethatinmostcasesinwhich it has broken down, it is because the financial system broke down. Inpractically every country where the financial system remained intact, thetransmissionmechanismoperatedwell,withactivityandassetpricesremainingsensitive tomonetary policy. Since almost every country has cut interest ratessignificantlytopreventlarge-scalecapitalinflowsandconsequentappreciationsof thecurrency,as interest rateshavefallen,housingdemandandhousepriceshaverisensharplyincountriesthatdidnothaveafinancialcrisis.Thishappenedin Singapore, in Hong Kong SAR, in Korea, in Australia, in Norway, inGermany,inIsrael,andelsewhere.

HousepricesandrentsinIsraelhavegoneupsignificantlysince2008.SinceJanuary 2008, nominal housing prices have risen 73 percent, and real housingpriceshaverisenbya littleover50percent.Rentshavegoneup35percent innominaltermsandabout19percentinrealterms.Thesearelargeincreases.

HereIwouldliketonoteafact,towhichIwillreturnlater:Israelihouseholdsarenotveryindebted.TheIsraelihouseholddebt-to-incomeratio is lower thanthatofGermany,whichisthelowestamongthemajorcountries,andfarbelowthatof theUnitedStates.Further, loan-to-value(LTV)ratiosonmortgagesarelow,onaverageabout53percent.Following recentchanges in regulations,noonecanobtainamortgagewithanLTVratioofmorethan75percent,andmostmortgageshaveanLTVratioof60percentorless.

There has been a great deal of press and popular interest in Israel in thequestionofwhetherthereisabubbleinhouseprices.Ihavetriedtoeducatethepress,andtelljournaliststhatabubbleisafeaturethathasatechnicaldefinitionandthatonecantestforit.OlivierBlancharddidfundamentalworkonthisissuelongago,asdidothers.

Nonetheless, the question of whether we are in a process that meets thetechnicaldefinitionofabubbleisnotreallyuseful.Rather,theusefulquestioniswhetherpricesaregoinguptoofast.Thefactthattwoyearsinarowpricesrosemore than16percent eachyear is an indication that pricesweregoingup toofast.

Sowehavetriedtodealwiththeproblem.Themacroprudentialmeasureswehaveundertaken are listed in table 7.1; additionalmeasures undertakenby thesupervisorofbanksarelistedintable7.2.InJuly2010werequiredthebanks,whicharethemainprovidersofhousingfinance,tomakeadditionalprovisionsforhousingloanswithhighLTVratios.AbitlaterwerequiredthemtoincreasethecapitalheldagainstfloatinginterestrateloansthatweregrantedwithahighLTVratio.

In May 2011, we imposed a constraint that was unquestionably the mosteffectiveof thevariousmacroprudentialmeasureswehave introduced.At thattime,bankswereprovidingmortgagefinancingthroughamortgagelinkedtotheBankofIsraelinterestrate,whichwasthen3percent,andthemarginonthoseloans, through competition among banks, had declined to 0.6 percent. Sohouseholdscouldobtainfinancingthroughafloatingratemortgage,atanominalinterest rate of 3.6 percent, indexed to the Bank of Israel interest rate. Themacroprudentialmeasureweimposedwastolimittheamountofamortgagethatcouldbe financed through that instrument toone-thirdof the financingof anyloan.Thatstephadasignificantimpactonthedemandformortgages,andonthedemandforhousing.

Table7.1

MeasurestoDealwithHousePriceIncreases

Table7.2

ResidentialMortgageMarket:AdditionalPrecautionaryMeasures

More recently, inNovember2012,we limited theLTVratiosonmortgages.That set of measures contained one element not fully consistent with risk

management:weallowedfirst-homebuyerstohavea75percentLTVratio,andthatwasincludedforpoliticalreasons.Wehadearlier(inOctober2010)comeunder considerable pressure because "wewere not allowing young couples tobuy housing"whenwe had raised the capital ratio on highLTV loans. So, inNovember 2012, we allowed firsttime buyers a slightly higher LTV ratio-75percent-andwereducedtheLTVratioparticularlyforinvestors(definedasthosebuyingasecondapartment).

Finally(sofar),inFebruary2013,wechangedtheriskweightsformortgages.Thiscameat the initiativeof thesupervisor,whowasbecomingworriedaboutthestabilityofthebanksifhousingpricesstartedtogodown.

NowIwanttodiscusstwoissues,thefirstrelatingtothefactthathousepricescontinuetorise.Wehaveimposedmanymeasures,andtheyhavenotyetdonethejob.Evidentlythemeasureswehaveimposedhavenotbeenstrongenoughtoreducethedemandforhousingsufficientlytostoppricesrising.

Theissueofhowstrongthemacroprudentialmeasuresshouldhavebeenisadifficult one. Fundamentally,we have two sources of growth in the economy.One is the construction sector, and theother is exports.Ourmandateprovidesthatso longas inflation isundercontrol,weshouldalso try tosupportgrowthand reduce unemployment, and we have consistently taken that obligationseriously,aswemust.Exportshavebeenunderconstantpressurebecauseoftheeffectofcapitalinflowsontheexchangerate.Exportsamountto37.5percentofGDP,withvalueaddedof26percentofGDP.Exportsarecriticaltogrowth,andwe accordingly are restricted in the use of the interest rate as a means ofreducingthedemandforhousing.

Constructionactivityamountstoabout8percentofthebusinesssectorGDP,andwehave tobecarefulnot tokill thehousingmarket.Wecouldhavebeenmuchtougherinthevariousconstraintsweimposed,butwedidnotwanttotakethe risk of overdoing the restraining effect of our macroprudential measures.Further,althoughhousingpriceshavebeenrisingfast,Israelihouseholdsarenot,byinternationalstandards,highlyindebted.

So, maybe we have done too little in imposing these measures, but at thisstageourknowledgeofhowmacroprudentialmeasuresworkisfarfromprecise.At some point, after further research, our knowledge should become moreaccurate.Inthemeantime,wewillapproachtheuseofmacroprudentialpoliciesintheareaofhousingwithcareandwithgradualism.

CentralBankIndependence

The second issue I want to discuss was raised by Olivier Blanchard in thematerial sent to speakers at the conference. If central banks are in charge ofmonetarypolicy,financialsupervision,andmacroprudentialpolicies,shouldwerethinkcentralbankindependence?

Thereasontoraise this issue is thatwhile therearegoodreasons to insulatetheinterestratedecisionfrompoliticalcontrol,thatmaybelesstruewithregardto macroprudential policies. Certainly we at the Bank of Israel have beensubjected to farmorecriticismoversomeof themacroprudentialmeasureswehave imposed than over our standard monetary policy measures. The naturaltendencyofeconomistswouldbe to say that if thegovernmentwants to favorone group or another's access to housing, it should simply subsidize theirhousingthroughthebudget.Thegovernmentmightwellrespondthatthecentralbankistypicallycriticalofsubsidies,andthatthecentralbankhasthemeanstosolvetheissue.

TheanswertoOlivier'squestionisyes,weshouldreconsidersomeaspectsofcentral bank independence for central banks that are also macroprudentialsupervisors.Asnotedabove,centralbankindependencein interestratesetting,whichisaclassicroleofthecentralbank,atleastoverthepast20years,remainsessential. But we should recognize that changes in the interest rate havedistributional effects.Whenwe raise the interest ratewe affect housing pricesand the exchange rate, among other variables and sectors.We and the publichave gotten used to that, and tend to accept it as a neutral use of monetarypolicy,whichitisnot.

Further,flexible inflationtargetingleavesacentralbankwitha lotofchoice

aboutthetrade-offbetweeninflationandgrowthatanymoment.Thatchoicetoohasdistributionaleffects, includingonunemployment.Muchof thepublicandpolitical pressure exerted on central banks over the years has related to theimpactof theirmonetarypoliciesongrowthandunemployment, aswell asoninflation.Weand thepublichaveover thecourseof timepersuadedourselvesthatwecanleavethattrade-offtothecentralbanks,andthatiswidelyaccepted.

Soifthatiswidelyaccepted,whyshouldweworrysomuchaboutthefactthatmacroprudential policy has clearly defined distributional effects? I believewehavetoworrybecausepoliticiansworryaboutthosedistributionaleffects.Thatmeans they will push harder on us to take their preferences in this area intoaccount-andweneedtofigureouthowbesttoreacttothat.

I'd like to make one other point here, that coordination betweenmacroprudential policies and monetary policy is necessary. There is in ourprofession,Ibelieve,aninvalid,quasi-AdamSmiththeoremthatreappearswithalarming frequency.The invalidAdamSmith theorem is, if each institution isgivenonetargettoachieve,andifeachseekstoreachthattarget,theoutcomeisasocialoptimum.Theassignmentis"onetargetperinstitution."

Theoutcomeofsuchaprocessisunlikelytobethateachinstitutionachievesits target, for therearebound tobe trade-offsamong thegoalsof thedifferentinstitutions. So the question becomes, who will coordinate and decide on thetrade-offs?Thatissueisrarelyexaminedexplicitly,butitisaquestionthatneedstobeanswered.

Idonot accept theview thathaving the same institutionbe responsible for,say, monetary policy and macroprudential policies would produce anunresolvableconflictinthecentralbank.Thetrade-offexists,andsomebodyhastodealwithit.Thecentralbankwouldlikelyhavetheadvantagethat itwouldunderstandthetrade-offanditsimplications,andmakeadecisionthattakesallthisintoaccount.

Further,itcandealwithanysuchtrade-off,astheBankofEnglandhasstartedtodo.TheirMonetaryPolicyCommittee (MPC)continues tooperateas it has

sinceitwassetup,andasMPCsinmanycountriesdo,inamannerindependentof politics. In addition, the Bank of England now has a Financial PolicyCommittee,inwhichgovernmentrepresentativesarepresentasobservers.

Someofmycolleaguessay,well,thecentralbank,orthegovernor,cannotbeindependent inoneroleanddependent inanother. I thinkyoucanbe. Idonotseethepresenceofgovernmentobserversinonecommitteeasimpingingontheindependenceofdecisionsinanothercommittee.

ConcludingComments

Inconclusion,Ireturntotheissueofwhetherallthemeasuresthatwedescribeasmacroprudential are that.Are themeasureswe have used in trying to keephousingprices from rising too fast trulymacroprudential?Certainly, there is agenuinefearthatifhousingpricescontinuetorise,andifatsomelaterdatetheyfall rapidly, we will have problems in the banking system. Thus, one of themotivesforouractionsisundoubtedlymacroprudential.

But I think there is something else going on,whichwe ought to recognize.Monetary policy has effects onmany relative prices and economic sectors. Itmaybethatwhatwearedoingisfindinganadditional instrumenttodealwithsomeaspectof thebusinesscycle,orofmonetarypolicy,and thatwearenowusingtwopolicyinstrumentstotargettwovariables.

Thatisawkward,forittakesusbacktowhatcentralbanksusedtodointhe1960s and the 1970s, intervening on too broad a front in too ambitious anattempttofine-tunetheeconomy.Thatwaywelosethesimplicityoftheuseoftheinterestrateinstrument,andweriskmakingmonetarypolicytoocomplicatedandultimatelylesseffective.

Thereisaquestionofwhatarepuremacroprudentialinstruments.Oneviewisthattheonlytrulymacroprudentialtooliscountercyclicalcapitalrequirements.Idoubt that, because there are other instruments aimed at ensuring financialstability,suchascontroloverLTVratios.

Finally, I have not discussed foreign exchange intervention as

macroprudential. As I read the literature, foreign exchange intervention issometimesdiscussedasmacroprudential,andyoucanmakeacaseforwhythatmightbeso,because,dependingonhowcapitalflowscomein,theycouldcreatesources of financial instability in the financial system itself. But foreignexchangeinterventionasmostcountrieshaveuseditisinfactprimarilyawayoftrying to affect the exchange rate to avoid damaging exports. I do not believethatistrulyamacroprudentialintervention.

Nowthatwebetterunderstandthepotentialconsequencesofadversesystemicinteractions, we are almost certain to increase our use of macroprudentialpoliciesinthefuture.Overtheyearswewillneedtodevelopourunderstandingofwhatmacroprudential policy is,what themost usefulmacroprudential toolsare,andhowtheywork.Wehavemadeconsiderableprogressintheperiodsince2008,butmuchmoreremainstobedonetodeveloptheirfullpotential,andtorecognizethelimitationsofthisnewsetofinstruments.

ChoongsooKim

Koreahasgoodexperienceinusingmacroprudentialpoliciestoaddressfinancialrisksinthehousingandforeignexchangemarkets.ThesetwomarketshavebeenkeysourcesofthesystemicriskinKorea,giventheirhighmarketvolatilityandsusceptibilitytobubbles,andwiththesignificantpotentialtowreakhavoconthebroader economy incaseofmarketdislocation. Indeed, after aprecipitous fallduring the 1997-1998 financial crisis, house prices rose at a rapid pace in thefirst half of the 2000s, aided by a strong expansion of credit to households.Similarly,theforeignexchange(FX)marketexperiencedanextendedperiodoflargeinflowsandexchangerateappreciationpriortotheglobalfinancialcrisis,bothofwhichthenreversedsharplyafterthecrisis.

Itwas considered inappropriate to contain the looming financial risks in thehousingorFXmarketbyusingmonetarypolicyalone,whichaffectsthebroadereconomy and thereforemay be too blunt a tool to address localized financialrisks.Amoretargetedpolicywascalledfor, inadditiontomonetarypolicy, toachieve both price and financial stability. As such, Korea resorted tomacroprudentialpolicieswhileatthesametimeadjustingthestanceofmonetarypolicy-butwithaneyeonthedevelopmentsinthebroadereconomy.

For housing market risks, loan-to-value (LTV) and debt-to-income (DTI)regulationsweredeployedwithaviewtolimitingthevolumeofbankfinancingofhomepurchases.Sincetheirrespectiveintroductionsin2002and2005,thesetools have been adjusted in a broadly countercyclicalmanner.With respect toFX-relatedrisks,Korealearnedvaluablelessonsfromtheglobalfinancialcrisisabout the danger of currency and maturity mismatches. As capital inflowsresumed in 2009 and afterward on the back of the highly accommodativemonetary policy in advanced countries, there was a need to prevent such

mismatches from developing again in the banking sector. To that end,macroprudential measures were imposed, including leverage caps on FXderivatives positions and a macroprudential stability levy on noncore FXliabilitiesofbanks.

This chapter reviewsKorea's experiences in usingmacroprudential policies,evaluates their effects, and discusses remaining challenges in the design andimplementationofmacroprudentialpolicies.

Background

Korea's housing market had gone through a dramatic price adjustment in theaftermathofthe1997-1998financialcrisis,possiblyovercorrectingitspreviouspriceincreases.Itbegantorecoverintheearly2000s,butthepaceofrecoverysoonaccelerated(figure8.1).Housepricesincreasedatanannualaveragerateof8.9percentduringtheperiodof2000-2002.Underlyingsuchsteepincreasesinhouse prices was households' greatly improved access to bank financing forhome purchases, combined with self-fulfilling market expectations of furtherprice increases.'Household loans, includingbankmortgage loans, increased atanannualaveragerateof13.9percentduringthesameperiod.

With the sharp increases in house prices and mortgage loans, there weregrowingconcernsoverpotentialoverheating in thehousingmarketandrelatedfinancial vulnerabilities that could translate into systemic risk.Bank financingforhomepurchaseswasinlargepartofferedintheformofvariable-ratebulletloans(withamaturityofthreeyearsorless)thatarevulnerabletorolloverrisks,particularly if house prices fall.Althoughmonetary policy had been tightenedfrom 2005 to 2007, partly to cool down the overheated housing market, itseffectsremainedlimitedastherapidpaceofhousepriceincreasescontinued.

Capital flows to Korea-in particular, bank flows-have been volatile andprocyclical on the back of high trade and financial openness (figures 8.2 and8.3).Forinstance,abouthalfofthebankborrowingsthathadflowedinoverthetwo-year period immediately before Lehman Brothers' collapse flowed outwithinjustfivemonthsafterthecrisishit.Moreover,largecurrencyandmaturity

mismatches emerged on the balance sheets of banks-particularly foreign bankbranches-inthemid-2000s(figure8.4).

Figure8.1

Household Loans and House Prices. a. Changes in household/homemortgageloans (US $billion).Dark bars represent changes in homemortgage loans (nodataavailablebefore2004).b.Rateof increase (%) inhouseprices (yearoveryear).

Source:BankofKorea.

Figure8.2

CapitalFlowVolatility.

Note: Shown is the 12-monthmoving standard deviation of capital flows as apercentageofGDP(annualized).

Source:BankofKorea.

Figure8.3

BankBorrowingsandtheBusinessCycle.

Note:Shownisthe12-monthmovingaverageofbankborrowings(US$billion).Shadedareaindicatescyclicalupswings.

Source:BankofKorea.

Figure8.4

CurrencyandMaturityMismatches(US$billion).

a. Domestic banks. b. Foreign bank branches. Currency mismatches arecalculated as foreign liabilitiesminus foreign assets.Maturitymismatches arecalculatedasshort-termforeignliabilitiesminusshort-termforeignassets.

Source:BankofKorea.

For theentirebankingsector,currencyandmaturitymismatchespeakedatUS$68 billion and US $85 billion, respectively, just before the global financialcrisis.

Themaindriverof suchmismatcheswas the swollenhedgingdemand frommajor shipbuilders amid strong market expectations of currency appreciation.Shipbuilders and other exporters sold dollars forward to banks. Then bankshedged their overbought position by foreign currency borrowings, mostly atshortmaturity.Theconsequencewasasurgeinbanks'short-termexternaldebtandrolloverrisks.Theglobalfinancialcrisiswasatestinggroundforsuchrisks,andKoreanbanksfailedthetest.

Capital inflows resumedfromthesecondhalfof2009on thebackofampleglobal liquidity and the two-speed recovery between advanced and emergingeconomies (figures 8.5 and 8.6).Although itwas awelcome development foreconomic recovery, policy concerns were raised as to the implications forfinancial stability, for short-term external borrowings by banks began to riseagainandportfolioinflowsalsosurged.

Figure8.5

NetNon-ForeignDirectInvestmentCapitalFlows(US$billion).

Source:BankofKorea.

Figure8.6

Pre-andPost-crisisCapitalFlows(MonthlyAverage,US$billion).

Source:BankofKorea.

MacroprudentialPolicyMeasures

HousingSector-RelatedMeasures

Twomajorpolicyinstrumentswereused,LTVandDTIregulations.TheywereintroducedinSeptember2002andAugust2005,respectively,inthemidstofahousing market boom. They have since been adjusted in a broadlycountercyclical manner, tightened or relaxed as warranted by cyclicaldevelopmentsinhousingmarketsandbanklending(table8.1andfigure8.7).

LTVratioswereadjustedatotalofninetimes(tightenedsixtimesandrelaxedthree times) within a 40 percent to 70 percent range, and they were alsodifferentiated depending on loan structure and across areas, with tighterstandards applied where the real estate market was thought to be plagued byspeculation.DTIratiosweresimilarlyadjustedatotalofeight times(tightenedsix times and relaxed two times)within a40percent to75percent range, andwere also differentiated according to borrower characteristics such as maritalstatus,houseprice,andgeographiclocationoftheproperty,amongothers.

Table8.1

MajorChangesinLTVandDTIRegulations

ForeignExchange-RelatedMeasures

SpecificmeasurestoaddressFXrisksweredesignedreflectingnewthinkingandevidence concerning the underlying causes of the global financial crisis andKorea'spastexperiencesaswell.Twopolicymeasureswereimposedonbanks:(1) leverage caps on banks' FX derivatives positions,which required banks tolimit their FX derivatives position at or below a targeted level (specified as apercentage of a bank's equity capital of the previous month), and (2) amacroprudentialstabilitylevy(MSL)onthenoncoreFXliabilitiesofbanks.2

LeveragecapswerefirstintroducedinOctober2010at250percentforforeignbank branches and 50 percent for domestic banks. Since then they have beentightenedtwice,inJuly2011andJanuary2013,andarepresentlyat150percentand 30 percent respectively for foreign bank branches and domestic banks(figure 8.8). The MSL was first imposed on banks in August 2011 and hasremainedunchanged.Thelevyratevariesfrom2basispointsto20basispoints,

withalowerlevybeingappliedtolonger-maturityliabilities(figure8.9).

Figure8.8

LeverageCapsonFXDerivativesPosition(asa%ofcapital).

Note:TheleveragewasfirstannouncedinJune2010.InOctober2010thecapwas 50% for domestic banks and 250% for foreign banks. Two subsequentadjustmentsweremade,inJuly2011(to40%and200%)andJanuary2013(to30%and150%).

Source:BankofKorea.

Figure8.9

MacroprudentialStabilityLevy:BasisPointsbyMaturity.

Source:BankofKorea.

PolicyEffects

CursoryLook

A cursory look at financial and economic indicators suggests that themacroprudential policies deployed thus far have produced the intended policyeffectsonhouseprices,mortgagelendingbybanks,andcapitalflows,atleastintheshortrun.

Lookingthroughafour-quarterwindowcenteredonthedateofpolicychange,atighteningofLTVorDTIregulationstendstobeassociatedwithastatisticallysignificant decline in the speed atwhich house price and/ormortgage lendingincreases(figure8.10).

Figure8.10

EffectsofHousingSector-RelatedPolicyMeasures.

a.Potential effects ofLTV.b.Potential effects ofDTI.From left to right, thefirstbaristhevolumeofincreaseinmortgageloansbeforetightening(leftaxis);

thesecondbaristhevolumeofincreaseinmortgageloansaftertightening(leftaxis);thethirdbaristherateofincreaseinhousepricesbeforetightening(rightaxis);thefourthbaristherateofincreaseinhousepricesaftertightening(rightaxis).ThedataareinquarterlyfrequencyandspantheperiodfromQ22003toQ32011.

Source:NICECreditInformationServiceDB,KookminBank.

Specifically, the rate of increase in house prices fell on average by 1.7percentage points afterLTV regulationswere tightened and by 0.8 percentagepoints afterDTI regulationswere tightened, comparedwith the correspondingrateofincreasesduringthepretighteningtwo-quarterperiod.Similarlymeasuredeffects on the volume of mortgage loans are also significant: the volume ofincrease inmortgage lendingfellonaverageby43.9percent (6.2 trillionwon)afterLTVregulationsweretightenedandby44.4percent(4.9trillionwon)afterDTIregulationsweretightened.

Macroprudential policies targeted for FX risks also appear to have workedwellbyandlarge,althoughlimiteddataavailability(owingtotheshorthistoryofpolicy implementation) constrains statistical analysis.As regards the effectsof leveragecaps,banks,particularly foreignbankbranches,have reduced theirFX derivatives position and hence short-term borrowings by a significantamount since May 2010, when the introduction of leverage caps was firstannounced (figures 8.11 and 8.12).3 MSL is a cost to banks and, therefore,ceteris paribus, shrinks the arbitrage margin (figure 8.13). A preliminaryestimatesuggeststhatthetotallevycollectedcouldbeaslargeas12percentofnet profits for foreign bank branches while being less than 1 percent fordomesticbanks (figure8.14).This is not surprising insofar asdomesticbanks'funding is predominantly in local currency while foreign bank branches fundmostlyinforeigncurrency.

Figure8.11

TotalFXDerivativesPosition(%ofbankcapital).

Source:BankofKorea.

Figure8.12

FXDerivativesPosition,RelatedtoShipbuilders(US$billion).

Note:Annualdataforperiodspriorto2008.

Source:BankofKorea.

Figure8.13

Incentives for Arbitrage Transaction (Foreign Bank Branches, PercentagePoints).

Note:Incentivesaremeasuredastheinterestratedifferential(3months)minustheswaprate(3months).

Source:BankofKorea.

Figure8.14

LevyasaPercentageofNetProfits(asofend2012).

Note:ProfitsareBankofKoreastaffestimates.

Source:BankofKorea.

Thematuritystructureofexternaldebthasalsoimproved,particularlyin thecaseof foreignbankbranches following the introductionof leverage caps andMSL (figure 8.15). During the two-year period after mid2010, foreign bankbranches'short-termborrowingsdecreasedbyUS$35billion,whiletheirlong-term borrowings increased by US $20 billion. To be specific, foreign bankbranches have resorted more to long-term interoffice borrowings in order toreducetheirMSLandalsotoincreasetheirforwardbuyingcapacityinresponseto a tightening of leverage caps on FX derivatives position.4 Consequently,foreignbankbranchesreducedshort-termborrowingsasashareoftotalexternalborrowingsfrom93percentto58percentduringtheperiodof2010-2012.

FormalAnalysis

Amore formal empirical analysis isundertaken togauge thepolicyeffectsontargeted risksandvariables. It shouldbenotedat theoutset,however, that the

analysis presented below is highly preliminary and subject to many caveatsassociated with the difficulty of unknown counterfactuals and limited dataavailability. Nevertheless, the empirical results appear promising for theeffectivenessofmacroprudentialpoliciesinachievingfinancialstability.

Figure8.15

MaturityCompositionofExternalDebt(%).

a.Domesticbanks.b.Foreignbankbranches.

Note:Blackandgrayvertical linesindicatethedatesof theintroductionof theleveragecapandthestabilitylevy,respectively.

Source:BankofKorea.

EffectsofLTVandDTIRegulations

Atwo-variablepanelvectorautoregression(PVAR)isestimatedfortherateofincreaseinhousepricesandthevolumeofbankmortgageloans,controllingfortheeffectsofLTVandDTIregulationsandotherrelevantpolicies(seeappendix8.1 for model specification and the full estimation results). Monetary policytightening is estimated to have the intended effects,while higher capital gainstaxesalsoappeartohavediscouragedthedemandformortgageloansbutraisedhouseprices, suggestingahigher tax incidenceonhomebuyers thanonhomesellers (which is partly attributable to the fact that the housing market was aseller'smarket rather than a buyer'smarketwhen high expectations of capitalgainsprevailed).

The coefficients of LTV and DTI dummies are negative, as expected, andhighlysignificantinmostcases;someDTIdummiesincludedintheequationformortgageloangrowthareestimatedimprecisely,indicatingthatLTVregulationsmayhavebeenmoreeffectivethanDTIregulations,atleastinKorea.Moreover,theestimatedcoefficientssuggestthattighterLTVregulations(withalowercapon LTV ratios) have stronger cooling effects on house prices and mortgagelending by banks. Last but not least, the actual effects of LTV and DTIregulations might be even stronger than estimated if the possible upwardendogeneity bias in the estimated coefficients (stemming from an endogenousresponse of policies to developments in housingmarkets and bank lending) istakenintoconsideration.

The cumulative effects of LTV and DTI regulations on house prices andmortgagelendingaresimulatedbyusingtheestimatedPVARmodel.Accordingtotheresultsofadynamicsimulation,housepricesandtheoutstandingstockofmortgage loanswouldhavebeen75percent and137percenthigher than theirrespective actual levels by the second quarter of 2012 if there had been noregulations inplace throughout the sampleperiod (table8.2and figure8.16).5Althoughtheestimatedpolicyeffects,especiallythoseonmortgageloans,mayappearsomewhatlargerthanonemightexpect,theyarenotunreasonableinthecontext of the galloping pace of credit expansion observed in many housingboomsofemergingeconomies.6

EffectsofLeverageCapsandMSL

The policy effects of leverage caps and the MSL are estimated by using aBayesianVARmodelforbanks'foreignborrowingsandotherrelatedfinancialvariablessuchasVIX7(seeappendix8.2formodelspecificationsandestimatedimpulseresponses).Thebasicassumptionis that leveragecapsarequantitativerestrictions and thus affect directly banks' foreign borrowings, whereas MSLaffectsbanks'foreignborrowingsmainlythroughchangesinfundingcostornetreturns.'Theestimatedmodelwasthenusedtoproducetwoconditionalforecastsof banks' foreign borrowings, one with macroprudential policies (policyscenario) and the otherwith nomacroprudential policies (no policy scenario).The difference between the two scenarios is taken as the estimate of policyeffects.

Table8.2

SimulatedMortgageLoansandHousePricesbyScenario

Leverage caps are estimated to have contributed to improving the maturitystructureofbanks' foreign liabilities.For foreignbankbranches, leveragecapsareestimatedtohavereducedshort-andlong-termforeignborrowingsbyabout0.6 percent and 0.2 percent of annualGDP, respectively, in cumulative termsover the one-year horizon from their introduction (figure 8.17). For domesticbanks, the effects are estimated tobeonlyone-tenthof those for foreignbankbranches.Suchadifferencebetweenforeignbankbranchesanddomesticbanksisnotsurprisingsinceleveragecapswerebindingfortheformeratthetimeoftheirintroductionbutnotforthelatter.

The estimated effects of MSL on the maturity structure of banks' foreignliabilities are smaller than those of leverage caps but nonethelessmoderate inlightofthefactthatthelevyratewassetatfairlylowlevels.Incumulativetermsover the one-year horizon, MSL reduced short-term borrowings of domesticbanksandforeignbankbranchesalikebyabout0.2percentofannualGDPwithnodiscernibleimpactonlong-termborrowings(figure8.18).

Figure8.16

GrowthRatesofMortgageLoansandHousePrices(Actualvs.Simulated,%).

a.Mortgageloans.b.Houseprices.

Source:BankofKorea.

Figure8.17

ImpactofLeverageCapsonExternalBorrowings(%ofGDP).

Note: Gray lines (dashed lines) refer to conditional forecasts under policyscenario(nopolicyscenario);solidblacklinesindicateactualvalues.

Source:BankofKorea.

ImplicationsforSystemicRisk

Evidence for the effectiveness of macroprudential policies may have positiveimplications for systemic risk. Indeed,LTVandDTI regulations seem tohavehad significant effects in mitigating the credit risk of mortgage loans. For

instance, delinquency rate and value-at-risk tend to fall after LTV or DTIregulations are tightened, and vice versa (figure 8.19). DTI regulations,introduced in late 2005, also appear to have affected the composition ofmortgageloans.Theshareofinstallmentloansintotalmortgageloanswaslessthan40percentatend-2005.Itbegantorisefrom2006andstoodat65percentby2012(figure8.20).Therisingshareofinstallmentloansindicatesthereducedrolloverrisksfacedbyborrowersandthuslessdefaultriskonmortgagelending.

Figure8.18

ImpactofMSLonExternalBorrowings(%ofGDP).

Notes: Gray lines (dashed lines) refer to conditional forecasts under policyscenario (no policy scenario); black lines indicate actual values. MSL,macroprudentialstabilitylevy.

Source:BankofKorea

Butmacroprudential policiesmaynot alwayshelp reduce systemic risk.Forinstance, theshareof long-termmortgage loanswithamaturityof10yearsorlongerhasbeenontherisesinceDTIregulationscameintoeffectinlate2005.Accordingly, the duration of mortgage loans increased from 3.8 years to 5.5years over the period of 2006-2012 (figure 8.21). Although a higher share ofinstallmentloansmayhelpreducedefaultrisk,thelongerdurationofmortgageloansmayhaveincreasedbanks'interestrateandliquidityrisks.

Figure8.19

CreditRiskofMortgageLoans.

a. Delinquency rate (%o). b. Value-at-risk. Value-at-risk is normalized to liebetween0and100.

Source:BankofKorea.

Figure8.20

ShareofInstallmentLoans(%).

Source:BankofKorea.

Figure8.21

WeightedAverageDurationofMortgageLoans,2006-2012(years).

Source:BankofKorea.

ConcludingRemarks

PreliminaryevidenceforKoreaseemstoofferstrongsupportfortheusefulnessand effectiveness of macroprudential policies as a tool to achieve and ensuremacrofinancialstability.LTVandDTIregulationshavehelpedstabilizehousingmarkets and keep credit expansion under proper control. Similarly,macroprudentialpoliciestargetingFXrisks,suchasleveragecapsandthebanklevy,appeartohavecontributedtoimprovingthematuritystructureofexternaldebtowedbybanksandreducingthelikelihoodofsuddenstops.

Several policy issues and lessons are emerging from Korea's experiences.First, country-specific circumstances may matter in important ways for theeffectivenessofmacroprudentialpolicies.Forinstance,thepowerfulroleplayedbyLTV andDTI regulations inKorea seems to have been aided by the large

presence of short-term bullet mortgage loans, which are treated as new loanswhenever rolled over. Second, the policy authorities should bear inmind andexercise vigilance concerning a variety of unintended consequences. LTVregulationscouldbecomeprocyclicalifadjustmentsinLTVcapslagfarbehindhouse-pricemovements.AlthoughDTIregulationshaveinducedbankstooffermore long-maturity installment loans, banks' funding maturity has failed toincrease accordingly resulting in largermaturitymismatch and higher interestrateandliquidityriskaswell.

Third,theissueofpolicycircumventionshouldbetakenseriouslybecauseitcould increase or even create systemic risk, not tomention undermine policyeffectiveness.As LTV andDTI regulationswere focused on banks,mortgagelendingby(lessregulatedandlesswell-capitalized)nonbanksincreasedrapidly.Last but not least,more research is called for to answer howbest to combinemacroprudentialandmonetarypolicies.

Appendix8.1

EstimatingtheEffectsofLTVandDTIRegulations:PanelVAR

Data andSpecificationApanelvector autoregression (PVAR) is estimated forhouse prices and the volume of mortgage lending by banks (both seasonallyadjusted), controlling for the effects ofLTVandDTI regulations.The sampleusedfortheanalysisisabalancedpanelthatcovers43areasovertheperiodofQ22003toQ22012.ThespecificationofthePVARisasfollows:

whereXisavectorofendogenousvariables(housepriceandmortgagelending),Pisavectorofpolicyvariables,andZrepresentsothercontrols,suchasnominalGDPanddummies for speculative areas and theglobal financial crisis.Policyvariables considered include the overnight call interest rate (as a proxy formonetarypolicystance),dummiesforLTVandDTIregulations,anddummiesforcapitalgainstaxpolicy(seetableA8.1).

TableA8.1

DescriptionofVariablesinTableA8.2

AsLTVregulationswerefirstintroducedwithacapsetat60percentandhavebeen in place throughout the entire sample period and across all areas, theestimatedcoefficientsofLTVdummiesincludedintheVAR,LTV4andLTV5,capture the policy effects relative to the baseline case of the LTV ratio beingcapped at 60 percent. For DTI regulations, three dummies-DTI4, DTI5, andDTI6-are used in theVAR as the sample includes the periods and/or areas inwhich no DTI regulations were in place. Therefore, their coefficients reflectpolicyeffectsrelativetothebaselinecaseofnoDTIregulations.LTVdummiesareconstructedbasedon thecap level applied tomortgage loansof3 - to10-yearmaturity, which account for the bulk ofmortgage lending by banks (theLTVcapforloanswithamaturityof10yearsorlongerwassetto70percentbutappears tobenonbinding).DTIdummiesare relatively free from thecoverageissue because DTI regulations are applied uniformly except for the first 10-monthperiod,whenDTIregulationswereappliedonlytolargemortgageloansexceeding600millionwon.

EstimationResults

Themodelisestimatedusingthegeneralizedmethodofmoments(GMM).TheoveridentifyingrestrictionsarenotrejectedforbothequationsiftheHansentest(whichismoreheteroskedasticity-consistent)isused.TheestimatedcoefficientsofLTVandDTIdummiesarestatisticallysignificantandhavetheexpectedsigninmostcases,althoughsomeDTIcoefficientsare impreciselyestimatedin theequation for mortgage lending. Monetary tightening (as captured by the callinterestrate) turnsout tohavebeeneffective incurbingthehousingboomandbanklending.Thetaxpolicyappearstohavereducedmortgageloangrowthbutincreasedhouseprices.

Appendix8.2

EstimatingtheEffectsofLeverageCapsandMSL:BayesianVAR

Thebaselinemodelfortheestimationisafour-variableBayesianVARmodel:

whereY is a vector of foreignborrowingsbydomestic banksor foreignbankbranches and other financial variables (OFVs) that are deemed to have aninfluence on FX risks. For domestic banks, OFVs include the FX derivativesposition ratio (as a percentage of bank capital), VIX (Volatility Index), andborrowingspreadoverLibor.Forforeignbankbranches,OFVsincludethesamevariablesexceptthattheborrowingspreadoverLiborisreplacedbythecoveredinterestparitydeviation(CID).

The sample used for the estimation is in quarterly frequency and spans theperiod from Q1 2003 to Q2 2012. Use of quarterly data is dictated by theavailabilityofdisaggregateddataonforeignborrowingsbydomesticbanksandby foreign bank branches (for a robustness check, however, themodel is alsoestimatedusingmonthlydataonproxyvariables).

TableA8.2

EstimationResults

FigureA8.1

ImpulseResponsestoSupplyandDemandShocks.

Notes: Estimation results are for domestic banks. Solid lines indicate medianvaluesofimpulseresponses,brokenlinesindicate68percentprobabilitybands.

Source:BankofKorea.

Four structural shocks of the model are identified by using both sign andexclusionrestrictionsassuggestedbyeconomictheoryandinstitutionalfeatures.First,aglobalriskperceptionshock(oraninnovationtoVIX)isassumedtobeorthogonaltootherstructuralshocks.Second,asupplyshock(orpushfactor)isassumed to move price (borrowing spread or CID) and quantity (foreignborrowings) in the opposite direction. Third, a demand shock (or pull factor)moves price and quantity in the same direction. Finally, a shock to the FXposition ratio is assumed tobe exogenous andorthogonal to contemporaneous

shockstopriceandquantityvariables,capturingchangesintheFXpositionratiounrelatedtoforeignborrowings.

Notes

The author especially thanks Jun Il Kim, Jong Ku Kang, Hoon Kim, SeungHwanLee,NamjinMa,ChanghoChoi,andYongMinKimattheBankofKoreafortheirhelpfulcontributionsandassistance.

1.Afterthe1997financialcrisis,domesticbanksshiftedtheirbusinessstrategyto household lending from corporate lending.The deregulation of the bankingindustry,theprivatizationoftheHousingandCommercialBankin1996(whosemain business was long-term home mortgage lending), and the reduced loandemandby large corporations induced commercial banks to aggressively enterthemortgageloanmarket.

2.The"globalbankingglut"viewexpoundedbyShin(2012),asopposedtotheglobalsavingsglutview,wasinstrumentalinthedesignandimplementationoftheMSL.

3.Atthetimeofpolicyannouncement,theaverageFXleverageratioofforeignbank branches exceeded 260 percent of equity capital. The same ratio fellsharplyto87percentbyJanuary2013.

4.Forforeignbankbranches,partoftheirlong-termborrowings(withamaturitylongerthanayear)fromtheirheadofficesisrecognizedasbankcapital(knownasCapitalB)andexemptfromtheMSL.Becauseleveragecapsaredefinedasapercentage of bank equity capital, higher Capital B ceteris paribus enablesforeignbankbranches tobuydollars forward ingreater amounts.CapitalBofforeignbankbranchesincreasedto16trillionwonbyend-March2012,upfrom6.1trillionwonatend-March2010.

5.FortheeffectsofLTVregulations,thescenarioof"noregulations"referstothecasewithLTVcapat60percentbeinginplace(seeappendix8.1forgreaterdetails).

6.Anomittedvariablebiasisanotherpossibilityforthelargerestimatedpolicyeffects,whichwarrantsfurtherscrutinyinfuturestudies.

7.VIXimpliesvolatilityinS&P500stockindexoptionprices.

8.Specifically,theborrowingspreadoverLiborisusedasaproxyfordomesticbanks' borrowing costs whereas the covered interest parity deviation (CID) isusedasaproxyforforeignbankbranches'netreturns.Thesetwopricevariablesare chosen in view of their centrality to banks' foreign asset and liabilitymanagement.

Reference

Shin,Hyun-Song.2012. "GlobalBankingGlutandLoanRiskPremium." IMFEconomicReview60:155-192.

SheilaBair

IwashonoredwhentheIMFaskedmetomoderatethe"FinancialRegulation"panel at this year's "Rethinking Macro Policy II" conference. And whilenaturallyIdeliveredoneofthemoreenlighteningandthoughtprovokingpolicydiscussionsoftheconference,Ididfailinmydutiesasmoderatortomakesurethepanelistscoveredalltheexcellentquestionsoursponsorssubmittedtous.Ofcourse,thiswastobeexpected,aspanelistsatthesetypesofeventsalmostneveraddress the topics requested of them (I certainly never do) but rather, likepresidential candidates, answer the questions they want to answer. However,acceptingresponsibilityformymismanagement,Iwillnowstepupandanswerthosequestionsmyself.

1. Does anybody have a clear vision of the desirable financial system of thefuture?

Yes, me. It should be smaller, simpler, less leveraged, and more focused onmeetingthecreditneedsoftherealeconomy.Andweshouldbanthespeculativeuseofcreditdefaultswapsfromthefaceoftheplanet.

2.IstheATMtheonlyusefulfinancialinnovationofthelast30years?

No. If bankers approach the business of banking as a way to provide greatervalueatlesscosttotheircustomers(andthatmightposedifficulties),technologyprovidesavirtualgoldmineforproductinnovations.Forinstance,Iamcurrentlytestingaprepaid,stored-valuecard that letsmedovirtuallyallmybankingonmyiPhone.Ittracksexpenses,tellsmewhenI'veblownmybudget,andletsmetemporarilyblockuseof the cardwhenmydaughter, unbeknownst tome,has

pulled it out ofmywallet to buy the latest jeans fromAeropostale. The card,aptlycalledSimple,wasengineeredbytwotechiesinPortland,Oregon.(Notetomegabanks: Ditch the pinstripes for Dockers and flip-flops. The techies arecomingforyounext.)

3.Does the ideaof a safe, regulated core set of activities and a less safe, lessregulatednoncoresetmakesense?

No.

Theideaofasafe,regulatedcoresetofactivitieswithaccesstothesafetynet(depositinsurance,centralbanklending)andalesssafe,moreregulatednoncoreset of activities that do not under any circumstances have access to the safetynet-thatmakessense.

4. How do the different proposals (Volcker rule, Liikanen's regulations,Vickers'sring-fencing)scoreinthatrespect?

Putthemalltogetherandyouaretwo-thirdsofthewaythere.TheVolckerruleacknowledges theneedfor toughrestrictionsonspeculative trading throughoutthe banking organization, including restrictions on trading of securities andderivatives in the so-called "casino bank." Liikanen andVickers acknowledgetheneedtofirewallinsureddepositsaroundtraditionalcommercialbankingandforcemarket fundingofhigher-riskcasinobankingactivities.Combining themwouldgiveusamuchsaferfinancialsystem.

But none of these proposals fully addresses the problem of excessive risktakingbynonbankfinancialinstitutionslikeAIG.TitleIoftheDoddFrankWallStreet Reform andConsumer ProtectionAct empowers the Financial StabilityOversight Council to bring these kinds of shadow banks under prudentialsupervisionbytheFed.Ofcourse,thatlawwasenactedthreeyearsago,andfornearlytwoyearsnowtheregulatorshavepromisedthattheywillbedesignatingshadow banks for supervisory oversight "very soon." This was repeatedmostrecently byTreasury Secretary JackLewonMay 22, 2013, before the SenateBankingCommittee(butthistimehereallymeantit).Forsomereason,theFedand theTreasuryDepartmentwereable to figureout thatAIGandGECapital

were systemic in ananosecond in2008whenbailoutmoneywas at stake, butwhenitcomestosubjectingthemtomoreregulationnow,well,hey,weneedtobecarefulhere.

5. How much do higher capital ratios actually affect the efficiency andprofitabilityofbanks?

Youdon'thavetobeveryefficienttomakemoneybyusingalotofleveragetojuiceprofits,thendumpthelossesonthegovernmentwhenthingsgobad.Inmyexperience,thebankswiththestrongercapitalratiosaretheonesthatarebettermanaged,doabetterjoboflending,andhavemoresustainableprofitsoverthelongterm,withtheaddedbenefitthattheydon'tputtaxpayersatriskandkeeplendingduringeconomicdownturns.

6.Shouldwegoforveryhighcapitalratios?

Yep. I've argued for a minimum leverage ratio of 8 percent, but I like JohnVickers's10percentevenbetter.

7. Is therevirtue in simplicity-for example, simple leverage rather thancapitalratios-orwillsimplicityonlyincreaseregulatoryarbitrage?

The late PatMoynihan once said that there are some things only a PhD canscrewup.TheBaselCommittee'srulesforriskweightingassetsareexhibitA.

These rules are hopelessly overcomplicated. They were subject to rampantgaming and arbitrageprior to the crisis and still are. (If youdon't believeme,read Senator Levin's report on the London Whale.) A simple leverage ratioshould be the binding constraint, supplementedwith a standardized system ofriskweightingstoforcehighercapitallevelsatbanksthattakeunduerisks.Itislaughabletothinkthattheleverageratioismoresusceptibletoarbitragethanthecurrentsystemofriskweightings,giventhewayriskweightsweregamedpriorto the crisis, for example, by moving assets to the trading book, securitizingloans to get lower capital charges, wrapping high-risk collateralized debtobligations in credit default swap protection to get near-zero risk charges,blindly investing in tripleAsecurities, loadinguponhigh-risksovereigndebt,

repofinancing....NeedIgoon?

8.Canwerealisticallysolvethetoo-big-to-failproblem?

Wehave to solve it. Ifwecan't, thennationalize thesebehemothsandpay thepeople who run them the same wages as everyone else who works for thegovernment.

9.Where do we stand on resolution processes, both at the national level andacrossborders?

Good progress has been made, but not enough. Resolution authority in theUnitedStatescouldbeoperationalizednow,ifnecessary,butitwouldbemessyandunduly expensive for creditors.Weneed thicker cushions of equity at themegabanks,minimumstandardsforbothequityandlongtermdebtissuancesatthe holding company level to facilitate the Federal Deposit InsuranceCorporation's single-point-of-entry strategy, andmost important-regulators whomake clear that they have the guts to put a megabank into receivership. Theindustrysaysitwantstoendtoobigtofail,buttheyaren'tdoingeverythingtheycan to make sure resolution authority works smoothly. For instance, industrygroups like the International Swaps andDerivativesAssociation could greatlyfacilitate international resolutions by revising global standards for swapdocumentation to recognize the government's authority to require continuedperformanceonderivativescontractsinaDoddFrankresolution.

10.Canweeverhopetomeasuresystemicrisk?

Yes. It's all about interconnectedness,whichmegabanks and regulators shouldbe able to measure. Ironically, interconnectedness is encouraged by those%$#@&Baselcapitalrulesforriskweightingassets.LendingtoIBMisviewedas five times riskier than lending to Morgan Stanley. Repos among financialinstitutionsaretreatedasextremelylowrisk,eventhoughexcessiverelianceonrepofundingalmostbroughtoursystemdown.Howdumbisthat?

Weneedtofixthecapitalrules.RegulatorsalsoneedtofocusmoreattentiononthecreditexposurereportsthatarerequiredunderDoddFrank.Thesereports

requiremegabankstoidentifyandquantifyforregulatorshowexposedtheyaretoeachother.Megabankfailurescenariosshouldbefactoredintostresstestingaswell.

(Since thesequestionsare related to financial regulation, Iwillnotopineonmeasuringsystemicriskbuildingasaresultofloosemonetarypolicy.)

11.Arebanksineffectdrivingthereformprocess?

Itsureseemsthatway.

12.Canregulatorseverbeasnimbleastheregulatees?

Yes. Read RogerMartin's Fixing the Game (Harvard Business Review Press,2011).FinancialregulatorsshouldlooktotheNFLforinspiration.

13. Given the cat-and-mouse game between regulators and regulatees, do wehavetolivewithregulatoryuncertainty?

Simple regulations that focus on market discipline and skin-in-the-gamerequirementsarehardertogameandmoreadaptabletochangingconditionsthanrules that try todictatebehavior.For instance, thickcapital cushionswill helpensure that whatever dumbmistakes banksmaymake in the future (and theywill), there will be significant capacity to absorb the resulting losses.Unfortunately,thetrendhasbeentowardcomplex,prescriptiverulesthatsmartbanking lawyers love to exploit. The industry generally likes the prescriptiverulesbecausetheyalwaysfindawayaroundthem,andtheregulatorsdon'tkeepup.

Youcanseethatdynamicplayingoutnow,wherethesecuritizationindustryisseekingtoundermineaDoddFrankrequirementthatsecuritizerstake5centsofeverydollarof lossonmortgages theysecuritize.Theysayriskretention isnolonger required because the Consumer Bureau has promulgated mortgagelending standards. But these rules are pretty permissive (no down paymentrequirementandawhopping43percentdebt-to-incomeratio),andI'msurethattheMortgageBankersAssociation is already trying to figureoutways to skirt

them.

Rules dictating behavior can sometimes be helpful, but forcing marketparticipantstotakethelossesfromtheirrisktakingcanbemuchmoreeffective.Oneapproachtellsthemwhatkindsofloanstheycanmake.Theothersaysthatwhateverkindofloanstheymake,theywilltakelossesifthoseloansdefault.

JeremyC.Stein

Iwill focusmy remarks on the ongoing regulatory challenges associatedwithlarge, systemically important financial institutions,orSIFIs. Inpart, this focusamounts to askingaquestion that seems tobeoneveryone'smind thesedays:Where do we stand with respect to fixing the problem of "too big to fail"(TBTF)?Arewemakingsatisfactoryprogress,oritistimetothinkaboutfurthermeasures?

I should note at the outset that solving the TBTF problem has two distinctaspects.First,andmostobvious,onegoalistogettothepointwhereallmarketparticipantsunderstand-withcertainty-that ifa largeSIFIwere to fail, then thelosseswouldfallonitsshareholdersandcreditors,andtaxpayerswouldhavenoexposure. However, this is only a necessary condition for success, not asufficient one. A second aim is that the failure of a SIFI must not imposesignificantspilloversontherestofthefinancialsystem,intheformofcontagioneffects, fire sales,widespread credit crunches, and the like.Clearly, these twogoals areclosely related. Ifpolicydoesabetter jobofmitigating spillovers, itbecomes more credible to claim that a SIFI will be allowed to fail withoutgovernmentbailout.

So where do we stand? I believe two statements are simultaneously true.We've made considerable progress with respect to SIFIs since the financialcrisis.Andwe'renotyetatapointwhereweshouldbesatisfied.

The areas of progress are familiar tomany.Higher andmore robust capitalrequirements, new liquidity requirements, and stress testing all should help tomaterially reduce theprobabilityofaSIFI finding itselfat thepointof failure.Andif,despitethesemeasures,aSIFIdoesfail,theorderlyliquidationauthority(OLA) in Title II of the Dodd-Frank Wall Street Reform and Consumer

ProtectionActnowoffersamechanismfor recapitalizingandrestructuring theinstitution by imposing losses on shareholders and creditors. In the interest ofbrevity,Iwon'tgointoalotofdetailaboutOLA.Butmyboardcolleague,JayPowell,talkedindepthaboutthistopicinaspeechlastmonth,andIwouldjustregister my broad agreement with his conclusion-namely, that the FederalDeposit Insurance Corporation's (FDIC's) so-called single-point-of-entryapproach to resolution is a promising one (see Powell 2013). The FederalReservecontinuestoworkwiththeFDIConthemanydifficultimplementationchallengesthatremain,butIbelievethisapproachgetsthefirst-ordereconomicsrightandultimatelyhasagoodchancetobeeffective.

PerhapsmoretothepointforTBTF,ifaSIFIdoesfail,Ihavelittledoubtthatprivate investorswill, in fact, bear the losses-even if this leads to an outcomethat is messier andmore costly to society than wewould ideally like. Dodd-FrankisveryclearinsayingthattheFederalReserveandotherregulatorscannotusetheiremergencyauthoritiestobailoutanindividualfailinginstitution.Andasamemberof theboard,Iamcommitted tofollowingboth the letterandthespiritofthelaw.

Still, we are quite a way from having fully solved the policy problemsassociated with SIFIs. For one thing, the market still appears to attach someprobabilitytothegovernmentbailingoutthecreditorsofaSIFI;thiscanbeseenin the ratings uplift granted to large banks based on the ratings agencies'assessmentoftheprobabilityofgovernmentsupport.Whilethisupliftseemstohave shrunk to some degree since the passage of Dodd-Frank, it is stillsignificant.' All else equal, this uplift confers a funding subsidy to the largestfinancialfirms.

Moreover,asInotedearlier,evenifbailoutswerecommonlyunderstoodtobea zero-probability event, the problem of spillovers remains. It is one thing tobelieve that a SIFI will be allowed to fail without government support; it isanother tobelieve that such failurewillnot inflict significantdamageonotherpartsofthefinancialsystem.Inthepresenceofsuchexternalities,financialfirmsmay still have excessive private incentives to remain big, complicated, andinterconnected, because they reap any benefits-for example, in terms of

economiesofscaleandscope-butdon'tbearallthesocialcosts.

Howcanwedobetter?Somehavearguedthat thecurrentpolicypathisnotworkingandthatweneedtotakeafundamentallydifferentapproach.2Suchanalternative approach might include, for example, outright caps on the size ofindividualbanksorareturntoGlass-Steagalltypeactivitylimits.

Myownviewissomewhatdifferent.WhileIagreethatwehavealongwaytogo, Ibelieve that theway toget there isnotbyabandoning thecurrent reformagenda but rather by sticking to its broad contours and ratcheting up itsforcefulness on a number of dimensions. In this spirit, two ideas meritconsideration:(1)anincreaseintheslopeofthecapital-surchargeschedulethatisappliedtolargecomplexfirmsand(2)theimpositionattheholdingcompanylevelofasubstantialseniordebtrequirementtofacilitateresolutionunderTitleII ofDodd-Frank. In parallelwith the approach to capital surcharges, a seniordebt requirement could also potentially bemade a function of an institution'ssystemicfootprint.

To illustratemy argument, let us take as given the central premise of thosewhofavorsizelimits,namely,thatsocietywouldbebetteroffifthedistributionofbankswerenotsoskewedtowardahandfulofverylargeinstitutions.(Tobeclear, I am using the word "size" as shorthand for the broader concept of aninstitution's systemic footprint, which in addition to size might reflectcomplexity,interconnectedness,andglobalspanofoperations.)Inotherwords,let's simplyposit that agoalof regulation shouldbe to leanagainstbank size,andask:Whatare thebest regulatory tools foraccomplishing thatgoal?As inmany other regulatory settings, this question can bemapped into the "prices-versus-quantities"frameworklaidoutbyMartinWeitzmannearly40yearsago.3Here a size cap is a formof quantity regulation,whereas capital requirementsthatincreasewithbanksizecanbethoughtofasakindofpriceregulation,inthesensethatsuchcapitalrequirementsareanalogoustoaprogressivetaxonbanksize.4

Akeychallengewithquantity-basedregulationisthatonehastodecidewheretosetthecap.Doingsorequiresaregulatortotakeastrongstandonthenature

of scale and scope economies in large financial firms.Moreover, even if onereadstheempiricalliteratureasbeingquiteskepticalabouttheexistenceofsucheconomiesbeyondacertainpointinthesizedistribution-apropositionthatitselfis debatable-the most that such large-sample studies can do is to make on-average statements about scale and scope economies.'These studies still leaveopen the possibility of considerable heterogeneity across firms, and that somefirmsareabletoaddconsiderablevalueinagivenlineofbusinessbybeingverybig-even if the average firm in the population is not. And such heterogeneityaloneisenoughtocreatesignificantdrawbackstoquantity-basedregulation.

Considerthefollowingexample.Therearethreebanks,A,B,andC.BothAandBhave$1trillioninassets,whereasCissmaller,withonly$400billioninassets. Bank A actually generates significant economies of scale so that it issociallyoptimalforittoremainatitscurrentsize.BanksBandC,bycontrast,haveverymodesteconomiesofscale,notenoughtooutweighthecoststhattheirsizeandcomplexity imposeonsociety.From theperspectiveofanomniscientsocialplanner,itwouldbebetterifbothBandCwerehalftheircurrentsize.

Nowlet'saskwhathappensifweimposeasizecapof,say,$500billion.ThissizecapdoestherightthingwithrespecttobankB,byshrinkingittoasociallyoptimalsize.ButitmishandlesbothbanksAandC,fordifferentreasons.Inthecase of A, the cap forces it to shrink when it shouldn't, because given thespecificsofitsbusinessmodel,itactuallycreatesasubstantialamountofvaluebybeingbig.AndinthecaseofC,thecapmakestheoppositemistake.Itwouldactually be beneficial to put pressure on C to shrink at themargin-that is, tomove it in thedirectionofbeinga$200billionbank insteadofa$400billionone-butsinceitliesbelowthecap,itiscompletelyuntouchedbytheregulation.

Supposeinsteadweattacktheproblembyimposingcapitalrequirementsthatareanincreasingfunctionofbanksize.Thisprice-basedapproachcreatessomeincentive for all three banks to shrink, but lets them balance this incentiveagainst the scale benefits they realize by staying big. In this case we wouldexpectA,withitssignificantscaleeconomies,toabsorbthetaxhitandchoosetoremain large, while B and C, with more modest scale economies, would beexpectedtoshrinkmoreradically.Inotherwords,price-basedregulationismore

flexibleinthatitleavesthesizedecisiontobankmanagers,whocanthenbasetheir decisions on their own understanding of the synergies-or lack thereof-intheirrespectivebusinesses.

This logic can be thought of as supporting the approach taken by theBaselCommittee on Banking Supervision in its rule imposing a common equitysurchargeondesignatedglobalsystemicallyimportantbanks.Theexactamountof the surchargewill range from 1 percent to 2.5 percent andwill depend onfactors that include a bank's size, complexity, and interconnectedness, asmeasuredbyavarietyofindicatorvariables.6Theseprogressivesurchargesareeffectively a type of price-based regulation and therefore should have theadvantagesIjustnoted.

However,aproponentofsizecapsmightreasonablyreply,"Fine,buthowdoIknow that these surcharges are actually enough to change behavior-that is, toexert ameaningful influence on the size distribution of the banking system?"After all, the analogy between a capital requirement and a tax is somewhatimperfectbecausewedon'tknowexactlytheimplicittaxrateassociatedwithagiven level of capital. Some view capital requirements as quite burdensome,whichwouldmeanthatevena2percentsurchargeamountstoasignificanttaxand,hence,astrongincentiveforabanktoshrink,whileothershavearguedthatcapital requirements impose only modest costs, which would imply littleincentivetoshrink.7

Thisuncertaintyabouttheultimateeffectofagivencapital-surchargeregimeonthesizedistributionofbankscouldpotentiallytipthebalancebackinfavorofquantity-based regulation, like size caps.And indeed, ifwewere facedwith astatic, once-and-for-all decision, I don't think economic reasoning alone couldgive us a definitive answer as to whether caps should be preferred to capitalsurcharges.ThisambiguityisinsomesensethecentralmessageofWeitzman'soriginalanalysis.

One way to resolve this tension is to refrain from putting ourselves in thepositionofhavingtomakeaonce-and-for-alldecisioninasettingofsubstantialuncertainty.Rather,itmightbepreferabletotrytolearnfromtheincomingdata

andadjustover time,particularlysince therecentchanges tocapital regulationalreadyonthebooksmayrepresentaninformativeexperiment.Inmyview,thisobservationabout thepotential for learning tips thebalance in favorofcapitalsurcharges.Forexample, thecapital-surcharge scheduleproposedby theBaselCommittee forglobally importantsystemicbanksmaybea reasonablestartingpoint;however,ifaftersometimeithasnotdeliveredmuchofachangeinthesizeandcomplexityofthelargestofbanks,onemightconcludethattheimplicittaxwastoosmallandshouldberatchetedup.8Inprinciple,thisturningup-the-dialsapproachfeelstomeliketherightwaytogo:Itretainstheflexibilitythatmakesprice-basedregulationattractivewhilemitigatingtheriskthattheimplicittax rate will be set too low. Of course, I recognize that its gradualist naturepresents practical challenges, not least of which is sustaining a level ofregulatorycommitmentandresolvesufficienttokeepthedialsturningsolongasthisistherightthingtodo.

Beforewrappingup,letmebrieflymentionanotherpieceofthepuzzlethatIthinkissometimesoverlookedbutthatstrikesmeashavingthepotentialtoplayan important complementary role in efforts to address the TBTF problem,namely,corporategovernance.Supposewedoeverything rightwith respect tocapitalregulationandsetupasystemofcapitalsurchargesthatimposesastrongincentive to shrink on those institutions that don't create large synergies.Howwould the adjustment process actually play out? The first step would be forshareholders,seeinganinadequatereturnoncapital,toselltheirshares,drivingthe bank's stock price down. And the second stepwould be formanagement,seekingtorestoreshareholdervalue,torespondbyselectivelysheddingassets.

Butasdecadesof research incorporatefinancehave taughtus,weshouldn'ttake the second step for granted. Numerous studies across a wide range ofindustries have documented how difficult it is for managers to voluntarilydownsizetheirfirms,evenwhenthestockmarketissendingaclearsignalthatdownsizingwouldbe in the interestsofoutsideshareholders.Often,changeofthissortrequirestheapplicationofsomeexternalforce,beitfromthemarketforcorporatecontrol,anactivistinvestor,orastrongandindependentboard.'Aswemove forward,we should keep these governancemechanisms inmind and dowhatwecantoensurethattheysupportthebroaderregulatorystrategy.

Notes

This chapter was originally delivered on April 17, 2013, as remarks at theInternational Monetary Fund's "Rethinking Macro Policy II" conference. Thethoughts that follow are my own. They are not necessarily shared by mycolleaguesontheFederalReserveBoard.Iamgratefultomembersoftheboardstaff,MichaelGibson,MichaelHsu,NellieLiang,andMarkVanDerWeide,fortheiradvice.

1. For example, in June 2012, Moody's described its ratings process for JPMorganChase as follows: "JPMorgan's ratings benefit from three notches ofuplift from the standalone credit assessment at the bank level, and from twonotchesofupliftattheholdingcompany,reflectingMoody'sassumptionsaboutaveryhigh likelihoodof support from theUSgovernment for bondholdersorothercreditorsintheeventsuchsupportwasrequiredtopreventadefault....Thenegative outlook on the parent holding company reflects Moody's view thatgovernment support forUS bank holding company creditors is becoming lesscertainandlesspredictable,giventheevolvingattitudeofUSauthoritiestotheresolution of large financial institutions, whereas support for creditors ofoperating entities remains sufficiently likely and predictable to warrant stableoutlooks"(Moody'sInvestorsService2012,pp.11-12ofthepdfversion).AlsoseesimilarstatementsbyMoody's in thesamedocument forBankofAmerica(pp.4-5)andCitibank(p.7).

2. See Fisher (2013), who said: "We recommend that the largest financialholdingcompaniesberestructuredsothateveryoneoftheircorporateentitiesissubject toaspeedybankruptcyprocess,and in thecaseof thebankingentitiesthemselves, that they be of a size that is `too small to save.' Addressinginstitutional size is vital to maintaining a credible threat of failure, therebyprovidingaconvincingcasethatpolicyhastrulychanged"(paragraph40).

3. SeeWeitzman (1974). Haldane (2010) also usesWeitzman's framework totalk about price-versus-quantity regulation in the TBTF context. It should benotedthattherearevarioushybridapproachesthatareneitherpurequantitynorpure price regulation. For example, Tarullo's (2012) discussion of limits onuninsured liabilities is not a rigid size cap, since it does not constrain aninstitution'sabsolutesize,totheextentthatitisabletoadjustitsfundingmix.

4. To be clear, this taxation aspect of capital requirements is not their onlyappeal, or even their primary one. Even if it were almost costless to imposehighercapitalrequirementsonbiggerbanks-sothatdoingsoprovidedessentiallyno disincentive to bank size-it might still be a good idea to do so for purelyprudential reasons. In other words, capital requirements serve as both aprudentialbufferandatax,andcanbeausefulregulatorytoolforbothreasons.

5.SeeHughes andMester (2011) for a recent contribution to the literatureonscaleeconomiesinbanking.

6.SeeBCBS(2011)foradescriptionofthemethodology.

7.Fordifferentestimatesofthecostsofcapitalrequirementstobanks,seeBakerandWurgler(2013),Admatiandothers(2011),andHanson,Kashyap,andStein(2011).

8.Again,itshouldbeemphasizedthattheunderlyingproblemisnotsimplyaninstitution'ssizebutratheritssystemicfootprint-which,inadditiontosheersize,isrelatedtoitscomplexity,interconnectedness,andglobalspanofoperations.

9.Jensen(1993)isaclassictreatmentoftheissues.

References

Admati, Anat R., Peter M.DeMarzo, Martin F.Hellwig, and Paul Pfleiderer.2011. "Fallacies, Irrelevant Facts, and Myths in the Discussion of CapitalRegulation: Why Bank Equity Is Not Expensive." Working paper, GraduateSchool of Business, Stanford University, Stanford, CA, and the Max PlanckInstitute for Research on Collective Goods, Bonn. http://gsbapps.stanford.edu/researchpapers/library/RP2065R1&86.pdf.

Baker, Malcolm, and Jeffrey Wurgler. 2013. "Would Stricter CapitalRequirementsRaisetheCostofCapital?BankCapitalRegulationandtheLowRiskAnomaly."Workingpaper,HarvardBusinessSchool,Cambridge,MA,andNYU Stern School of Business, New York. http://people.stern.nyu.edu/jwurgler/papers/Bank%20Capita1%20Regulation.pdf.

BaselCommitteeonBankingSupervision(BCBS).2011."GlobalSystemically

ImportantBanks:AssessmentMethodologyandtheAdditionalLossAbsorbencyRequirement."BankforInternationalSettlements,Basel,November.http://www.bis.org/publ/bcbs207.pdf.

Fisher,RichardW. 2013. "Ending "TooBig to Fail."' Speech delivered at theConservative Political Action Conference, National Harbor, MD, March 16.http://www.dallasfed.org/news/speeches/fisher/2013/fs130316.cfm.

Haldane,AndrewG.2010."The$100BillionQuestion."SpeechdeliveredattheInstitute of Regulation & Risk, Hong Kong, March 30. http://www.bis.org/review/r100406d.pdf.

Hanson, Samuel G., Anil K.Kashyap, and Jeremy C.Stein. 2011. "AMacroprudential Approach to Financial Regulation." Journal of EconomicPerspectives 25 (1): 3-28. http://www.people.hbs.edu/shanson/hanson_kashyap_steinjEP.pdf.

Hughes, JosephP., andLoretta J.Mester.2011. "WhoSaidLargeBanksDon'tExperience Scale Economies? Evidence from a Risk-Return-Driven CostFunction." Working Paper 11-27, Federal Reserve Bank of Philadelphia,Philadelphia, PA, July. www.phil.frb.org/research-and-data/publications/working-papers/2011/wp11-27.pdf.

Jensen, Michael C. 1993. "The Modern Industrial Revolution, Exit, and theFailureofInternalControlSystems."JournalofFinance48(3):831-880.http://onlinelibrary.wiley.com/doi/10.1111/j.1540-6261.1993.tb04022.x/full.

Moody's Investors Service. 2012. "Moody's Downgrades Firms with GlobalCapital Markets Operations." Press release, Moody's Investors Service, NewYork, June 21. www.moodys.com/research/ Moodys-downgrades-firms-with-global-capital-markets-operations--PR_248989?WT.mc_id=BankRatings2012.

Powell, JeromeH. 2013. "Ending `Too Big to Fail."' Speech delivered at theInstitute of International Bankers 2013 Washington Conference, Washington,DC, March 4. http://www.federalreserve.gov/ newsevents/speech/powe1120130304a.htm.

Tarullo, Daniel K. 2012. "Industry Structure and Systemic Risk Regulation."

Speech delivered at the Brookings Institution Conference, "Structuring theFinancial Industry to Enhance Economic Growth and Stability." BrookingsInstitution, Washington, DC, December 4. http://www.federalreserve.gov/newsevents/speech/tarullo20121204a.htm.

Weitzman,MartinL.1974."Pricesvs.Quantities."ReviewofEconomicStudies41(October):477-491.vs_quantities.pdf.

JeanTirole

Aftermathsofbanking,sovereign,andothercrisesoften lookalike:afteryearsofneglectandquasi-laissez-faire leadingtoacrisis,policymakersandscholarswork assiduouslyonnew schemes thatwill prevent thenext crisis.While thisconstitutes a useful reaction, the process also reflects political immediacy asmuchasalong-termperspective.Thetitleoftheconference,"RethinkingMacroPolicyII:FirstStepsandEarlyLessons,"modestlybutrightlyremindsusofthelimitsofourknowledgeintheseareas.

Thischapterbrieflydiscussesthreekindsofreformsinthemakinginfinancialregulation:structuralreforms(amongwhichIwouldincludegovernanceandtheinteraction with shadow banking), solvency and liquidity regulation, andinstitutional/supervisoryreforms.

Before I begin, twopreliminary remarks are inorder.First,weneed to stepbackamomentandthinkaboutwhythebankingsectorisregulatedinthefirstplace, regardless of its degree of competition. There are multiple reasons forregulation. The most universal one is to protect depositors, or the depositinsurancefundandthetaxpayer,ifdepositsareexplicitlyorimplicitlycovered.The state represents the interestsofdepositors,whodonothave theexpertise,the information,or the incentive tomonitor thebalance sheet andoff-balance-sheet activities of their banks. This "representation hypothesis" explains whyeven tiny banks, without market power and systemically unimportant, aresupervised, and why pension funds and insurance companies are subject tosimilarprudentialrequirements.'Italsoexplainswhyinvestmentbanks,withnoretailinvestors,havesofarbeenleftunregulatedorlightlyregulated.

The representation hypothesis, however, does not account for the recentbailoutoftheinvestmentbanksandofAIG'sholdingintheUnitedStatesorthat

of Long-TermCapitalManagement (LTCM) in 1998.Nor does it explain thecurrent emphasis on regulating systemically important financial institutions(SIFIs).Systemicriskisofteninvokedasanargumentforpayingcloseattentiontofinancialinstitutions,andsotherepresentationandsystemicriskrationalesarerelated.Whenlargeinstitutionsfail,unpaidliabilities(counterpartyrisk)andfiresalescreateexternalitiesontheprudentiallyregulatedsectorandmayputatrisksomeretailbanksorinsurancecompanies.2Theseexternalitiesare,tobecertain,highlyendogenous.Wereretailbanks,forinstance,tohavelimitedcounterpartyexposuretoinstitutionsoutsidetheprudentiallyregulatedsector,theexpostcasefor rescuing institutions in the latter sectorwouldbemuchweaker. I return tothispointlater.

Thesecondpreliminaryremarkisthatweneedtomakeprogressonageneralequilibrium view, and not only for the standard macroprudential reasons.Makingbalancesheetssaferisalaudablegoalinlightoftherecentexperience,but itcannotbe theonlyconsideration.Otherwise,prudential regulationwouldbe a rather simple matter: it would suffice to impose very high capitalrequirements, demand CoCos-contingent convertible bonds3-in large amounts,orrequirebankstoinvestonlyinGermanbunds.Thereisademandforthesafe,liquidinstrumentscreatedbythebanks'liabilityside,asillustratedinthetheoryof aggregate liquidity, but so far the latter has been poorly connected toprudential regulationanalysis.There isalsoanincidenceofbankingregulationon the volume of credit to the economy, as well as implications for theincentivesofmanagersandshareholders.

StructuralReforms

The rationale for separating retail from investmentbanking is to insulatebasicbankingservices-and thereby the trilogyofdepositors, theguarantee fund,andtaxpayers-from investment banking risks. It is a simple refusal to engage incross-subsidies and thereby promote value-decreasing marginal investmentbankingservices.

In a sense, structural reform, coming on top of the Basel III prudentialregulation, is an admission that society will always lose in the prudential

regulationcat-and-mousegame;prudentialregulationoughttobecapableatthevery least of duplicating structural reforms (which often are a special case ofprudentialregulationwithoutseparationbutwithinfiniteweightsonsomeassetsor activities) and should in theory do better. Thus, either regulators are nottrustedtochooseandenforcelessextremeriskweights,orthereisafeelingthatoutrightprohibitionhasverydifferenteffectsfromhighriskweights.Underthelatter hypothesis, a specificity of banking is that themousemoves really fast;indeed, there are few industries in which the balance sheet can be changedsubstantiallywithinamatterofdays.Thebankingindustry,at least in itsmost"innovative"segments,ishighlycomplex,andasymmetricinformationwiththeregulator is paramount.There is latency in conforming to capital and liquidityrequirements,which,ifveryriskyinstrumentsareavailable,maygivescopeforgamblingforresurrectionwhenthebankisunderwater.

The various proposals for structural separation have been described andcommentedoncarefullyelsewhere.4HereIwillmakejustafewremarksontheeconomicsofsuchrulestopointoutwhereourknowledgeoughttobeimprovedon.

For the sake of illustration, let's consider the Vickers rule.' In essence, itcreates a ring-fenced subsidiary (the retail bank) with a limited scope ofactivities:Itcanlendonlytohouseholdsandnonfinancialfirmsandtradehigh-quality securities. It can hedge the risk on corresponding exposures-theIndependent Commission on Banking (ICB)6 report calls this the "Treasuryfunction."All other activities are not allowedwithin the ring-fenced bank butcanbeperformedbytherestofthebank(theinvestmentbank).Thering-fencedbankhasoperational independenceandisprohibitedfromprovidingsupport totheinvestmentbank.

As the ICB report emphasizes, ring-fencing is no substitute for capitaladequacyandliquidityrequirements.Actually,thereportcallsforhighercapitaladequacyrequirementsforring-fencedbanks thannowprevail,whichamountstosayingthateventhering-fencedbankmaynotbethatsafe.

TheVickers rule has a number of desirable properties, one ofwhich is the

facilitationofresolution.Asinthecaseofprudentialregulation,thisargumentissomekindofadmissionthattherequirementofalivingwill(i.e.,theprovisionbyabankofdetailedinformationonhowauthoritiescandismantlethebankandproceed through an orderly resolution) will not function properly or at leastneedstobereinforcedbystructuralseparation.

Although the aims of structural separation are well grounded, there arenonethelessseriousconcernsthattherulemaynotachieveitspurposes.Hereareafewpotentialreasons,onwhicheconomistsshouldprobablydomoreresearch:

•Theretailbankcanbuildlargemacrorisksonitsretailbook,suchasrealestaterisk (thinkof Irish,Spanish,andUSbanks),or throughrathersimple financialproductsofferedtoretailcustomers,suchasguaranteesofminimumreturnsoninvestmentportfolios(asgrantedbysomeEuropeanbanks).Indeed,anumberofrecent failures had to do with institutions then abiding by the separationrequirements,primarily-retailbanksandlargeinvestmentbanks.

• The ICB report recognizes that such risks exist and, accordingly, allows theretail bank to hedge. Hedging is a well-known double-edged sword. If notcarefully monitored, it can enable institutions to (voluntarily or involuntarily)increaserisk.HereIrecallthatJPMorgan'strader"LondonWhale"wasactuallyusingveryriskycreditdefaultswapsaspartofthehedgingfunction.

•Itmustnotbethecasethatauthoritiesfeelcompelledtorescuetheinvestmentbank.Theremightbe tworeasonsforsucha rescue.One is reputationrisk fortheretailbank,althoughthisargumentcutsbothways(i.e.,theinvestmentbankmight feel compelled to rescue the retail bank). Another ismore standard: In2008 the US government bailed out all investment banks (except LehmanBrothers),aswellasAIG,becauseitwasconcernedaboutsystemicrisk.7Directandindirectexposuresofretailregulatedentitiestofailingentities,aswellasthepossibilityoffiresales,probablycontributedtothisveryunfortunateoutcomes

RelationshiptoShadowBanking

Thelevelplayingfieldintheshadowbankingsector(theunregulatedsectorthatperformsmaturitytransformationand,becauseoftheabsenceofretailliabilities,

hasnoaccess topublicsectorenhancementssuchascentralbankliquidityanddepositinsurance)hasbeendestroyedintwooppositeways.First,inthecaseofintegrated banks, retail deposit activities and public sector enhancementsmayundulycross-subsidize investmentbankingactivities; this isakeyrationaleforintroducing structural separation.Second,non-integrated investmentbanksandother shadow banking players can avail themselves of massive access totaxpayermoney,aswesawin2008.9

Asoftoday,theexacttighteningofrulesforshadowbankingisstilluncertain(Adrian andAshcraft 2012). Particularly prominent in the last few years havebeen the proposals for regulating SIFIs.While I agree with the premises andpurpose of this reform, I have reservations. First, in an environmentcharacterized by complex products and balance sheets and fragmentedinformation,itisratherhardtoidentifysystemicallyimportantactors.1°WouldAIG have been deemed systemically important? LTCM? Second, there arecurrently too few regulators to oversee retail institutions; carefully overseeinginvestment banks, hedge funds and other potential SIFIs would require asubstantialexpansioninregulatoryresources.Besides,activitieskeepmigrating;tomorrownewentities (suchasenergycompanies,whichalreadyareactive infinancial markets) might expand their financial involvement in response to atight regulation of investment banks and hedge funds. This migration issuecompounds the difficulty of achieving widespread financial regulation withscarceresources.

While the jury on this issue is clearly out,myown, nonmainstreamview isthat rather than attempting ubiquitous regulation, it might be preferable toinsulate prudentially regulated entities (retail banks, insurance companies,pension funds) fromcounterparty riskwith unregulated financial entities.Thatwould contribute to bringing to a halt the soft budget constraint enjoyed by anumber of unregulated entities, which secure cheap refinancing thanks to theexpectationthattheywillberescuedbypublicmoneyifpushcomestoshove.Afaster migration toward the use of centralized exchanges (which need to beprudentially regulated) and further disincentives for prudentially regulatedentities to use over-the-counter (OTC) markets would be desirable in thisrespect.

TheVickersruleisasteptowardthisinsulationstrategy:theretailbankwillhavelimitedexposuretoitsowninvestmentbank.Theextensionoftheruletothe insulation from the impact of failures of non-affiliated entities remains aquestionmark,asexposuresmayreside in theTreasury function throughOTCtransactions; to be sure, the rule's proponents seem sympathetic to the use ofcentralcounterpartyclearinghouses(CCPs)intheCCP/OTCdebate.'

AssetIncomeRuns

Structural separation is an instance of asset income run; that is, thecompartmentalization of the balance sheet through the earmarking of specificassetstospecificlenders.Whilestructuralseparationisdemandedbyauthoritiesand ismeant toprotect retail depositors and taxpayers against certain typesofrisks on the asset side, asset income runs are usually carried out by privatelenders. To this extent, earmarking has always existed: repos (whose useincreased tremendouslywhen legal uncertainty concerning the effectivenessoftheearmarkingguaranteewaslifted)andcoveredbondsarecasesinpoint.Andin rough times, asset income runs in various guises (shortening of maturitystructure,highercollateraldemands)increase.Thisisparticularlythecasetoday.

Earmarking specific assets to specific liabilities has two benefits (and, topursue the analogy, both benefits are mentioned by structural separationproponentstomaketheircase).First,itfacilitatesresolution.Relatedly,lendersneed toascertainonly thevalueof thespecificasset that ismatchedwith theirliabilityandnotthequalityoftheentirebalancesheet,apossiblydauntingtaskfor the lender; by contrast, asserting the value of a single asset (say, of agovernmentsecurity)maybeano-brainer.Alternatively,thelendersmaybringtheir specialized knowledge to assess the value of the collateral; they therebycertifyapieceofthebalancesheet.

Earmarking also has costs. First, it is known that theremaybe less adverseselectionorlessmoralhazardonabundlethanonindividualassets.12Second,and currently verymuch to the point, it createsmore scope for a rat race, inwhicheachlendertriestoobtainpriorityoverotherlenders."Indeed,oneofthecurrentconcerns is that theDeposit InsuranceFundwouldfaceanemptyshell

whentryingtocollect,afterallgoodassetshavebeencollateralizedwithprivatelenders; put differently, such asset income runs make bail-ins infeasible. Onewouldthereforeexpectinthenearfuturemoreprudentialmovestowardlimitingassetincomeruns;acaseinpointistherecentAustraliandecisiontolimitto8percentthefractionofassetsthatcanbeusedtocreatecoveredbonds.

PrudentialRegulationReforms

This is no place to review the very extensive discussions and alterationsconcerning various aspects of regulatory reform: capital adequacy,countercyclicalbuffers, treatmentofderivatives,andsoon.Similarly, Icannotcover the many questions remaining today, including the future of internalmodels or the homogenization and coordination of resolution processes. Iwillcontentmyselfwithafewlightremarksconcerninganinnovationofthecurrentreforms: the introductionof liquidity requirements into theBaselprocess,withtwo ratios in themaking-the liquidity coverage ratio (LCR) and the net stablefunding ratio (NSFR).Reflections on the former aremore advanced; theLCRwillbecomeoperativein2019(withalightversionfor2015).

TheLCRwillbebasedonstresstests:banksmustclaimaccesstoasufficientbufferofhigh-quality liquidassets tobeable tofaceoutflowsduringamonth.TheNSFRwillrelatelonger-termfundingtolonger-termliquidity.

The academic contributions on liquidity have helped clarify the need for aliquidity requirement on top of a solvency one. They have emphasized threemain rationales: agency, microeconomic externalities, and macroprudentialconcerns.Theagencyrationalehasastandardprudentialflavorandrelatestotheneedforconstrainingaspecificformofrisktaking:engaginginanexcessivelyunbalancedmaturity transformation by not hedging against short-term shocks,market collapse, or increases in interest rates. The externalities rationale is toprevent propagation to other financial institutions through counterparty risk orfiresales.Thethirdrationalealsotakesuponthefirstoneandemphasizes thestrategic complementarities in balance sheet mismatches. Authorities (firstmonetary,thenfiscal)aremoretemptedtobailoutthefinancialsectorifmoreofitsmembers have engaged in dangerous levels ofmaturity transformation; the

classic example of a collective bailout is themassive transfer of wealth fromsaverstoborrowersduringtimesoflowinterestrates.14

Where, inmyview,academic research fallsbehind is inprovidingguidanceon technical but important issues in liquidity regulation. Should liquid assetsreceiveriskweights in liquidityratiosasgeneralassetsdoincapital ratios?(Itwould seem so.) Should central-bank-eligible assets, which are liquid bydefinition,becountedas liquidassets,given that thechoiceofwhat counts asliquidity in the computation of the liquidity ratio and central banks' choice ofacceptable collateral are two (imperfectly substitutable) public sectorinterventions in the provision of aggregate liquidity? Could the focus on acertainhorizongiverisetodetrimentalcliffeffectsandgamingstrategies(e.g.,automatic rollover/substitution of liabilities with maturity just above thehorizon)?Towhatextentshould thebufferbeusable?Howdowe incorporategeneral equilibrium considerations and country specificities? For example,countries differ in their stock of government debt and in the amount of retaildeposits, which will be deemed largely more stable than wholesale depositsunder theLCRregulation.Also, the implicationsofbankdistressdifferacrosscountriesaccordingtotheindebtednessofthecountryitselfortheautonomyofthe monetary policy. Should asset quality be assessed through ratings, as iscurrentlyproposed,despitetheDodd-FrankActhostilitytoratings?Thislist isfarfromexhaustive.

This discussion points to a blatant need for academic research in the area.Overthepast20yearseconomistshavemademuchprogressinconceptualizingliquidity,buttheyhavenotyetreachedthepointatwhichtheirtheoriescanbeoperationalized for prudential purposes. At best, the new knowledge suggestseducatedguesses,butthisisinsufficient.

StrengtheningRegulatoryInstitutions:Europe'sBankingUnion

Regulationisprettyhopelesswithoutgoodsupervision.Iwillthereforeconcludewith a fewwords on a specific institutionalmatter, Europe's nascent bankingunion.MyhunchhereisthatEuropeinthisareaisdefinitelymovingintherightdirection, and yet the new institutions are not only incompletewith respect to

resolution and deposit insurance but also are unlikely to function well ascurrentlydesigned.

There are three rationales for abandoning domestic regulation. The first isexpertise.Inthecat-and-mousegamediscussedearlier,itisratherunlikelythattherewillbeadequateexpertiseinthe27nationalsupervisors;poolingresourcesmayenableregulatorstohaveaccesstomoreexpertise.Second,therearestrongcross-border externalities of a bank failure inEurope, including the impact onforeigncounterpartiesandborrowers,theeffectontheforeigndepositinsurancescheme (for subsidiaries, although not for branches), ring-fencing, andappropriation of liquidity. Third, bank failures may increase government debtandlaterforceEuropetorescuethecountryitself.TheMaastrichtTreatydidnotintegratethefactthatprivatebankdebtcanbecomepublicdebt.

Despite the recent move to create a supranational supervisor located at theEuropean Central Bank (ECB), there are still strong concerns that nationalinterestsremainpowerfulunderthenewsystem.Inmyview,thereisaneedtoprovide the key national supervisors in each countrywith aEuropeanmissionand status.Otherwise, there is a real danger that informationwill bewithheldfrom the ECB precisely when such information should be acted on to triggerearly intervention or resolution. Such a European status would alsomake theboard less captured by national interests. We have observed how decision-making bodies composed primarily of national representatives can be quiteineffective at taking difficult decisions; a case in point is the Economic andFinancialAffairsCouncil(Ecofin)intherun-uptotheEuropeancrisis.

Europe'sbankingunionisstillbyandlargemissingaresolutionauthority.TheCompetitionDirectoratecurrentlysubstitutesforsuchanauthoritybutonlyforsomeaspects;itmakesbailoutsmoredifficultandtriestopreventthedistressedbanks'useofgovernmentfundstogambleforresurrection.Butitdoesnothaveany money to create bad banks or facilitate purchase and assumption. In thefuture, we will need the treatment of individual failures to be financed by adeposit insurance fund collecting fees from insured institutions. SystemicfailureswillrequireastrengtheningoftheEuropeanStabilityMechanism'sfiringpower.

ThecreationofaEuropeandepositinsuranceschemeshouldbemadefeasibleby credible, European-oriented supervision. At this stage, though, losses onunderwater legacy assets could imply substantial cross-subsidies amongcountriesunderamutualizeddepositinsurancescheme.Puttingtheseassetsintobadbanksthatbecomethepropertyofcountriesisfair,andstartingagainwithacleanslatecanmakecommondepositinsurancehappen.Tobecertain,thesebadassets will worsen the countries' financial situation. But in a restructuringsituation, there is no distinction to be made between public debt due toprofligacyandpublicdebtduetonegligenceinprudentialsupervision.

Notes

I am grateful to the conference participants and discussants, and to JoshuaFelmanforhelpfulresponsestotheideasdevelopedinthischapter.

1. See Dewatripont and Tirole (1994) for a more detailed argument on therepresentationhypothesis.

2. Another common source of systemic risk created by bank failures is thecreation of sovereign risk through bailouts and recessions, as dramaticallyillustratedbytherecentexamplesofSpain,Ireland,andCyprus.

3. CoCos, or contingent convertible bonds, are bonds that are converted intoequitycontingentontheoccurrenceofaspecifiedevent,suchasaninsufficientcapitaladequacyratio.

4.See,forexample,Vickers(2012).

5.AsJohnVickersnotedduringtheconference,theUSVolckerrulehassomedrawbacks, which led Europeans (most notably the U.K.Independent BankingCommission, chaired by John Vickers, and the Liikanen Commission, at theEuropeanlevel)tolookfordifferentapproachestoinsulatingretailbankingfrominvestment banking (Liikanen 2012). In a nutshell, the Volcker rule rules outproprietary trading, the ownership of private equity and hedge funds, andactivities leading toaconflictof interest.At the same time, it allowshedging,theproprietary tradingofUSgovernmentsecurities,underwriting,andmarket-making. (Note that market-makers' inventory risk is in essence proprietary

tradingandthatunderwritingisthesaleofaputoption.)

6.SeeICB(2011).

7. TheAIG insurance companywas relatively healthy, aswell as ring-fencedrelative to the investment bank. Thus, we are discussing the bailout of aninvestmentbankratherthanofaprudentiallyregulatedentity.

8. This is not an exhaustive list. For example, one might think of rescues ofshadowbanksthatlendtopoliticallysensitiveentities.

9.Traditional banking activities aremore andmoreperformedby a variety ofplayers, suchashedge fundsor specialpurposevehicles (SPVs).For instance,hedgefundslendtomid-caps.

10.Eachregulatorknowsonlyabouthisorherownjurisdiction.Domesticturfissues, national interests, and mere overload impose limits on the sharing ofinformationamongregulators.

11. Links can also occur through fire sales and wholesale funding (althoughtherewillberestrictionsonthelatter),andnotjustderivativetransactions.Thisinsulationfromthenonretailsectorwouldnotbecomplete,althoughitwouldbeseriouslyreduced.

12.See,forexample,Diamond(1984)andFarhiandTirole(2013).

13.See,forexample,BizerandDeMarzo(1992)andBrunnermeierandOehmke(2013).

14. See Farhi and Tirole (2012). The transfer and concomitant diversion ofsavingstootherusesisnottheonlycost.Lowshort-termratesencouragefuturematuritymismatchesand thereforeacontinued fragilityof the financial sector.Furthermore, institutions that have committed to certain returns for theircustomersarehighlyincentivizedtogambleforresurrection(intheparlanceoffinance,toreachforyield).

References

Adrian, T., and A.B.Ashcraft. 2012. "Shadow Banking: A Review of the

Literature." Federal Reserve Bank of New York Staff Report 580, FederalReserveBankofNewYork,NewYork,October. http://www.newyorkfed.org/research/staffreports/sr580.pdf.

Bizer,D.S.,andP.M.DeMarzo.1992."SequentialBanking."JournalofPoliticalEconomy100:41,61.

Brunnermeier,M.,andM.Oehmke.2013."TheMaturityRatRace." JournalofFinance68(2):483-521.

Dewatripont, Mathias, and Jean Tirole. 1994. The Prudential Regulation ofBanks.Cambridge,MA:MITPress.

Diamond, D. 1984. "Financial Intermediation and Delegated Monitoring."ReviewofEconomicStudies51:393-414.

Farhi,E.,andJ.Tirole.2012."CollectiveMoralHazard,MaturityMismatchandSystemicBailouts."AmericanEconomicReview102:60-93.

Farhi, E., and J.Tirole. 2013. "LiquidBundles." Faculty paper,Department ofEconomics, Harvard University, Cambridge, MA, and Toulouse School ofEconomics, Toulouse, France. http://files.conferencemanager. dk/medialibrary/51432ddb-bbe3-4327-85f4-be3493077470/ images/liquid_bundles_07_08_13.pdf.

Independent Commission on Banking (ICB). 2011. "Final Report:Recommendations." European Corporate Governance Institute, London,September.http://www.ecgi.org/documents/icb_final_report_12sep2011.pdf.

Liikanen,E., and theHigh-LevelExpertGrouponReforming theStructureoftheEUBankingSector.2012."FinalReportoftheHigh-LevelExpertGrouponReforming the Structure of the EU Banking Sector." European Commission,Brussels, October 2. http://ec.europa.eu/internal-market/ bank/docs/high-level_expert-group/report_en.pdf.

Vickers, J. 2012. "Some Economics of Banking Reforms." Discussion Paper632,DepartmentofEconomics,OxfordUniversity,Oxford.

JohnVickers

To start, we need concrete progress with the tooimportant-to-fail conundrum.Weneedagloballeveldiscussionoftheprosandconsofdirectrestrictionsonbusinessmodels.

-ChristineLagarde,Toronto,October25,2012

Aninternationaldiscussiononthestructureofbankinghastakenalongtimetogetgoing.IntheUnitedStates, thestructuraldebatesincetheglobaleconomiccrisisof2008-2009has focusedonhow to implement theVolcker rule,whichprohibitsbanksfromengaginginproprietarytradingthatisnotcustomer-related.In the United Kingdom, the discussion has concentrated on ringfencing-theseparationwithinbankinggroupsof retailand investmentbanking.FranceandGermany are developing the mild hybrid of ringfencing some proprietarytrading. The first international consideration of structural reform was theLiikanen (2012) report for theEuropeanUnion,which, like theUKapproach,recommended separation, within bank holding companies, between depositbankingandtrading.

How are these structural reform initiatives related? What do they seek toachieve?Thischapterconsidersthesequestions,butfirstitputsstructuralreforminthecontextofbankingreformmoregenerally.

StructuralReforminContext

Thecaseforstructuralreformofbanksisnotthatitsolvesthetooimportant-to-failconundrumbyitself.Rather,itisthatstructuralreformisanelementintheoptimal policy package, at the heart of which must also be a much greatercapacitybybankstoabsorblosses.

Apurestructuralistargumentmight runas follows: (1) retailbanking is low

risk and essential, (2) investment banking is risky and inessential, and (3)separating the two confines the public safety net to the low-risk and essentialactivities,leavingmarketforcestodisciplinerisktakingininvestmentbanking.

If only things were so simple! In reality, unless defined so narrowly as toexclude all lending, retail banking, contrary to the first proposition, is risky.Lendinginvolvescreditrisk,muchofwhichislinkedtopropertyvalues,whichcangodownaswellasup.Andastothesecondproposition,evenifinvestmentbanking is separated fully from retail banking, the occurrence of investmentbanking risks can, through various channels, jeopardize essential bankingservices.

Banking reform-both for retail and for investment banking-must thereforehave greater loss absorbency at its core. The crisis exposed a double problemwith banks' loss absorbency. First, banks had nowhere near enough equitycapital-so far toomuch leverage-for the risks theywere running.Second,afterthe thin equity layer had gone, bondholders bore remarkably little loss. Theywould bear losses in a bankruptcy, but to preserve continuity of essentialservices,governmentsstavedoffbankruptcybyprovidingtaxpayersupport.Sobondholdersgenerallycameoutwhole,withthepublicpursebearinglossesthatthe bondholders should have taken. Moreover, where depositor preference isabsent,inabankruptcy,bondholderswouldhavelegalrightstobetreatedonparwithordinarydepositors.

Under theBasel III agreement,which is to be implemented by 2019, bankswillneedtomaintainequitycapitalofatleast7percentofriskweightedassets,orupto9.5percentforbanksofglobalsystemicimportance.Butriskweights,whichwere such a failure in the run-up to the crisis, are typicallywell belowunity, so these ratios allow bank balance sheets still to be a largemultiple ofequity capital.Abackstop limiting leverage to 33 times capital is proposed toaddressthisissue.'

In a commonsense view, 33 is a lot of leverage, and way above corporatenormsintherestoftheeconomy.Remembertoothatthisisforthepost-reformworld. It is true thatmanybanksweremuchmore leveragedpre-crisis, so the

BaselIIIcapitalreforms,whichalsoaddressissuesaboutriskweights,representa fundamental shift fromwhere thingswere. But given the scale of the crisistheir intended destination is hardly ambitious, and might not be reached ifimplementationfallsshort.Itisseriouslyopentoquestionwhetheritiswisetooperate market economies with banking systems as lowly capitalized as theBaselreformswillstillallow.2

Apartial remedy to thisproblemmightcomefrombail-indebt.This isdebtthat is required to be converted to equity, or be wiped out, once low capitaltriggersarebreached.Thus the regulator "bails in"bondholders rather than, astended to happen in the crisis, taxpayers bailing them out. But for this to becredible, the debt subject to regulatory bail-in needs to be unsecured, to havesignificanttime(e.g.,atleastayear)tomaturitysoasnottobepronetorunrisk,andtobejuniortootherdebt,particularlyretaildeposits.Andforbail-indebttomakeamajorcontributiontolossabsorbency,banksneedtohaveenoughofitinaddition to equity (or else far more equity than the Basel minima). Theinternationaldebatehassofarbeensilentonhowmuchbail-indebttorequire,which is currently an important missing piece of international policy on lossabsorbency.

Against this background, what can structural policy-forms of separationbetween retail and investment banking-add? There are three potentialcontributions. First, structural reform can help contain, but cannot whollyprevent,kindsofwithin-bankcontagionfrominvestmentbankingshockstovitalretail services, the continuous provision of which is an economic and socialimperative. An external shock, such as that in 2008 from US subprimemortgagesandrelatedderivatives,directly jeopardizedordinarydomesticHighStreet banking in universal banks lacking internal structure (e.g., RBS).Separation provides an internal firebreak, together with independent lossabsorbencyforretailbanking.Separationalsocreatesthepossibilityofdifferentloss-absorbencyrequirementsforretailandforinvestmentbanking.

Second, separation helps crisismanagement and resolution.Unlike in 2008-2009, it would allow policymakers to pursue different policies for retail andinvestment banking rather than an indiscriminate policy for the entirety.Thus,

separation helps resolvability, which is universally acknowledged to be anessential ingredientofbankingreform. Indeed, it ishard toseehowresolutionplans for large and complex institutions can be credible without ex anteseparationmeasures.'Forrelatedreasons,separationcanassistbanksupervisioninnormaltimes.

Third,bycurtailing the implicit taxpayerguarantee, separation togetherwithhigherloss-absorbencyrequirementsimprovesexantedisciplinesonrisktaking.In particular, if investment banking can no longer be funded by (government-insured)retaildeposits, itwillfacemorerisk-reflectivefundingcosts, justasitshould,andgovernment fundingcostswillbenefitcorrespondingly.Dependingon the design of separation, this in turn should promote lending to the realeconomybeingfundedby,amongotherthings,deposits,includinginsuredretaildeposits.

StructuralReformOptions

Inbroadtermsthestructuralreformofbanksinvolvestwoissues.Oneiswheretodrawtheline(ormultipleslines)betweendifferentkindsofbankingactivity.Theotherishowstrongtheseparationshouldbe.

UnitedStates

TheUnited States provides a useful benchmark for comparing current reformefforts in Britain and Europe. From 1933 through the 1990s,US law had theGlass-Steagall prohibition of affiliation between deposittaking banks andcompanies "engagedprincipally" in securitiesbusiness.Thisprovided for totalseparation between retail/commercial banking and investment banking. Overtime, regulatory permissiveness and market developments led to the effectiveerosionoftheprohibition,aheadofitsrepealin1999.Sincethen,affiliationhasbeen allowed, but a substantial body of law provides for regulation of howdeposittaking banks may and may not deal with their investment bankingaffiliates,notablysection23AoftheFederalReserveActanditsimplementingRegulationW.Theseregulationswerenotalwaysappliedstrictlyinthepast,andtheDoddFrankActof2010makesprovisionforthemtobetightened.

Thus, the repeal of the Glass-Steagall prohibition was not a move fromseparationtounstructureduniversalbanking.Onthecontrary,itwasamovetouniversalbankingwithstructuralregulation.

The Volcker rule, which is also and famously part of the DoddFranklegislation, is an addition to that kind of structural regulation.Like theGlass-Steagalllegislationofold,itisaprohibition.Itsaimistopreventinstitutionsthatdobankingfromengaginginspeculativeproprietarytrading-thatis,tradingthatisnotcustomer-related.The rule likewiseprohibits substantialbankownershipofhedgefundsandprivateequity.Itspurposeisnottoprohibitbankinggroupsfromdoingcustomer-relatedproprietary tradingsuchasmarket-making,whichis among the exemptions from the rule. It has, however, proved difficult todistinguishinpracticebetweenkindsofproprietarytrading,andimplementationoftheruleisbehindschedule.4

Britain

Britain had a bad banking crisis, and is suffering severe economic and fiscalconsequences,inpartbecauseofthesizeofitsbanksinrelationtotheeconomy-roughly five times its annual GDP. In response, the UK is enacting a far-reachingreformprogram, following the reportof the IndependentCommissiononBanking(ICB2011).5

The ICB was asked to make recommendations on structural measures-"including the complex issue of separating retail and investment bankingfunctions"-and "related nonstructural measures" to promote financial stabilityand competition. Competition was another casualty of the crisis but is notconsideredfurtherherebeyondnoting that the tooimportant-to-failconundrum,ifunsolved,isaproblemforcompetitionaswellasforfinancialstabilityandthepublicfinances.

Astotherelatednonstructuralmeasures,theICBfocusedonlossabsorbencyand recommended higher-than-Basel capital requirements for the major retailbanks, accordingly tighter leverage caps, quantified requirements for primaryloss-absorbing capacity (e.g., long-term unsecured debt subject to bail-in

powers)on topof theequity requirement,anddepositorpreference for insureddeposits.

Onstructuralreform,theICBconsideredarangeofoptionsandrecommendedringfencing,requiringthatretailbankingbeconductedinanentityseparatefrominvestmentbanking,withself-standingequityandotherloss-absorbingcapacity,and appropriately independent governance.The retail entity couldbepart of awider banking group but with tight restrictions on its dealings with, andexposuresto,therestofthegroup.

Onwhere to draw the line between retail and investment banking, the ICBapproachwas to identify (1) core retail activities, the continuous provision ofwhichisessential,whichcanbecarriedoutonlyinaringfencedentity,and(2)prohibitedactivities,whichcannotbecarriedoutintheringfencedentity(butarepermitted in the rest of the group).Deposits and overdrafts to individuals andsmallbusinessesarecoreactivities.Prohibitedactivitiesincludetrading(notjustproprietary trading in theVolcker sense), underwriting, derivatives (other thanhedging retail risks), lending to other kinds of financial institution, andmuchoverseas business outside the European Economic Area. A wide swath ofcommercial banking-for example, mortgage lending and lending to largernonfinancialcorporates-isneithercorenorprohibitedintheseterms.IntheICBapproach,banksandtheircustomersarefreetodecidewhetherornottoconductsuchbusinessfromtheretailentity.Inthatsense,thefenceisflexible.Butitisstrongwithrespecttoindependencebetweentheringfencedentityandtherestofabankinggrouptowhichitmaybelong.

TheICBconsideredbutdidnotrecommendtotalseparationbetweenretailandinvestment banking. That would entail considerably higher costs, and for anuncertain financial stability gain. There are circumstances in which totalseparationwouldprotect retail banking fromexternal shocksmore surely thanwould ringfencing. But equally there are situations in which total separationwould be detrimental to financial stability. It would create a stand-aloneintercorrelated domestic banking sector without recourse to support fromelsewhereinbankinggroupsintheeventofadomesticcrisis,suchasapropertycrash. Such crises are perfectly possible, and might well be more acute if

diversificationhasbeenlost.

NeitherdidtheICBfavortheVolckerruleasanalternativetoringfencing.ItwouldhaveappliedtoaverysmallproportionofUKbankbalancesheetsandbyitselfwouldhavegonenowherenearfarenough,intheICB'sanalysis,toshieldretail banking and the public finances from investment banking risks. Yet itwouldhavebeencomplextoimplement.

TheUKgovernmenthasacceptedtheICBrecommendations,includingthosepertainingtocompetition,inverylargepart,6andabankingreformacthasbeenpassed by Parliament.7 The draft legislation has been scrutinized by theParliamentaryCommissiononBankingStandards,whichwasestablishedinthesummerof2012followingtheLondoninterbankofferedrate(LIBOR)scandal.Among other things, that commission has proposed strengthening ringfencingand its durability by way of "electrification"-reserve powers for regulators,subjecttoaframeworkofaccountability,fullytoseparateindividualinstitutions,and even the sector as a whole, in the event that ringfencing were beingundermined. The Parliamentary Commission has also examined the case foradding a form of the Volcker rule to ringfencing but concluded againstlegislating for that now, partly in the light of claims by theBritish banks thattheydonotengageinsuchproprietarytradinganyway.

The casual observer might think that UK reform, with ringfencing, is on adifferenttrackfromreformintheUnitedStates,whichisadoptingtheVolckerrule.But this is to overlook the fact that theUnitedStates alreadyhas, and isstrengthening, a framework of separation between deposit banks and theirinvestment banking affiliates within bank holding companies. Thus, the truerpicture is one of UK-US convergence-on structured universal banking. TheUnited States moved from separation to structured universal banking as theGlass-Steagall prohibition eroded and was finally repealed. It is adding aVolcker rule to that model. The UK is legislating for structured universalbankingagainstabackgroundofalackofstructure.(FortheUKtherewouldbelittlegainbutsomecostinaddingaVolckerruleaswell.)Thus,theUKandtheUnitedStateshavecomefromdifferentdirectionstoabroadlysimilarplace.

Europe

TheEUismovingforwardwithavarietyoffinancialsectorreforms,notablyindirectivesoncapitalrequirementstoimplementBaselIIIandonbankrecoveryand resolution.' The question of EU-wide structural reformwas opened up in2012 by the creation byCommissionerBarnier of the high-level expert groupchairedbyErkkiLiikanen,governoroftheBankofFinland.

The recommendations in the Liikanen (2012) report havemuch in commonwiththeUKreformsoutlinedabove,eventhoughthegroup'sremitwaspan-EU.Itrecommendedthatforbanksabovecertainsize thresholds, tradingshouldbeseparatedfromdepositbanking,butthetwokindsofactivitycancoexistwithinbankholdingcompanies. Inaddition, thereshouldbepowers torequirefurtherseparation if judged necessary for resolvability. It also made a number ofrecommendationsonlossabsorbency,forexampleinrelationtobail-indebtandriskweights.

EspeciallyinlightofthewidergeographicscopeoftheLiikanenreview,ithasaremarkableamountincommonwiththeUKreforms.Bothenvisageamovetostructured rather thanunstructureduniversalbanking. (Separating trading fromdepositbankingisessentiallythesameasringfencingretail/commercialbanking,justasbuildingafencetoseparatelionsfromdeeris thesameasbuildingonearoundadeerparktokeepoutthelions.')NeitherfavorsaddingaVolckerrule,still less adopting that instead of ringfencing. And like the ICB, the Liikanenreportstresseslossabsorbencyanditsinterrelationwithstructuralreform.

There are differences, someofwhich reflect the different geographic scope.Thus, forexample,acommonEuropeanapproach to trading-bookriskweightsmakes sense, whereas a single-country approach could have detrimentalconsequences for geographic arbitrage.Amore surprising difference concernssecurities underwriting, which the Liikanen report would allow with depositbanking. Besides the historical observation that this is the opposite of theprovisions of the Glass-Steagall Act, economic logic suggests that securitiesunderwriting belongs with trading, as it involves writing large put options.Corporate customers could still obtain a comprehensive array of banking

services, including underwriting, from a single bank, but with the risk lyingoutsidethedepositbankentity.

WhereastheLiikanenandUKapproachesarelargelyinharmony,legislationbefore the French and German parliaments takes a minimalist approach tostructural reform. In essence, it requires thatmajor banks conduct speculativeproprietarytradinginaseparatesubsidiary.Verylittleofthebalancesheetoftheaffectedbanks-perhaps1percentorsoseemslikelytobeaffected,yetiftheUSexperiencewiththeVolckerruleisaguide,thedefinitionalandimplementationdifficultiesarelikelytobeconsiderable.Still,thesearestepsinthedirectionofstructural reform. It remains to be seen whether the farther-reaching reformsrecommendedintheLiikanenreportwillbeadoptedacrossEurope.ThereportiscurrentlybeingconsideredbytheEuropeanCommissioninBrussels.

Itsimportanceisheightenedbythewiderquestionofabankingunionfortheeuroarea,which iswidelyseenascrucial for futureeuroareamacroeconomicperformance.Whetherornotabankingunion-ofaformthatentailsadegreeofmutualizationofthecontingentliabilitiesofeuroareabanksbycommondepositinsuranceorotherwise-isagoodideadependsverymuchonthefutureresilienceofthebanksinquestion.Abankingunionhaslargepotentialadvantagesiftheyarewellcapitalizedandsensiblystructured,butcouldbeamistakenenterpriseifnot. Indeed, a union among unstructured and thinly capitalized banks couldmake the tooimportant-to-fail conundrum worse, and put more euro areataxpayersmoreonthehookformorebankingrisks.

Conclusion

The financial crisis began more than five years ago. The prolongedmacroeconomic and fiscal costs of financial crises aremanifest to all.Yet theprogress of banking reform has been mixed, and much unfinished businessremains.With regard to loss absorbency there are concerns that the Basel IIIcapital requirements are unambitious and liable to be watered down inapplication.On theother hand, there are voices, including some in theUnitedStates, calling for substantial stiffening beyond Basel III. There has beensurprisingly little public discussion, outside the UK and Switzerland, of the

complementaryissueofhowmuchbail-indebtmajorbanksshouldberequiredtohave.

Aninternationaldebateonstructuralreformnearlydidn'thappenatallbutisnowunderway.Structural reform initiatives in theUnitedStates (theVolckerrule)andtheUK(ringfencing)mightappeartocontrast,butthedeepertruthisthat both are converging, though coming fromdifferent directions historically,on structured universal banking. The EU's adoption of the Liikanen report'srecommendationswould further promote international convergence, aswell asbeingwise for Europe, especiallywith a banking union on the horizon.Moreresilient banks-with greater loss absorbency, safer structures, and lowercontingenttaxpayerdependence-are,moreover,aglobalpublicgood.

Notes

IamgratefultomyformercolleaguesattheUK'sIndependentCommissiononBanking for many helpful discussions on the topics in this chapter, whichprovidedbackgroundtomyremarksattheIMF's"RethinkingMacroPolicyII"conferenceinWashington,D.C.,onApril16-17,2013.AfullerdiscussionoftheeconomicissuespresentedinthechapterisinVickers(2012).

1.Thedefinitionofcapitalforthisleveragecapiswiderthanequitycapital,sothisallowscloserto40timesleverageinrelationtoequitycapital.

2.AneconomiccaseforverymuchhighercapitalrequirementsthanBaselIIIismadebyAdmatiandHellwig(2013).

3.This is sowhether the resolution strategy involves a "singlepoint of entry"(i.e., typically into the parent company by the home regulator) or "multiplepointsofentry"(e.g.,byseveralregulatorsonageographicorfunctionalbasis).

4.Duffie(2012)providesacriticalanalysisoftheissues.

5.Ichairedthecommission,whichworkedfromJune2010toSeptember2011.TheothermemberswereClareSpottiswoode,MartinTaylor,BillWinters,andMartinWolf.

6.SeeHMTreasury(2012,2013).

7.TheFinancialServices(BankingReform)ActwaspassedinDecember2013.

8.InEuropebeyondtheEU,reforminSwitzerlandisnoteworthy.LiketheUK,Switzerland has (two) very large banks in relation to GDP, and established acommission on banking reform. The Swiss Commission recommended muchhighercapitalrequirementsthanthatfortheBaselIIIbaseline.

9.AmetaphorbyMartinTaylor.

References

Admati, A., andM.Hellwig. 2013. The Bankers' NewClothes. Princeton, NJ:PrincetonUniversityPress.

Duffie, D. 2012. "MarketMaking under the ProposedVolcker Rule." Facultypaper,RockCenterforCorporateGovernanceandGraduateSchoolofBusiness,Stanford University, Palo Alto, CA. http://www.gsb.stanford.edu/news/packages/PDFNolcker_duffie_011712.pdf.

HM Treasury. 2012. Sound Banking: Delivering Reform. Cm 8453. London:HMSO,October.

HM Treasury. 2013. Banking Reform: A New Structure for Stability andGrowth.Cm8545.London:HMSO,February.

Independent Commission on Banking (ICB). 2011. "Final Report:Recommendations." European Corporate Governance Institute, London,September.http://www.ecgi.org/documents/icb_final_report_12sep2011.pdf.

Liikanen,E., and theHigh-LevelExpertGrouponReforming theStructureoftheEUBankingSector.2012."FinalReportoftheHigh-LevelExpertGrouponReforming the Structure of the EU Banking Sector." European Commission,Brussels, October 2. http://ec.europa.eu/internal_market/ bank/docs/high-level_expertgroup/reporten.pdf.

Vickers, J. 2012. "Some Economics of Banking Reform." Rivista di PoliticaEconomica4:11-35.

AdairTurner

In 2009, Queen Elizabeth visited the Economics Department of the LondonSchoolofEconomics,discussedthefinancialcrisis,andaskedasimplequestion:"Whydidnooneseeitcoming?"Itwasagoodquestion,butonethatcouldnowbeexpanded,becausethereweretwofailures.

First, therewas a failure to foresee the crisis comingat all-a failure in, say,2005 or 2006 or 2007 to understand that we were heading toward a majorfinancial crash. There were some notable exceptions. To different degrees,NourielRoubini,RaghuRai an, andBillWhite issued somewarnings.Butonthewhole, theworld'scentralbanksand regulators, financeministries,and theIMF not only did notwarn of impending disaster but in general propagated athesisthatfinancialinnovationandincreasingfinancialintensityhadmadecriseslesslikely.

And thena second failureoccurred in spring2009,once the severecrisisof2008was behind us: a failure to foresee how difficult and slow the recoverywouldbe.Noofficial forecastanticipatedanything like thescaleand lengthofthesubsequentrecession,andalmostnooneanticipated thescaleof thepolicystimulus we would deploy in an attempt to offset recessionary forces. Therewere, tomyknowledge,noofficialsectorormarketforecaststhatpolicyrates,having fallen to thezero lowerbound,would stay there for fouryears (so far)andprobablyforseveralyearsmore.

How did this double failure occur? There were many specific reasons. Inparticular, theflawedstructureoftheeuroareaplayedakeyroleinexplainingwhytheinitialcrashproducedasubsequentmajoraftershock.ButinthischapterI present the argument-not original but I think so fundamental that it isworthstressing-that central toourpoor foresightwasa sustained failureover severaldecades to appreciate and focus on the central importance to both financial

stabilityandmacroeconomicstabilityofthescaleofdebtcontracts-thelevelofleveragebothwithinthefinancialsectorandwithintherealeconomy,andintheprivatesectorquiteasmuchasthepublicsector.

Ialsoarguethatoncewedofocusonthefundamentalimportanceofdebtandleverage,wemayneed toconsidermoreradicalpolicies,bothmacroprudentialandmonetary,thanwehavedonesofar.

Mervyn King pointed out in a lecture last autumn that the dominant newKeynesian model of monetary economics "lacks an account of financialintermediation,sothatmoney,creditandbanksplaynomeaningfulrole"(King2012).

In retrospect, thatwasaveryoddomission.Afterall, sincemonetarypolicyworks through an interest rate, itmust presumablywork through the financialintermediation system, debt contracts, and banks. But it was not only an oddomission,itwasalsoverydangerous,becausedebtcontractshaveveryspecificfeaturesthathavemajorimplicationsforfinancialstabilityandmacroeconomicstability.

Those featureswereobvious to and amajor focus for severalmidtwentieth-centuryeconomistswritingamidthewreckageof the19291933financialcrashand the subsequentGreatDepression,writers such as IrvingFisher andHenrySimons (Fisher1933,1936;Simons1936).And if one reads those economistsagain,what isstrikingis thecentral role theyascribe todebtcreationanddebtdestruction in the origins and development of theGreat Depression, and howradicalthepolicyprescriptionswerethattheyproposedinresponse.

Henry Simons is typically thought of as a foundational figure of Chicagoschool economics and laissez-faire policy prescriptions. But he argued that"private initiative has been allowed too much freedom in determining thecharacter of our financial system and in directing changes in the quantity ofmoneyandmoneysubstitutes."Hewantednotonly tomake fractional reservebankingillegalbuttoseverelyrestraintherolethatevennonbankdebtcontracts,asopposedtoequitycontracts,couldplayintheeconomy.

Simons and Fisher believed that the more debt contracts there were in theeconomy,themorefragileitwouldbecomeandthemorevulnerabletoharmfulfinancial cycles of the sort that Claudio Borio describes in chapter 6 of thisvolume.There are at least five interlocking reasonswhy that couldbe so-whydebtcontractsaredifferentfromequity.

FirstisthephenomenonGennaioli,Shleifer,andVishny(2010)havelabeled"local thinking": the tendency to ignore in thegood times thedownside tailofthedistributionofpossibledebt returns.Whenweholdanequitycontract, theobservedmovementofequitypricesremindsuseachdaythatreturnscangoupordown,thatweholdariskyinvestment.But ifweholddebtcontracts, in thegoodtimesweobserveonlythenondefault,full-payoutresult.Thereisthereforeadangerthatinvestorswillcometobelievethatinherentlyriskydebtcontractsare risk-free investments, and therefore a danger that the aggregate value ofapparently lowrisk debt instruments created will exceed the value that couldactuallyberisk-free,giventheunderlyingrealeconomicrisksfacingcompaniesandhouseholds.Asaresult,asGennaioli,Shleifer,andVishny(2010)putit,afreefinancialsystemcanmanufacturelargevolumesofdebtsecurities"thatowetheirveryexistencetoneglectedrisk."

A second reason is the rigidities and fragilities created by default andbankruptcy processes.AsBenBernanke noted, "In a completemarketsworld,bankruptcy would never be observed." Instead, we would see smooth, nonjumpy, continual redefinition of debt terms-steadily converting to equity-ascreditworthiness declined (Bernanke 2004). And as Charles Goodhart andDemetriusTsomocosexploredinthe2011MayekawaLecture,oneofthemaindeficiencies of modern macroeconomic theory, with its representative agentfiction,istheabsenceofthepossibilityofdefault-ofcompanies,households,orbanksthemselves(GoodhartandTsomocos2011).Butintherealworld,wedoseedefaults,andasbothBernankeandIrvingFisherdescribed, theiroperationcanplayamajorroleinthepropagationofrecessionsanddepressions.

Third,debtisquitedifferentfromequitybecauseithastobecontinuallyrolledover.One could imagine aworkingmarket economy inwhich the new equityissue market closed entirely for, say, five years. There would obviously be

disadvantages, but such closure would not in itself tip the economy into arecession because the preexisting equity investmentswould still exist-they arenot continually repaid. But debt contracts continually mature, so thatmacroeconomic stability depends crucially on the smooth continuity of newcredit supply, a vulnerability that made Simons so wary of short-term debtcontracts that he hankered after an economy in which only long-term debtcontracts(hesuggested50yearsormore)wouldbeallowed.

Fourth,akeyfocusofSimonsandFisheristhefactthatbanksdonotjust,astoomanyeconomicstextbookswronglyimply,intermediateexistingmoneyintocredit.Rather,theycreatenewcreditandmoneydenovo,andintroducematuritytransformationrisks.

And fifth, thepotential exists forcredit extensionagainstassetsinparticular,realestate-toitselfinfluencethevalueofthoseassets,aprocessthatcanunleashMinsky-typecyclesinwhichmorebankcreditcreationcanbegetyetmore,withboth lenderandborrower incentivesandeconomicscreatingstrongprocyclicaleffects. Inotherwords,acreditandasset-pricecyclecomes intobeing that,asClaudioBoriomentionedinchapter6,isnotjustapartbutthecentralfeatureofthefinancialcycle.

Thesedistinctivefeaturesofdebtcontractshave,Ibelieve,twoimplications:

• First, we cannot assume that the freemarket, left to itself, will arrive at anoptimalbalanceofdebt andequity contracts. Indeed, itwill havea systematiclong-termtendencytocreatetoomuchdebt,toomuchleverage.

• Second, themore leverage-probably bestmeasured as the debt-toGDP ratio-beyond some point, the more potentially fragile become both the financialsystemandthemacroeconomy.Thatiswhattheoryshould,Ibelieve,tellus,andwhatsomeempiricalresearchisbeginningtentativelytoconfirm.

In an important paper from the Bank for International Settlements, SteveCecchetti and EnisseKharroubi (2012) suggest that the debt-toGDP ratio andlong-termgrowthratesmayberelatedinaninverseUfunction,withgrowthfirstincreasingover some rangeof increasing financial intensitybut thendeclining

abovesometurningpoint.

Thosefindingswouldnothavesurprisedthemidtwentieth-centuryeconomistswho wrote in the wake of the 1929-1930 crash. But we somehow forgot theimportanceofdebtstocklevelsorwronglydismissedthemasunimportant,andasa result ignoredor assumedbenign thehuge increases in real economyandfinancial system leverage, in both the formal banking system and the shadowbankingsystem,thatoccurredoverseveraldecadesaheadofthecrisis.

And in our regulatory response to the crisis, I still fear that we have notrecognizedthescaleofthefinancialfragilityrisksthatdebtcontractscancreate.AndIfearthatwehavenotyetbeenadequatelyradicalinourpolicyresponse.

We have significantly increased bank capital requirements. But there are, Ibelieve, persuasive arguments-as, for instance, set out in Anat Adman andMartinHellwig's(2013)newbook-thatoptimalbankcapitalratios(theratioswewould set ifwewere benevolent dictators of a greenfield economy)would bemuchhigher,morelike25percentto30percent.

Even more fundamentally, our regulatory response still does not overtlyrecognizethatthelevelofrealeconomyleverageisapotentiallyvitalvariable.

Take, for instance, the indicativemeasure proposed for the operation of theBasel III countercyclical buffer. It is that the buffer should be raised if creditgrowthisrunningsignificantlyabovepast trendandreducedif it isbelowthattrend.Thatimpliesthataslongascreditgrowthisinlinewithtrend-aslongasitis smooth growth-that is fine, even if the trend growth is more rapid thannominal GDP growth. And even, therefore, if the debt-to-GDP ratio isrelentlesslyrising.

I believe that measure is inadequate and that we have to start treatingaggregateleveragelevels(privateasmuchaspublic)asacrucial indicatorandseek policy measures to contain that level through, for instance, limits onmaximumallowedloan-to-incomeratios.

DeflationintheDownswingoftheCycle

Wefailedtoseethecrisiscoming,then,becausewetreatedrelentlessfinancialdeepeningaseitherneutralorbenignratherthandangerous.Butwhataboutthefailure inearly2009,once thecrisishadalreadyoccurred, toseehowdifficultrecoverywouldbe?

Here again I think our crucial blindness was related to debt: a failure toanticipate thestrengthof thedeflationary impetuscreatedbyattemptedprivatesectordeleveragingintheaftermathofanexcessdebtcrisis-afailureofforesighteventhoughJapanhad,overtheprevious20years,providedastrongillustrationof that effect, well described in Richard Koo's account of a balance sheetrecession(Koo2009).

• Left by the upswing with debt stock levels they now believe excessive,constrained companies or households become determined to deleverage. Theyseek to generate financial surpluses with which to pay down debt. And theirinvestmentandconsumptiondecisionsbecomehighlyinelastic toreductions ininterestrates.

•Policy rates at the zero lowerbound, therefore, have little stimulative effect,andprivatedemandfalls.Publicdeficitsrise,usefullyprovidinganoffsettolowprivatedemandbutattheexpenseofrisingpublicdebtlevels.

•Attheaggregatelevel,leveragedoesn'tactuallyfallbutsimplyshiftsfromtheprivate to the public sector-the pattern clearly seen in Japan over the last 20yearsandintheUnitedKingdom,Ireland,Spain,andtheUnitedStatesoverthelastfour.

Thatdescriptionis,Ithink,persuasive.EggertssonandKrugman's2012paperprovidesaformalmathematicalexplanationoftheprocessesatwork,integratinginto a new Keynesian framework the assumption that some agents areconstrained by debt stock concerns and, as a result, make different marginaldecisionsthannonconstrainedagentsdo.

Again,whatissurprisinginretrospectishownovelthatintegrationisandhowabsent debt stock levelswere frompreviousmodels.After all, in the arena ofinternational monetary policy we have discussed for decades the potentially

deflationary impact of an asymmetry of response between debtor and creditornations,theformerconstrainedtocutbackdemand,thelatternotconstrainedtostimulate. What Eggertsson and Krugman do is simply explore the samepotential asymmetry asbetween constrainednet debtors andunconstrainednetcreditorswithinaneconomy.Thatasymmetry,deleveraging,andbalancesheetrecessionprocess is, Ibelieve,acrucial factor inexplaining theslowrecoverysince2009.

So the question is, what policies will best help navigate this inherentlychallengingdeleveragingenvironment?Again,Isuggestthatweneedtobeopentomore radicalpolicies thanhavebeendeployedso far.Let's suppose thatwewant to stimulate aggregatenominaldemand, toproduce amore rapidpathofnominal GDP growth, as Michael Woodford proposes in chapter 4 of thisvolume Of course, we might not want to. Mervyn King warns in chapter 3against assuming that deficient aggregate demand is the only problem, and Icertainlyagreewith that,particularly in theUK.But let's suppose it isat leastpartoftheproblem:howbesttostimulate?Thepredominantcurrentapproachisbyusingunconventionalmonetarytools.

Policy rates have been at the zero lower bound for four years, butunconventionalpolicies-quantitativeeasing,twists,crediteasing,creditsubsidy,central bank liquidity support, forward guidance-are available and have beendeployed.Allthesetoolsworkviaoneoftworelatedtransmissionmechanisms:either (1) they seek to influence a wider set of interest rates than the currentpolicyratealone:longaswellasshort,expectationsofforwardratesaswellascurrentrates,interestratesactuallypaidbyendborrowersintherealeconomyaswellasthepolicyrateinwholesalemarkets,or(2)theyworkbyassetpriceandportfoliobalanceeffects:higherbondorequitypricesproducingwealtheffectsandsearchforyield.

Thebestevidenceseemstobethatthesepolicies,workingviathesechannels,havehadsomepositive impactsonbothprice levelsandrealoutput.But theremustbetwoconcernsaboutthesepolicies.

•First,suchpolicies,workingthroughtheseindirectandexpectationalchannels,

must be potentially subject to decliningmarginal effectiveness in the specificcircumstances of debt overhang in a balance sheet recession. Because if wereally do have debt-constrained companies or households focused ondeleveraging,theymayberelativelyinelastictoreductioninlong-termratesortotheratestheyactuallypay,aswellastothecurrentpolicyrate.

• Second, a long, sustained period of low interest rates must have adverseconsequences of the sort JeremyStein highlights in chapter 10 andwhich thelatest Global Financial Stability Report describes-with financial stability riskgrowingasfinancialmarketagentsseektotakeonleverage,towriteputoptions,and to engage in leveraged carry trades. The elasticity to lower current andexpectedinterestratesofresponsesfocusedonassetspeculationandsearchforyield via financial innovationmay turn out to be greater than the elasticity ofresponse of real economy investment and consumption. A sole reliance onmonetarypolicystimulus,workingviatheseindirectandexpectationalchannels,maythereforecarrydangers.

An alternative, or a complement, is fiscal stimulus-directly injectingpurchasingpowerintotheeconomyratherthanoperatingviaindirectchannels.The classic argument against this is that the first-roundeffects of the stimulusare offset by crowding out, by the central bank response, and by Ricardianequivalenceeffects,makingfiscalmultiplierslow.

But Brad DeLong and Lawrence Summers's (2012) recent paper providespersuasive arguments for believing that in the current conditions of debtoverhangandprivate-sectordeleveraging,andwithcentralbankscommitted tomaintain interest rates at the zero lower bound for several years ahead, fiscalmultipliers are bound to be far higher. And thattogether with the potentiallimitations and dangers of monetary policy working entirely via indirectchannels-suggeststheneedforcautionaboutapolicyprescriptionthatcombinesrapid fiscal consolidation offset, it is supposed, by unconventional monetarystimulus.

Butequally,wecannotbeunconcernedbydramatic increases inpublicdebtlevels.RichardKoomayberightthatwithoutthelargeJapanesefiscaldeficits

over the last 20 years, the Japanese economy would have suffered a greatdepression.ButJapanesegovernmentdebtlevelsover200percentofGDPandrisingcannotsimplybeignored.AndtotheextentthatJapaneseconsumersandcompanies are aware of that debt burdenwhich, beyond some level, theymustsurelybe-suchdebtlevelsmayindeedgenerateRicardianequivalenceoffsetstoconfidenceandthustodemand.

Asa result, it seemspossible thatbalancesheet recessionscanplaceus inapositionwheretheauthoritiesrunoutofammunition-thepuremonetarybulletsineffectiveorendangeringadversesideeffects, thefiscalmagazineempty.Butfiscalandmonetaryauthoritiescombinedneverrunoutofammunition.TheycanalwaysdowhatBenBernankeproposedforJapanin2003.

•Heproposed"ataxcutforhouseholdsandbusinessesthatisexplicitlycoupledwith incrementalBankof Japanpurchasesofgovernmentdebt, so that the taxcutisineffectfinancedbymoneycreation."

• He stressed that it would be important to be clear that "much or all of theincreaseinthemoneystockisviewedaspermanent."

•Hesuggestedthatbusinessesandcompanieswouldwillinglyspendthemoneyreceivedsince"nocurrentorfuturedebtserviceburdenhasbeencreated"(i.e.,noRicardianequivalenceoffsetwouldlogicallyarise).

• He argued that the debt-to-GDP ratio would fall, since there would be noincreaseinnominaldebtbutariseinnominalGDP.

This ishelicoptermoneyor,as I labeled it ina recent lecture,"overtmoneyfinance of an increased fiscal deficit" (Turner 2013). It is an available policyoptionand,intechnicalterms,therearenoreasonsforbelievingthatitwouldbemore inflationary (i.e., would produce a less favorable balance between priceandoutputeffects)thananyotherpolicythatwouldbesuccessfulinstimulatingnominalGDP.

Buttherecouldcertainlybepowerfulpoliticaleconomyreasonsforexcludingthisoption,fortreatingitastaboo.Becauseifthetaboowerebroken,politicians

might want to use the option in excess and all the time, rather than in smallamountsandinthespecificextremeconditionsofbalancesheetrecessions.Andthat may argue for not being explicit about debt monetization. As MichaelWoodfordcommentedinconversationattheLondonBusinessSchool(Reichlin,Turner,andWoodford2013):

• If you inject a fiscal stimulus against the background of a central bankcommittedtomaintaininglowinterestratesforseveralyears,

• and that commitment ismade credibleby aprice level ormoneyGDP leveltarget,

• and if you accept the possibility-indeed, the likelihood-that some of theincreaseinthemonetarybasewillturnoutpostfactotobepermanent,

• then you have a strategy substantially very close to Bernanke's helicoptermoneybutwithoutbreakingthepotentiallyvaluablepoliticaleconomytaboo.

However, that strategystilldoes requireadegreeofcoordinationof fiscalandmonetarypolicy, and an acceptanceby the central bank that it is facilitating afiscalpolicystimulusrather thanoffsettingfiscalausterityviamonetarypolicystimulusworkingentirelyvia interest rate, expectational, andportfoliobalancechannels.

Thatcoordinationismadeessentialbythespecificconditionsofthepostcrisisbalance sheet recession. Itwouldhavebeenbetter ifwehadnevergotten intothissituationinthefirstplace,neverallowedexcessiveleveragetodevelop.Butwedid,andweneedtodesignpolicytodayinthespecificconditionscreatedbythosepastpolicymistakes.

Thecrisisoccurredandwasnotforeseenbecausewefailedtoappreciatethefundamental importance of aggregate leverage. And our failure to foresee theslow and difficult recovery reflected the fact that ourmacroeconomicmodels,whileincorporatingstickypricesandwages,largelyfailedtoreflecttheperhapsstillmoreimportantrigiditiesintroducedbydebtcontracts,debtstocklevels,anddefaultprocesses.

Integratingfinancialstructure,debtcontracts,default,andthebankingsystemintomacroeconomicmodelsisthereforecriticalbutstillatanearlystage.

Whichmakesthesubtitleoftheconference,"FirstStepsandEarlyLessons,"verywellchosen.

References

Admati, Anat, andMartin Hellwig. 2013. The Bankers' New Clothes:What'sWrong with Banking and What to Do about It. Princeton, NJ: PrincetonUniversityPress.

Bernanke,BenS.2003."SomeThoughtsonMonetaryPolicyinJapan."Speechbefore the Japan Society of Monetary Economists, Tokyo, May. http:llwww.federalreserve.gov/boarddocs/speeches/2003/20030531.

Bernanke, Ben S. 2004. "Non-Monetary Effects of the Financial Crisis." InEssaysontheGreatDepression.Princeton,NJ:PrincetonUniversityPress.

Cecchetti,StephenG.,andEnisseKharroubi.2012."ReassessingtheImpactofFinance on Growth." BIS Working Paper 381, Monetary and EconomicDepartment, Bank for International Settlements, Basel, July. http://www.bis.org/publ/work381.pdf.

DeLong, Brad, and Lawrence Summers. 2012. "Fiscal Policy in a DepressedEconomy." Brookings Papers on Economic Activity, Brookings Institution,Washington, DC, Spring. http://www.brookings.edu/ -/media/Projects/BPEA/Spring%202012/2012a_DeLong.pdf.

Eggertsson,Gauti B., and PaulKrugman. 2012. "Debt, Deleveraging, and theLiquidity Trap: A Fisher-Minsky-Koo Approach." Quarterly Journal ofEconomics127(3):1469-1513.

Fisher, Irving. 1933. "The Debt-Deflation Theory of Great Depressions."Econometrica1(4):337-357.

Fisher, Irving. 1936. "100%Money and the Public Debt." Economic Forum,Spring,406-420.

Gennaioli,Nicola,AndreiShleifer,andRobertVishny.2010."NeglectedRisks,Financial Innovation, and Financial Fragility." NBER Working Paper 16068,NationalBureauofEconomicResearch,Cambridge,MA.June.http://www.nber.org/papers/w16068.

Goodhart, Charles, and Dimitrios Tsomocos. 2011. "The Role of Default inMacroeconomics." Mayekawa Lecture, Institute for Monetary and EconomicStudies, Bank of Japan, Tokyo, November. Monetary and Economic Studies,vol.29.http://www.imes.boj.or.jp/research/abstracts/english/me29-4.html.

King,Mervyn. 2012. "TwentyYears of InflationTargeting." StampMemorialLecture,LondonSchoolofEconomics,London,October.

Koo,Richard.2009.TheHolyGrailofMicroeconomics:LessonsfromJapan'sGreatRecession.NewYork:JohnWiley&Sons.

Reichlin, Lucrezia, Adair Turner, and Michael Woodford. 2013. "HelicopterMoney as a Policy Option." VoxEu, May 20. http://www.voxeu. org/article/helicopter-money-policy-option.

Simons,Henry.1936."RulesversusAuthoritiesinMonetaryPolicy."JournalofPoliticalEconomy44(1):1-30.

Turner,Adair.2013."Debt,Money,andMephistopheles:HowDoWeGetOutof This Mess?" Lecture at Cass Business School, February.http://www.fsa.gov.uk/library/communication/speeches/2013/0206-at.

JaniceEberly

Whereasmuchofthepolicyinnovationduringthefinancialcrisisof2008-2009focused on monetary policy-and especially unconventional monetary policy-there has since been renewed attention to fiscal policy. The monetary focusmanifestedinpartbecausemonetaryauthoritieswereabletomovequicklyanddecisively during the crisis, whereas fiscal policy tends to be slower, both indecision making and in implementation. But the renewed attention to thepotentialpoweroffiscalpolicycameastheseverityofthecrisisdemandedthatall available tools be brought to bear and also as the limits of monetaryeffectivenessmayhavebeenreached.Thenatureofthefinancialcrisisandtheuseofunconventionalmonetary toolsbroughtaboutaconceptual rethinking inmonetaryeconomics,particularlytoincorporatefinancialmarketsintomonetarymodels.Therehasnotbeensuchaground-uprethinkingsofarinfiscalpolicy,although recentwork suggestswemight consider fiscal policydifferently in aworldwheremonetarypolicyisatthezerolowerbound.

Iwillbreakoutthequestionsaddressedinthisandthenextfewchaptersintothe phases of the business cycle,which implicitly focuses our thinking on thecyclicalroleoffiscalpolicy.However,Iwillreturntothepotentialroleoffiscalpolicyineconomictrendsattheconclusionofthisintroduction.

Starting with the pre-recession or boom period in an economy, how muchheadroomshouldacountrymaintain in itsdebt-to-GDPratio?Thisquestion isthe counterpart to the corporate finance concept of debt capacity and thequestionofhowmuchsparedebtcapacityshouldbeheldinabeyance.Thecrisistaughtusthatwhilepolicymakerscuttaxes,increasedspending,andrandeficitsduringcyclicalbooms-eitheroutofpoliticalexpedienceorasa strategy toputdownwardpressureonthesizeofgovernment-theyusedupborrowingcapacity

that was needed later when the crisis eventually hit. Downturns are naturallyassociatedwith largerdeficits, at the least fromautomatic stabilizers andevenmorefromactivefiscalpolicy,andwiththeneedtotapdebtmarkets.Thistaskistypicallyeasierandcheaperwhenthedebt-to-GDPratioinheritedfromboomtimes is relatively low.Such a precautionary approach to debt capacity is lesscontroversial than the thorny quantitative question that follows: how muchborrowing capacity should a country build up outside downturns, and if itdepends on state-specific factors, what determines how large this capacityshouldbe?

Second,coming to thedownturn,how largeare the fiscalpolicymultipliers,and if they are state-contingent, what do they depend on? Research so faremphasizes that fiscal multipliers can be much larger when monetary policyreaches the zero lower bound. Other work has shown that excess capacity oroutputgapsintheeconomyincreasefiscalmultipliers.Thisanalysisiscrucialtopolicymakers'approachingfiscalconsolidationwhenaneconomyisstillweak,eitherintheaftermathofacrisisorworse,duringacrisisitselfassociatedwithstructural fiscal imbalances. Policymakers face the essential conundrum thatfinancialmarketsare lookingforcredible improvements in thestructural fiscalstanceatthesametimethattheeconomywillsuffermostfromfiscaltightening.

Oneapproachtothesequestionsistoputinplacefiscalrules, justascentralbanks have proposed monetary rules to develop credible policy. In principle,suchrulescanallowfordeficitspendingwhentheeconomyisweakestandthemultipliersarelargest,withacrediblecommitmenttofiscalsustainabilityastheeconomyimproves.Rulescanalsohelp imposedisciplineduring"good times"to rebuilddebt capacity andbringdown thedebt-to-GDP ratio after a cyclicaldownturn.Similarly,suchrulescanallowforcyclicaldeficitsduringdownturns,alongthelinesofautomaticstabilizers,whichnaturallyrolloffastheeconomystrengthens. Rules can also ease the decision-making lags that limit theeffectivenessoffiscalpolicyattypicalbusinesscyclefrequencies.Theeffectsofthese stabilizers can be large: The Congressional Budget Office (CBO 2013)estimates that spending and forgone revenue associated with automaticstabilizersaveragedover$350billionperyearintheUnitedStatesfrom2009tothepresent(FY2013),totalingmorethan2.5timesthetotalstimulusincludedin

theAmericanRecovery andReinvestmentAct.Although the use of automaticstabilizers iswellestablished, strengtheningandbroadening fiscal rulescanbeproblematic.Moregeneralfiscalruleshavewell-knownproblemsbecausetheyarerigidbydesignanddifficulttocustomizetoparticulareconomicconditions;indeed, allowing judgment or "escape hatches" to accommodate unforeseencircumstancescanunderminethecredibilitythataruleisdesignedtodevelop.

Finally,oncethedownturnhaseased,howrapidshouldfiscalconsolidationbeduring a recovery? Policymakers often undertake rapid fiscal consolidationunderpressurefromfinancialmarketstoreducesovereigndebt,oratleasttheirborrowing. Countries not under immediate pressure argue that a reduction inborrowing is needed to forestall such pressure eventually. When shouldpolicymakersstarttotakethisconcernintoaccount?Shouldpolicymakersreactbefore reachinganupperboundon thedebt-to-GDP ratio that they shouldnotbreach, or similarly, that financial markets will not tolerate? Concern aboutinvestors'appetiteforsovereignriskcanleadpolicymakerstotightentooearlyortoomuch,becausethecostsareperceivedtobeasymmetric:Waitingtoolongandriskingawithdrawaloflendingcanbeverycostly.

Thesequestionsarealmostentirely focusedoncyclicalmanagementofdebtanddeficits.Yettheseverityandpersistenceoftherecentrecessionandtheslowpaceofsubsequentrecoveryraisetheconcernthatthecrisiswillhavealastingeffect on the macroeconomy, perhaps by reducing potential output. Thepossibilitythatoutputmaybepermanentlyaffectedprofoundlyraisesthestakesofadownturn,yet themechanism throughwhicha financial crisisora severedownturn affects the productive capacity of the economy is not yet wellunderstood.Thereisevidencethatlong-termunemploymentcausesdeteriorationof a worker's skills and reduces the probability of re-employment. Otherpotentialmechanismsarelessclear-cut.Isthereareductionininnovationorintechnologicalchangeembodiedininvestmentinphysicalcapital?Doesreducedlabor mobility or reallocation hurt productivity? Some of these mechanismswould benefit from fiscal intervention, either broadly to increase economicactivity, or specifically, for example, to promote skill development and jobmatching.Yetwithoutunderstandingwhatcauses the long-termdamage to theeconomy,anypolicyinterventionislargelyspeculative.

Finally, I began by emphasizing the ground-up rethinking in monetaryeconomics to explicitly incorporate financial markets. Monetary policy hasembracedcreditmarkets as amechanism formonetary stimulus, and the samemay be true for fiscal policy. Fiscal use of credit policy has been employedprimarily for microeconomic reasons: to remedy market failures and supportsocialbenefitswhere theyexceedprivatebenefits. Inpractice,however, creditpolicy was widely used in the crisis by fiscal authorities, particularly in thehousingmarketandtosupportsmallbusiness.Aswerethinkfiscaleconomics,policymakers could likely benefit from a better conceptual and empiricalunderstanding of the potential role of credit and macroprudential policy incountercyclicalfiscalpolicy.

Reference

Congressional Budget Office (CBO). 2013. "The Effects of AutomaticStabilizerson theFederalBudgetasof2013."CBO,Washington,DC,March.http://www .cbo.gov/sites/default/files/ cbofiles/attachments/43977_AutomaticStablilizers3-2013.pdf.

AndersBorg

The theme of this conference is rethinkingmacro policy, and I would like tooutline some important lessons from theglobaleconomiccrisis since2008 forthedesignoffiscalpolicy,drawingonmyexperienceasSweden'sministerforfinanceandasamemberoftheEcofinCouncil(EconomicandFinancialAffairsCounciloftheEuropeanUnion)overthepastsixyears.InthischapterIfocusonthecurrentroleoffiscalpolicyandonanumberofimportantlessonsthecrisishas taught us, and concludewith a discussion of the problems Europe has totackle.Thischapterhasthreebasicmessages:

•First,fiscalpolicyshouldbeusedasastabilizationtool.Itisveryimportanttorealize that institutions have to be reformed so that fiscal policy can play anactiveroleinstabilizingtheeconomy.

• Second, debt is important. High debt is a constraint on rational economicpolicy.

•Third,themaindifferencebetweenfiscalpolicyviewsintheUnitedStatesandinEuropehastodowiththefactthattheEuropeanproblemsarefundamentallystructural.Thisdoesnotmean there isnodemand shortage inEurope,but theproblemsaretoalargerextentstructuralthanarethoseintheUnitedStates.

TheCurrentSituation

If we take a longer view of this crisis, one stylized fact is that Europe hasactuallygrownslowly-lessthanhalfapercentagepointoverthelastsixorsevenyears. Even in a 10-year perspective, Europe has growth of close to onepercentage point. Fundamentally, then, there is a long-term, not a short-term,probleminEurope.Thishasalsofedintounemployment.Unemploymentinthe

euroareaisnowcloseto12percent.TheemploymentratesintheUnitedStatesand Europe are at least 5 to 10 percent lower than they should be in a well-functioningeconomy.Andontopofthatwehaveaveryheavyincreaseinthedebt level (seefigures15.1a-d).Whatmightbe theappropriate fiscalstance inthesecircumstances?

Figure15.1

KeyIndicatorsfortheUnitedStates,Japan,theEuroarea,andSweden

a.GDPgrowth.Averagegrowthrate,2007-2012(in%).b.UnemploymentrateasofQ42012(in%).Euroareadatarepresent17countries.c.Employmentrate,all persons ages 15 to 64, as ofQ4 2012 (in%). Euro area data represent 17countriesandareasofQ32012.d.GeneralgovernmentgrossdebtasapercentofGDP.

Sources:a-c.OECD.d.EuropeanEconomicForecast,spring2013.

FiscalStance

Weshouldbeverycautiousaboutassumingthatthecrisisisover.Thereisstilla

need for both fiscal andmonetary policy to remain expansionary. At present,fiscalpolicyconsiderations involvestrikingadifficultbalanceamongtheneedforconsolidationofpublicfinances,supportinggrowth,andtacklingpersistentlyhighunemployment. In striking thisbalance, it is imperative thatgovernmentsnot lose the confidence ofmarket actors, businesses, and citizens. This policybalancewill obviously vary from country to country. Letme illustratewhat Imean.

Italy and Spain and other countries under considerable pressure from themarketsmustmeettheirfiscalpolicycommitmentsatasufficientpacetorestoremarketconfidence.There isaclear risk thatbacktrackingonmeasuresalreadydecidedoncouldleadtoareturnofmarketuncertainty.Butitcouldbearguedthat weaker growth than expected should not be met with additionalconsolidationmeasures.

Countries in a somewhat stronger position but still facing considerablechallenges, such as France and the UK, should let automatic stabilizers workfully.ItisclearthattheFrenchgovernmentwillneedmoretimetoreachthe3percent deficit level. This is a reasonable position on the part of the Frenchgovernment and the European Commission. Letting the automatic stabilizerswork fully at this time is a sound approach. It is also important to implementstructural reforms to increasecompetitivenessand improve the functionalityofthelabormarkettoboostgrowth.

Forcountrieswithsufficientmarginsinpublicfinances,anactivefiscalpolicyto support recovery is desirable. I would argue that the situation in my owncountry, Sweden, calls for an expansionary fiscal policy. In Sweden the debtlevelisabout40percentofGDP,publicfinancesareinstructuralbalance,andthere isahighdegreeof long-runsustainability.Butat thesametime,Swedenhasahighunemploymentrateasaresultofthecrisis.Inthesecircumstancesitseemsreasonabletocontinuetotakemeasuresonboththeexpenditureandtherevenuesidestostimulatetheeconomy.

ThesituationisdifferentinGermany.Germanyhasan80percentdebtlevel,andunemploymentisaround5percent.Ontopofthis,therearesomelong-term

issuesconcerningpublic finances. If there is fiscal space inGermany, Iwouldargue that a substantial part of this space should be kept as a buffer-a safetymargin-for two reasons.One is that theremightbeother countries in the euroarea thatwill need new programs in the short ormedium term; therefore, theanchorof thewhole euro systemmustbeon the safe side andondryground.Theremight alsobe aneed tobuild a fuller bankingunion,which couldhavefiscalimplications.AmajorpartofthefiscalspaceinGermanyshouldbeusedasasafetybuffer,eitherforuseinfutureprogramsforcountriesindifficultyoras part of a backstop arrangement, rather than to support a short-termstabilizationpolicyinGermany.However,therearealsoargumentstoconsideras to whether Germany and others can do more to increase growth. From ageneralperspective,lowertaxesforlow-incomeearnersandincreasedspendingongrowth-friendlyexpenditures,suchaseducationandinfrastructure,couldbeconsidered.

WhatHasThisCrisisTaughtUsaboutFiscalPolicy?

FiscalPolicyIsImportantforStabilizingtheEconomy

Thecrisishasshowntheimportanceoffiscalpolicysupportingmonetarypolicyin counteracting significant downturns, particularly when there are large gapsandthepolicyinterestrateapproacheszero.ThisisespeciallytrueforEuropeaneconomies with deep structural problems and ongoing private sectordeleveraging.However, theability touse fiscalpolicyeffectively isdependenton whether fiscal policy is deemed credible by the public and the financialmarkets. The long-term sustainability of public finances is therefore crucial iffiscalpolicyistoplayalargerroleinstabilizingtheeconomy.

Itisasomewhatironicconclusionthatwhilewearenowconvincedthatfiscalpolicy should be used actively, many countries have debt levels that do notpermitsuchactions.Ifyoubelievethatfiscalpolicyhasaroletoplay,itshouldbeveryclear that in the long run,weneed tobringdowndebts toa level thatenablesanactiveroleforfiscalpolicy.

Whenitcomestothetraditionalviewofstabilizationthroughfiscalpolicy,the

general conception has been thatmeasures should be temporary, targeted, andtimely. I see fiscal policy somewhat differently. Fiscal policy has three maintargets:distribution,stabilization,andgrowth.Fiscalpolicyshoulddealwithallthreeof these targetsat thesame time.Particularly inEuropeancountrieswithstructural problems, it is very important that short-term measures aimed atstabilizingtheeconomyalsosupportincreasedlong-termgrowth.Whenitcomesto discretionary measures, temporary measures are preferable in theory; inpractice, however, there are substantial risks that temporary measures willbecomepermanent,sotheyshouldbeusedwithcaution.

FiscalPolicyHasItsOwnProblems

Ifwe are going to use fiscal policy,we need to realize that fiscal policy is adifficultanimaltolivewith.Oneofthekeydifficultiesoffiscalpolicyisthatitis preceded by long decision lags, particularly regarding structural reforms.Structuralreformsareoftencomplexanddemandlongpreparation.WecantaketheSwedishbudgetfor2013asanexample.Inlightoftheweakergrowthintheeuro area and clear downside risks, together with increased internationalcompetition for the large Swedish export sector, there was a strong case forincludingstructuralmeasurestosupportgrowth.Afterthebudgetfor2012waspresented to the Riksdag (the Swedish parliament) in September 2011, theMinistry ofFinance started to consider appropriatemeasures to strengthen theeconomy. InNovember, thebasicelementsof the reformagendawerechosen.Thethreemaincomponentswereincreasedspendingoninfrastructure,increasedspending on research and development, and cutting corporate taxation andbroadening the taxbaseby reducing interest ratedeductions.The taxproposalthen had to be prepared by a government commission and was subsequentlyassessedbyourlegalcounselbeforeitcouldbepresentedtotheRiksdag.IttookuntilAprilbeforetherewasaclearviewonhowthisproposalwouldworkandwhatitwouldcost.InAugustwedecidedhowlargethetaxcutwouldbe.

Ifyouwanttocombinefiscalpolicywithstructuralmeasures,youneedtosetupanadequateplanningframework.Youneedtodecideayearaheadwhatthebasic proposals in the budget will be, and you need to work with scalablemeasuresthatcanbeadaptedtochangesintheamountofroomforreform.

Incontrast,changesinmonetarypolicyaredecidedandimplementedalmostinstantly.Thisisafundamentaldifferencebetweenfiscalandmonetarypolicy.Itisthereforeimportanttothinkthroughhowtodealwiththelongdecisionlags.Many of the political economy issues that you normally encounterwith fiscalpolicy can be circumvented and solved by setting up institutions that arewellequippedtohandlelong-termdecisionmaking.Oneimportantfeatureistoavoidautomaticindexingofexpenditures;forexample,transfersthatarefullyindexedtowages.InSweden,wehavereducedtheautomaticadjustmentofexpendituresover the past 20 years. This reduction has served us well. Fully indexedexpenditures tend to lead touncontrolledexpendituregrowthandultimately tobudget problems. Such a feature also reduces the pressure for change in thepublic sector and ties the hands of democratically elected politicians to setpriorities in the annual budget. A prudent trend for expenditures also makessense in terms of stabilization policy. If expenditures are reduced semi-automatically, it ismore credible that fiscal policywill return to a sustainablepathwhentheeconomyrecovers.

There is something deeply democratic about fiscal policy. Fiscal policy isaboutnegotiationsandreconcilingconflictsbetweendifferentprinciples. Ithasto deal with different interest groups. Interest groups represent ways ofchanneling different points of view in the debate and are a part of a well-functioning democracy. Economists have to accept that there is a need to seefiscalpolicymakingasanegotiationprocess inwhichdemocraticforcesplayacentralrole.However,itisalsoclearthatwhensettingtheframeworkforthosenegotiations,youareaffectingtheendresults.Withawell-organizednegotiationprocess and awell-organized budget process, conflicts between principles andinterestsaremucheasiertosolve.Thisleadsmetomylessonaboutcredibility.

CredibilityIsKey

Myfinallessonisthatcredibilityiskeytoasuccessfuleconomicpolicy.Iwillmake a different argument than is normally made about credibility. Modernsocietiesarebuilton trust.Citizensmustbeable to relyon the socialcontractbetween individuals and the state. The social contract is based on everyindividualinsocietybeingabletorelyonthefactthathisorhercontributionto

societywillyielda returnwhenneeded.Thismustapply inboth theshortandthelongterm.ThisisparticularlythecaseinSweden,wherefiscalpolicyhasaprofound influence on citizens' life situations, given the substantial publicprovision of welfare services. Take, for example, health care or education.Regardlessofthebusinesscycle,youasacitizenmustbeconfidentthatifyouhaveaheartattack,youwillgetgoodhealthcare,andthatyourchildrenwillgetagoodeducationwhentheygotoschool.

If there is uncertainty about the ability to handle fiscal policy in a welfarestate,citizenswillhavetrustproblems.Countriesthatincreasedeficitsanddebtwillcreatecredibilityproblemsthatriskbreakingdownthistrust.Ifwebelievethat a welfare state is more successful in creating good living conditions forpeople,itismoreimportanttodealprudentlywithdeficitanddebtissues.Inthisway, European countries differ significantly from countries with substantiallylowerlevelsofpubliclyfinancedwelfareprovision,suchastheUnitedStates.

Europe'sProblemsRunDeep

Wheredoesthisleaveusintermsoftheeurocrisis?AsIseeit,howcountriesemergefromthecrisis isnotprimarilyaboutshort-termstabilizationbutratheraboutthelong-termprospectsforgrowthandsocialcohesion.Itisimportanttoacknowledge the structural differences between the US and Europeaneconomies. These differences are probably one of the main reasons for thepolarized views on stabilization policy, both within academia and amongpolicymakers.

Ifyoutakestockof theUSeconomy, it isveryhardtoseeanyfundamentallabormarketproblems.TheUSunemploymentratewillprobablygodownto5or6percent againwithout anymajor structural reforms.Theremightbe someproblems with hysteresis in the labor market, but the fundamental factors-product market regulation, employment protection, unemployment benefits,marginal taxes, wage setting-all point in the direction of a low structuralunemploymentlevel.ThisisnottrueofmanycountriesinEurope.Thestructuralunemploymentlevelintheeuroareaisprobablyslightlybelow10percent,andtherearestrong fundamentalarguments for thishigh level. InEurope,markets

generally are much more regulated. Also, the tax rates and unemploymentbenefitsarehigher(figures15.2a-d).

Myargument isnot thatweshould transformtheEuropeaneconomies tobelike that of theUnitedStates in every respect. I tend to like the socialmarketeconomyandthewelfarestate.Europeneedstoreform,butthereformsshouldbe in linewith theEuropean socialmodel.However, it is important to realizethat under these circumstances there is a much higher risk of high structuralunemployment. Ifyouevaluatestructuralpublicfinances, it isalsomuchmorelikelythat theUnitedStateswill loweritsdeficit level,andthereforedealwithitsdebtproblemautomaticallywhen the economy is reinvigorated.This isnotthecaseinEurope.

Given the problems of high indebtedness, low long-run growth, and highstructuralunemployment,thebulkoffiscalmeasuresinEurope,especiallyintheeuroarea,needtoaddresstheseproblems.Weneedtoconfrontthefundamentalstructural problems in the euro area, first and foremost the functioning of thelabormarket.Ifwearegoingtousefiscalpolicy,it'simportanttokeepinmindthat fiscal policy has goals for distribution, growth, and stabilization. Whendealingwith stabilization issues,weshouldalso try todo things that stimulategrowth in the long term. Rather than being temporary and timely, measuresshould be directed toward confrontingmore fundamental issues thatwill trulydetermine long-term growth prospects and the scope of our European socialmodel.So,weshouldusefiscalpolicy,butitmustbeaconstrainedactivismthatinspires credibility not only among markets but also in terms of the trust ofsociety.

Figure15.2

DifferencesbetweentheUS,Japanese,andEuropeanEconomies

a. Product market regulation. Shown is the OECD overall product marketregulation indicator for2008on an index scaleof0 to6,where6 is themostrestrictive.

b. Strictness of employment protection. Shown is the OECD strictness ofemploymentprotectionindicatorfor2008onanindexscaleof0to6,where6isthe most restrictive. c. Marginal tax rate. Marginal tax wedge as figured forprincipalearner(in%).d.Structuralunemploymentrate,percentoflaborforce,2012.

Note: Euro area refers to a weighted average of GDP weights (2011) andrepresents15OECDcountries.

Sources:aandb.OECDandowncalculations. c.OECD.d.OECDEconomicOutlook2012.

Conclusion

• Fiscal policy should support monetary policy, particularly in dealing withspecificstructuralanddistributionalproblems.Butfiscalpolicyhasitsownsetofproblems,particularlylongdecisionlagsinherentinstructuralreforms.

• Long-term sustainable public finances are a fundamental condition for

conducting fiscal policy and maintaining confidence, especially for countrieswithhighwelfareambitionsandlesseffectivelaborandproductmarkets.

• Europe must maintain its focus on structural reforms and tackle thefundamentalissuesthataffectlong-termconditionsforgrowth,increasedsocialcohesion,andincreasedconfidence.

Note

I thankPhilipLof for his help in drafting the speechonwhich this chapter isbased.

RobertoPerotti

Ideally, a governmentwould like to use fiscal policy as a countercyclical toolwhileatthesametimeconvincingthemarketsthatitremainssolventand/oritsgovernment debt remains sustainable.' What it can do depends on the initialsituation.InthischapterIconsiderapolicymakerthathassome"fiscalspace."In other words, there is room for some, perhaps temporary, increase in thedeficit.Inmoremundaneterms,the"debtvigilantes"arenotactive.Admittedly,these are loose terms; Iwill not try to define them formally, and even less todetermineathresholdforwhen"fiscalspace"isavailable.

Iwillexplorethreepopularoptions;theemphasiswillbeonpolicyissues,butIwillalsodealwiththeensuingchallengestomacroeconomicthinking.

Option1: Ignore the availability of fiscal space, cut government spending andtaxes. Then GDP, private consumption, and private investment will startgrowing,andthiswilltakecareofanyissueswithsustainabilityandsolvency.

This is the notion of the "expansionary effects of fiscal consolidations."Obviously, itsproponentshave rarelyargued that spendingcutsby themselveswill spur growth; accompanying structural reforms of labor, goods, and creditmarketsarealsonecessary.

Whataretheempiricalfoundationsofthisview?Afirstmethodtoinvestigatethe effects of a fiscal consolidation is simply to extrapolate from currentestimatesoftaxandspendingmultipliers,fromtime-seriesvectorautoregressionstudies. There are several recent surveys of existing results, mostly based onU.S.data,andestimatesoffiscalmultiplierswereextensivelydebatedinthefirstconferenceonrethinkingmacropolicy.Therefore,Iwillconcentrateonstudiesthatfocusexplicitlyonepisodesoffiscalconsolidation.

A method that has been widely used to study the effects of fiscal

consolidationsconsistsofasimplecomparisonofmeansofvariablesovertime.Specifically: (1)definea "fiscal consolidation," for instance, as a country-yearwhen the discretionary2 decline in the primary deficit is more than, say, 1.5percentofGDP,orastwoconsecutivecountry-yearswhenitisatleast1percenteach year; then (2) take amacroeconomic variable of interest, such as privateconsumption,andcomparetheaverageofthatvariableinthetwoyearsafter(orduring) the consolidation with the average in the two years before theconsolidation. This "mean comparison" approach would provide unbiasedestimatesof the average effectsof consolidations if the latterwere completelyrandom events (in which case it is essentially a difference-in-differenceestimator).

This is themethodology applied byAlesina and Perotti (1995) andAlesinaandArdagna(2010)withcyclicallyadjusteddata,andbyAlesinaandArdagna(2012)withthenarrativeIMFdataofDevriesandothers(2011).3

The typical result is that spending-based consolidations (where thediscretionarydeclineinthedeficitconsistsofatleast50percentspendingcuts)tendtobelonger-lastingandareassociatedwithanincreaseinGDPgrowthorasmall recession, while tax-based consolidations are short-lived and areassociatedwithaslowdowningrowthorevenarecession.Withsomevariations,allofprivateconsumption,investment,andexportsdisplaythispattern.Also,ingeneral,thesevariablesareparticularlyresponsivetocutsinsocialspendingorspending on public wages and salaries-the two largest government spendingitemsinallOrganisationforEconomicCo-operationandDevelopment(OECD)countries.

Fiscal consolidations are typicallymultiyear events. In thismethod, a fiscalconsolidation lasting four years would appear as three consecutive two-yearconsolidations;moreover,agivenyearcanappearinallof the"pre,""during,"and "post" groups at different dates. It is not clearwhat themean comparisonmethoddeliversinthesecases.

A second problem with this approach is that it is difficult to control forconcomitant effects. For instance, one typical result is that spending-based

consolidations are associated with real depreciation of the exchange rate andimprovementinrelativeunitlaborcosts.Isthisaconsequenceofspending-basedconsolidations or the result of policies typically implemented together withspending-basedconsolidations?Asalways,causalityisdifficulttoascertain.

Theaccompanyingpoliciesmighttakeseveralformsthatmightbedifficulttocapturewithoneortwovariables:consider,forinstance,labormarketreforms,orchangesinexchangerateormonetarypolicyregimes.Finally,thegovernmentbudgetsandaccompanying technicaldocumentsneed tobestudied indepth inordertodeterminethediscretionarymeasureswithaminimumofconfidence.

Forall thesereasons,it isusefultocomplementtheexistingevidencewithadifferent approach. Perotti (2013a) presents a detailed discussion of the fourlargestspending-basedconsolidations-Denmarkin1983to1987,Irelandin1987to 1989, Finland in 1992 to 1996, and Sweden in 1993 to 1997-based on theoriginalbudgetdocuments andon contemporarydiscussion, such asOECDorIMF annual reports, and countryspecific sources.4 I focus on two questions.First, is thereevidence that largebudgetconsolidationscanhaveexpansionaryeffects in the short run?Second,howuseful is the experienceof thepast as aguidetotoday'seuroareacountries?

ThemainconclusionsofthecasestudiesIpresentarethefollowing:

1.Actualconsolidationsweresmallerthanpreviouslythought,andnotspending-based.

Alltheseconsolidationshavelongbeenconsideredquintessentialcasesoflarge,"spending-basedconsolidations."Twooftheseweretrulyenormous:Asshownintable16.1inthecolumnlabeled"IMF,"inFinlandthediscretionaryprimarydeficit fell by 11.5 percent of GDP over 5 years (all of them spending cuts),according to the IMF narrativemeasure of Devries and others (2011), and inSweden by 10.6 percent of GDP over 5 years (of which almost 7 percentagepointswereGDPspendingcuts).

But a closer look at the budget documents shows that in many cases theannounced spending cutswere not implemented, or else theywere undone by

supplementary budgets during the fiscal years. As a result, the decline in thebudgetdeficitinFinlandwasonlyabout5percentagepointsofGDP;moreover,nearly all of it consisted of revenue increases (see table 16.1, column labeled"Actual"). My estimate of spending cuts in Sweden is about half the IMFestimate. Only in Ireland were spending cuts larger than revenue increases.sThese conclusions are corroborated by contemporary policy documents anddiscussions,whichdonotshowanyconsciousnessof living througha"budgetbloodbath."6

Table16.1

LargeFiscalConsolidationsinEurope

2.Depreciationandtheroleoftheexchangerateandmonetaryregimes.

Tables16.2and16.3displaythebehaviorofthemultilateralnominalexchangerate andofmultilateral unit labor costs inmanufacturing, respectively.During

the consolidations,Denmark and Ireland used the exchange rate as a nominalanchor, by committing to a hard peg within the European Exchange RateMechanism (ERM).Denmark had repeatedly devalued its currency before theconsolidation,thusenteringtheconsolidationwithadepreciatedexchangerate,but at the cost of very high interest rates (up to 23 percent). During theconsolidation phase, Ireland also benefited from the large appreciation of thecurrencyofitsmaintradingpartner,theUK,whichwasnotpartoftheERM.

Table16.2

NominalEffectiveExchangeRate

Table16.3

RelativeUnitLaborCostsinManufacturing

FinlandandSwedenalsodevaluedrepeatedly,butthenfloatedtheircurrenciesjustbeforetheconsolidation,experiencingafurtherdepreciation.Overall, their

currenciesdepreciatedbybetween15and25percentinmultilateraltermsinthefirst twoyearsof theconsolidation.Asanominalanchor,FinlandandSwedenintroducedinsteadinflationtargeting.Thereissomeevidencethat,whilealmostcompletely new at the time, this approach was regarded as credible from thestart:According to insiders'accountsat the time, it"hadaprofound impactonthe behavior of labormarket participants" (Jonung,Kiander, andVartia 2008,37).

3.Incomepolicieswerekey.

Fiscal consolidations were accompanied by explicit income policies, wherebythe government, trade unions, and industrialists' organizations reached anagreement to exchange wage moderation for lower income taxes and socialsecurity contributions. Ireland returned to a tripartitewage settlement in 1987(seetable16.3),whichsetamaximumincreaseinwagesof2.5percentin1988,1989, and 1990. Finland andSweden signed tripartitewage agreements at thestartof theconsolidations,and then,after somewageslippage threeyears intothe consolidation (see table 16.3), the government summoned the unions andindustrialists' associations again to sign other wage agreements. Thesedevelopmentswereregardedasverysignificantbycontemporaries:AsJonung,Kiander,andVartia(2008,35)write,basedoncontemporaryaccounts,"Perhapsthebiggestchangeinthe1990sinFinlandwastheadoptionandwideacceptanceofapolicyoflongtermwagemoderation."

Incomepolicieswereparticularlyexplicit inDenmark.Here thegovernmentrenouncedanydepreciationoftheexchangerateandreliedinsteadonaninternaldevaluation: It suspendedwage indexation, cappedcontractualwage increases,and froze unemployment subsidies and transfers, all in exchange for lowerincometaxesandsocialsecuritycontributions.

Wage moderation, which was made possible by income policies, wasinstrumental in maintaining the benefits of the nominal depreciations and inreducinginflationexpectationsandinterestrates.

4.Recoveriesweremostlyexport-driven.

AllstabilizationswereassociatedwithlargeincreasesinGDPgrowth,typicallyatabout4percentforafewyears(table16.4).

Thesourceoftherecoveryiscrucialintryingtoshedlightonthemechanism.Most models posit that a fiscal consolidation raises consumers' and investors'confidencevia awealth effect or other channels, and therefore should cause aquick increase in private consumption and investment. However, except inDenmark (where the recovery was already under way at the time of theconsolidation), private consumption typically started recovering six to eightquarters after the beginning of the consolidation; in Sweden, in the first year,domesticdemandcollapsed,with investment fallingby15percent (tables16.5and16.6).Initially,andagainwiththeexceptionofDenmark,therecoverywasexport-driven(table16.7):InFinland,Ireland,andSweden,exportsincreasedatratesabout10percentperannumforseveralyearsaftertheconsolidation.

Table16.4

GDPGrowth

Table16.5

PrivateConsumptionGrowth

Table16.6

PrivateInvestmentGrowth

Table16.7

ExportGrowth

Table16.8

Long-TermInterestRates

Thiswasmadepossibleby thecombinationofnominaldepreciationandwagemoderation:Between1992and1995,multilateralunitlaborcostsinSwedenfellbyalmost40percent(seetable16.3).

Denmark, which alone pursued a hard peg policy, experienced all thehallmarksofthe"exchangeratebasedstabilizations"studiedinalargeliteratureintheeightiesandnineties(see,e.g.,Ades,Kiguel,andLiviatan1993):domesticdemand initiallyboomedas inflation and interest rates fell fast, but as incomepoliciesbythemselvesproveduntenableafterabouttwoyears,competitivenessand the current account worsened; eventually, growth ground to a halt andconsumption declined for three years. The slump lasted for several years (seetable16.4).

5.Highanddeclininginterestrates.

Inallcountriestheconsolidationswereaccompaniedbylargeandfastdeclinesinnominalinterestratesfromveryhighlevels.InDenmarkthe10-yearinterestratefellfrom21percentin1982to11percentin1987,inFinlanditfellfrom12percentin1992to7percentin1996,andtheratefellbysimilaramountsintheother twocountries (table16.8).Thiswasmadepossiblebywagemoderation,whichinturnmadethenominalanchorscredible.

InDenmarkandIreland,thedeclinesininterestratesledtolargeincreasesinhouse prices and possibly a large wealth effect on households. According toGiavazziandPagano(1990),thiswaslargelyresponsibleforthespurinprivate

consumption.

What are the key policy conclusions we can derive from these episodes?Causalityisdifficulttoestablishineconomics,allthemoresofromafewcasestudies.Butafewpatternscanbedetectedthatcouldprovideusefulinsights-andcaveats-forthecurrentsituation,particularlyofeuroareacountries.

1.Theoft-citedexpansionaryconsolidationsofEuropein the1980sand1990sweresmallerandlessspending-basedthanpreviouslythought.

2.Allstartedatveryhigh levelsofnominalandreal interest rates,which thendeclined quickly. Interest rates are at historical lows today except in thosecountriesontheperiphery,wheretheyincludeadefaultriskpremium.

3.WagemoderationwasthekeytoacrediblepeginDenmarkandIreland,andtomaintaining the benefits of devaluations in Finland and Sweden. Butwageinflation is hardly a problem in today's low-inflation scenario. In addition,incomepolicieswereinturninstrumentalinachievingwagemoderation.Butforpolitical and perhaps cultural reasons, income policies are not on the agendatoday.

4.ExceptinDenmark,exportsweretheprimefactoroftherecoveryforseveralquarters, and thereafter kept growing at a sustained pace for several years;domesticdemandinitiallystalledorevenfell.Allcountries(includingDenmarkandIreland,whichpeggedtheexchangerateduringtheconsolidation-themorerelevant case for today's euro area members) devalued repeatedly before theconsolidations. This option is obviously not available to euro area membersexcept vis-a-vis non-euroarea members. Ireland also benefited from theappreciationof the currencyof itsmain tradingpartner, theUK.On the otherhand, the Danish expansion was short-lived, as it quickly ran into a loss ofcompetitivenessthathamperedgrowthforseveralyears.

The observations above suggest that the notion of "expansionary fiscalausterity"intheshortrunisprobablyanillusion:Atrade-offdoesseemtoexistbetweenfiscalausterityandshort-rungrowth.

Howdoes this evidence comparewith the econometric evidence from time-seriesstudies?Contrarytowidespreadopinion,thereisverylittledisagreementon the fact that a positive shock to government spending causes totalGDP torise.Thedisagreement concerns the effect onprivate consumption andprivateinvestment, and on privateGDP. I do notwant to revisit this debate here butwouldliketopointout that theevidencethatgovernmentspendingcrowdsoutprivate GDP comes from shocks to defense spending or to total governmentspendingwhendefensespendingshocksdominate(e.g.,duringWorldWarII).TheresponseofprivateGDPanditscomponentstocivilianspendingshocksispositiveandlarge(figure16.1).

Figure16.1

Effects of Different Government Spending Shocks, Government SpendingversusPrivateGDP

Source:Perotti(2013b).

Thereisoneimportantcaveattoallthis,however:Weknowverylittle,indeednext to nothing, about multipliers on government transfers, as opposed togovernmentspendingongoodsandservices.

Option2:Expandgovernmentspendingtemporarily.

Thesecondoptionforagovernmentwithsomefiscalspacewishingtousefiscalpolicy as an antirecessionary tool is to increase government spending, andconvincemarketsthatitwillcutbackonspendinglater.Theadvantagesofthis

policy are clear. It reconciles solvency and sustainability with the use ofcountercyclical fiscal policy; and it takes advantage of the higher spendingmultipliers at the zero lower bound (ZLB) or, more generally, in a recession(moreonthislater).

This option has often been proposed as particularly suited for the currentsituation. But one has to be careful. Multiyear budget plans are dangerousobjects:Forpoliticians,thereisanobviousincentivetoannouncespendingcutsfor the future,andwhen the futurecomes, topostpone them.Thiscanbeseenclearly in the Clinton budgets (eventually the problem was solved by theeconomic recovery); and also in the experience of Finland and especiallySweden in the early 1990s, described above, when, at the beginning, a newgovernmentwouldannounceplansforlargespendingcutsinthefollowingfiveyears,whichcuts,however,materializedonlyinaverysmallpart.7

Twoimportantstipulationsoftenaccompanythisproposal:budgettargetsandcommitments shouldbe set in cyclically adjusted terms, anda "fiscal council"should be set up to monitor the implementation of the plans. In theory, bothmake a lot of sense. In practice, however, we should not forget that cyclicaladjustment is an artmore than a science; the dispersion of cyclically adjustednumbersonthesameaggregatebydifferentinstitutionsatagivenpointintimeand by the same institution at different points in time should be an importantwarning.Becauseofthisuncertainty,cyclicallyadjustedfigureswillbepronetopolitical manipulation and to endless bickering, and in the end may be moreconfusingtomarketsthantherawfigures.

Fiscalcouncils,too,makealotofsenseintheory.Theparallelisoftendrawnwiththemovetomakecentralbanksindependentinthepastthreedecades.Butthereisonefundamentaldifference.Centralbankscontrolveryfewinstruments(and just one until a few years ago, the policy interest rate), with limiteddistributional effects-or at least with distributional effects that are notimmediately apparent. Fiscal policy consists of amyriad of instruments, withenormous distributional implications. No government and no parliament willever relinquish it completely. This is not to say that fiscal councilsmight nothave a role, but it is hard to escape the conclusion that they will work in

countrieswhere theculturalandpoliticalenvironment isalreadywelldisposedtoaresponsiblefiscalpolicy.

Thepolicyofincreasinggovernmentspendingtotakeadvantageoftheextrakick from the ZLB is gaining momentum in Europe, as a backlash against"German austerity" also gains strength everywhere. Academically, it issupported by some empirical evidence to the effect that government spendingmight have a highermultiplier during recessions (see, e.g., Eggertsson 2010).However, this is only indirect evidence concerning the ZLB itself, as not allrecessionscoincidedwithinterestratesclosetozero.Inaddition,themechanismatwork is not clear. Theoretically, fiscal policy ismore powerful at the ZLBbecause, in a stickypricemodel, apersistent increase ingovernment spendingcausesariseinexpectedinflationandthereforeadeclineintherealinterestrateifthenominalinterestrateisfixedatzero;fromtheEulerequation,thiscausesprivateconsumptiontoincreasetoday.However,atfirstsight,theexperienceofthose countries that did increase government spending with very low interestratesatthebeginningofthecrisis,orofJapaninthe1990s,seemshardtosquarewiththismechanism.

Inaddition,ahighermultiplierdoesnotmeannecessarilyhigherwelfare.Inarepresentativeagentmodel,itiseasytoseethatwelfarecoulddeclineevenifthemultiplierattheZLBislarge,andinanonrepresentativeagentmodelingeneralsaverswillbeharmedbythispolicy(seeBilbiie,Monacelli,andPerotti2013).

Option3:Cuttaxestemporarily.

The results based on tax data from Romer and Romer suggest that the taxmultipliercanbeverylarge(RomerandRomer2010;MertensandRavn2013),althoughprobablynotaslargeasintheoriginalRomerandRomerstudy(Perotti2012).AnotheradvantageisthatthispolicymightbeParetoimproving,whereasapolicyofincreasinggovernmentspendingisnot.Althoughsaversarehurtbyan increase in government spending, even at theZLB, theymight still benefitfromatemporarytaxcut.Whenthereareliquidityconstrainedindividuals,ataxcut is likea transferfromsavers toborrowerstoday,offsetbyatransfer in theopposite direction tomorrow.Both borrowers and savers can benefit from this

operation(seeBilbiie,Monacelli,andPerotti2013).

However,intherealworld,ataxcuthasonedisadvantage:itmighthavebaddistributional effects, as it does not benefit the low end of the distribution ofincome.Thissuggests twoareas thatweneed to learnmoreabout.Thefirst isthegeneralequilibriumeffectofalternativetypesoftaxchanges,withdifferentdistributional effects. The empirical evidence we have from the Romer andRomerdataconcernsaggregatechanges in taxes(MertensandRavn[2013]dostudy the different effects of personal and corporate income taxes). Moregenerally, we need to understand better how to design economically andpolitically feasible policies to protect the low end of the distribution duringperiodsoffiscaladjustments.

Notes

1. A government is solvent when future taxes are enough to cover futurespendingplustheexistingdebt.Agovernmentthatisnotcurrentlyperceivedassolventmustthereforereducedeficitsnoworlater(oruseinflationtoreducethecurrentvalueofdebt).Thecurrent levelofdebt issustainableif,at thecurrentprimary deficit, GDP growth is sufficient to keep the current debt/GDP ratioconstant. A government whose debt is not perceived as sustainable mustengineerhighergrowthoralowerinterestrateifitdoesnotwanttoreducetheprimarydeficit.

2.The"discretionary"changeinthedeficitisthepartofthechangeinthedeficitthatisnotduetotheautomaticresponseofthedeficittotheeconomiccycle.Inthis sense, it can be interpreted as the part of the change in the deficit due tointentionalactionsbypolicymakers,suchaschangesintaxrates,inreplacementrate for unemployment benefits, in defense spending,etc. The same definitionappliestoeachindividualbudgetcomponent.

3.There are twomethods to obtain "discretionary"measures of a change in abudgetvariable.First,the"cyclicaladjustment"method:Estimatetheelasticitiesofthatbudgetvariableto,say,outputandinflation,andsubtractfromtheactualchange in the budget variable the change in output multiplied by the outputelasticity and the change in inflation multiplied by the inflation elasticity.Second, the "narrative" method, pioneered by Romer and Romer (2010) for

revenuechanges:Usebudgetdocumentstoinferthediscretionarychangeintaxrevenuesorspendingenactedbyanylawthathasconsequencesforthebudget.Devriesandothers(2011)computeyearlydiscretionarychangesingovernmentspendingandrevenuesduringperiodsofdeficitreductionsin15countries.

4. The pros and cons of case studies vs. an econometric approach are well-known.Hence,Iwillnotrevisitthisdebatehere.

5. All this still excludes bank support measures. For instance, in Finland thegovernmentspentabout10percentofGDPtosupportthebankingsector.

6.Forinstance,anofficialintheIrishadministrationatthetimelaterwrote:

Briefly,therewasnosignificantreductionintherealvolumeofcurrentspendingasa resultof [theexpenditurereviewsetupby thenewgovernment in19871.There was a further squeeze on capital spending, amistake in retrospect, butmostof theadjustmentcameon therevenueside.The"slashandburn"storiesabout1987,referencestothefinanceministeras"MactheKnife,"decimationofpublic services and so forth are just journalistic invention. It never happened.(McCarthy2010,45)

7.Mauro(2011)attributesthisundoingoftheoriginalplansmostlytoshocks.Ibelievethatinmanycases,itisjustasimpleissueoftimeinconsistency.

References

Ades,A.,M.Kiguel,andN.Liviatan.1993."Exchange-Rate-BasedStabilization:Tales fromEuropeandLatinAmerica."PolicyResearchWorkingPaper1087,WorldBank,Washington,DC.

Alesina,A.,andS.Ardagna.2010.LargeChangesinFiscalPolicy:TaxesversusSpending. In Tax Policy and the Economy, vol. 24, ed. Jeffrey R.Brown.Cambridge,MA:NationalBureauofEconomicResearch.

Alesina,A.,andS.Ardagna.2012."TheDesignofFiscalAdjustments."Facultypaper,BocconiUniversity,Milan.

Alesina,A.,andR.Perotti.1995."FiscalExpansionsandAdjustmentsinOECDEconomies."EconomicPolicy21:207-247.

Bilbiie, E, T.Monacelli, and R.Perotti. 2013. "Tax Cuts vs. GovernmentSpending: Welfare at the Zero Lower Bound." Faculty paper, BocconiUniversity,Milan.

Devries,P.,J.Guajardo,D.Leigh,andA.Pescatori.2011."ANewAction-BasedDataset of Fiscal Consolidation." IMF Working Paper 11/128, InternationalMonetary Fund, Washington, DC. Data set available at www.imf.org/externa]/pubs/cat/longres.aspx?sk=24892.0.

Eggertsson,G.2010."WhatFiscalPolicyIsEffectiveatZeroInterestRates?"InNBERMacroeconomicsAnnual 2010, vol. 25, 59-112.Cambridge,MA:MITPress.

Giavazzi, E, and M.Pagano. 1990. "Can Severe Fiscal Contractions BeExpansionary? Tales of Two Small European Countries." In NBERMacroeconomicsAnnual1990,vol.5,75-122.Cambridge,MA:MITPress.

Jonung,L.,J.Kiander,andP.Vartia.2008."TheGreatFinancialCrisisinFinlandand Sweden: The Dynamics of Boom, Bust, and Recovery, 1985-2000."Economic Papers 350, Directorate-General, Economic and Financial Affairs,European Commission, Brussels, December. http://ec.europa.eu/economy_finance/publications/publication13551_en.pdf.

Mauro,P.,ed.2011.ChippingAwayatPublicDebt.NewYork:Wiley.

McCarthy,C.2010. "Ireland'sSecondRoundofCuts:AComparisonwith theLastTime." InDealingwithDebt:LessonsfromAbroad,ed.J.Springford,41-54.CentreForum(CentreforReform)Canada,Ernst&Young.

Mertens, K., and M.Ravn. 2013. "The Dynamic Effects of Personal andCorporate Income Tax changes in the United States." American EconomicReview103(4):1212-1247.

Perotti,R.2012."TheEffectsofTaxShocksonOutput:NotSoLarge,butNotSmallEither."AmericanEconomicJournal:EconomicPolicy42:214-237.

Perotti,R.2013a."TheAusterityMyth:GrowthwithoutPain?"InFiscalPolicyaftertheGreatRecession,ed.A.AlesinaandEGiavazzi.Chicago:UniversityofChicagoPress.

Perotti, R. 2013b. "DefenseGovernment Spending Is Contractionary; CivilianGovernment Spending Is Expansionary." Faculty paper, Bocconi University,Milan.

Romer,C.,andD.Romer.2010."TheMacroeconomicEffectsofTaxChanges:Estimates Based on a New Measure of Fiscal Shocks." American EconomicReview100(3):763-801.

NourielRoubini

The recent global financial crisis has brought the attention of analysts andpolicymakersbacktotheroleoffiscalpolicyduringthecrisisanditsaftermath.Several important questions need to be addressed: What is the relationshipbetweenlevelsofpublicdebtandeconomicgrowth?Whatarethecausesofhighdebtanddeficits-loosefiscalpolicyorweakeconomicgrowth?Whatisthesizeof fiscalmultipliers, andhowdo theydependonbusinesscycleconditions? Isthere a risk of fiscal dominance? How can we reduce a debt overhang andachieveasmootherdeleveragingfromhighdebtratios?Whatistheoptimalpaceoffiscalconsolidation?Thequestionoffiscalpolicyisacriticalone,soIwilltotrytoanswerthesequestions.

Thefirstquestiononeneedstoaddressconcernstherelationshipbetweenhighpublic debt and economic growth, and the costs of that high debt. Economictheorysuggeststhatatsomepoint,highpublicdebtcanhaveanegativeeffectoneconomicgrowth.Itcanleadtohighrealinterestratesandcrowdoutinvestmentandconsumption.Itcan increase theriskofadebtcrisiswithall thecollateraldamageofadebtdefaultoccurring.Itcanforcepolicymakerstoincreasetaxesto avoid a debt crisis, but high taxes cause distortions that negatively affecteconomicgrowth.

Sothosearethefactorsthatcanleadtolowergrowthinthepresenceofhighand rising public debt levels. Recent research work by a number of scholars,especially CarmenReinhart andKennethRogoff, (Reinhart andRogoff 2010)suggests that there could be a significant empirical relationship between highpublic debt and lower economic growth. According to them, the criticalthreshold in advanced economies occurred when debt was greater than 90percentofGDP(lowerinemergingmarkets).Thisiswhenthenegativeeffectson economic growth become significant. Indeed, the median debt ratio foradvancedeconomieshasgonefromabout60percentofGDPbeforethecrisisto

alevelcloserto100percenttoday.

When one considers Reinhart and Rogoff's empirical results, there are twoimportantthingstokeepinmind.Oneis,ofcourse,thecausalityissue.Isitthehigh debt ratios that cause lower economic growth, or shocks that lead tosignificantrecessions,orfinancialcrisesthatcauseeconomicweaknessandthusleadtoanincreasingdebtratio?Sothefirstquestionisthecausalityone.

Recent academic work (Herndon, Ash, and Pollin 2013) has shown someserious methodological problems with Reinhart and Rogoff's results: Therelationshipbetweenhighdebtandeconomicgrowthmightnotbeasrobustastheythoughtitwas.Thus,therecentconventionalwisdomthathighdebtleadstolowergrowthrateshasbeenseriouslychallengedempirically.

AsOlivierBlanchard,GiovanniDell'Ariccia,andPaoloMaurosuggestintheintroductiontothisbook,anotherbigriskandpotentialeffectofhighdebtistheriskofmultipleequilibria.Ifyouhavealotofpublicdebtandithastoberolledover,thereisariskthatyoumayendupinasituationinwhichthereisaself-fulfilling run on public debt. This will make the spreads higher andunsustainable,andyoumayendup ina situation inwhicha liquidityproblemleadstoinsolvency.Anilliquidbutsolventsovereignmightendupindefaultifsucharundoesoccur;thisisabadequilibriumthatyoucannotruleout.Allelseequal, having a lower debt ratio and less of a liquidity risk with a longer-maturitydebtcanreducetheriskofsuchabadequilibrium.

Ofcourse, there isasolution toa liquiditycrisis.Weknowit in thecaseofbankruns,butthesamethingoccursinthecaseofarunongovernmentdebt.Ifyou have a lender of last resort-a central bank that can provide liquidity to asovereignbymonetizing itsdebt-youcanavoid thatbad runequilibrium.Thiswas certainly the situation with the euro area in the summer of 2012, wheninterestratesonItaliandebtrosetoalmost7percentandthoseonSpain'sdebtwere closer to 8 percent. When Mario Draghi gave his "whatever it takes"speechandtheEuropeanCentralBank(ECB)announceditsOutrightMonetaryTransactions(OMT)program,asignificantreductioninthoseyieldsandspreadsoccurredasboththecurrencyredenominationriskofaeuroareabreakupandthe

riskofarunonpublicdebtweresharplyreduced.Thissuggestsanelementofself-fulfilling bad equilibrium in the case ofSpain and Italy in the summer of2012,whenthespreadkeptonrisingtounsustainablelevels.Theexistenceofalenderoflastresortcanhelpavoidthosebadequilibria.

Similarly,areratesintheUnitedStatesandJapanlowbecausethesemarketsare safehavensduringperiodswhen tail risks arehighand risk isoff?Or arethose lowrates the resultof large-scalequantitativeeasing that iseffectivelyaformofdebtmonetizationthatreducestheriskofrunsonpublicdebtandkeepslongrateslowerthantheyotherwisewouldbe?

Of course, one needs then to address the moral hazard problems that suchintervention/insuranceagainstliquidityrunsonpublicdebtmayinduce.

Thesecondquestionthatisworthdiscussingconcernsthecausesofhighdebtproblems. The policy answer on what we should do about high debt ratiosdependsinpartonwhatleddebttoincreaseandreachsuchahighlevel.Whenonelooksattheexperienceofthelastfewyears,ofcourseonefindsexamplesofcountries inwhich fiscal policywas loose and reckless. The obvious examplemight be Greece, which was running very large and unsustainable budgetdeficitsuntil theonsetofitsdebtcrisis in2010.Policymakersthereeffectivelycheated and lied about the true size of the deficit, which turned out to be 15percentofGDP,muchhigherthanpreviouslyannounced.

Of course, if you run very large budget deficits for reasons that have to dowith political distortions that lead to an increase of debt that becomesunsustainable,youare ina typical situationwhereyou'regoing tohaveadebtcrisis owing to reckless fiscal policy. And those high debt ratios will havesharplynegativeeffectsoneconomicgrowth.

However,theexperienceofthepastdecadesuggeststhatmanyofthefinancialcrisesledtoalargeincreaseinpublicdebtanddeficitsthatstartedwithprivate-sector, not public-sector, financial excesses. These were credit, housing, andasset bubbles of one sort or another that eventually burst, andonce theyburstthey caused a significant increase in budget deficits and public debt. These

increases were first driven by the ensuing recession that triggered automaticstabilizers.Thus,revenuesfellsharplyovertimewhilespendingrose.

Second,wheneverafinancialcrisisdoesoccur-suchastheonein2008-2009-there's also the risk that a great recession might turn into another GreatDepression.Therefore, theoptimal policy response to avoid such adepressionwasa fiscal stimulus-mostnecessary ina situation inwhichprivatedemand iscollapsing.Ifprivateconsumptionandinvestmentareinfreefallandyoudon'thavealargefiscalstimulus(anincreaseinspendingorareductionintaxesoracombinationofthetwo),aneconomycouldfallintoadepression.Infact,oneofthe lessons of theGreatDepressionwas that you need a fiscal stimuluswhenprivatedemandiscollapsing.

Third, the fiscal cost of cleaning up, bailing out, and backstopping thefinancial system, or even corporations (the GM and Chrysler bailouts) orhouseholds, implies therewill be large costs. In these financial crises, lots ofcontingent liabilities will emerge, and that will be the source of additionalincreasesinpublicdebt.

Again, the experiences over the last few years, perhaps especially theexamplesofIreland,Iceland,Spain,theUnitedKingdom,andtheUnitedStates-and even emerging markets such as Dubai-were all essentially private-sector-induced excesses that led to bubbles. Those excesses eventually led to a bust,andtheresultingincreaseinpublicdebtanddeficitwastheresultofthatseverefinancialcrisis.Thisimpliesthatperhapsthepolicyresponsetoabalancesheetcrisis might be different from the response to a situation in which there wasrecklessfiscalbehaviorinthefirstplace.Inbalancesheetcrises-wherethebustleads the private sector to sharply deleverage by cutting spending onconsumptionand investment-a large fiscal stimulus isnecessary toprevent thecollapseofprivatedemandfromcausinganevenmoresevererecession.Thus,fiscal stimulus (along with aggressive monetary easing) rather than fiscalcontractionistheappropriatepolicyresponsetoaprivate-sector-inducedbalancesheet crisis. This is the policy that economists such as Richard Koo havecorrectlyidentifiedasappropriatefollowingbalancesheetcrises.

Thethirdquestiontodiscusshastodowiththesizeoffiscalmultipliers.Doesa fiscal expansion increaseGDP?How large are those fiscalmultipliers?Aretheygreaterthan1ornot?Therelatedquestioniswhetherfiscalconsolidationisexpansionary.Thereisapopularhypothesisthatafiscalconsolidationwillhaveapositiveconfidenceeffect(a"confidencefairy")oneconomicgrowth;thatis,thatreducingthefiscaldeficitwillincreaseeconomicactivityevenintheshortrun.Isthatviewcorrect?

Ifone looksat theempiricalevidence, four resultsemerge.First,asRobertoPerottishowsinchapter16,thereisnorealevidencethatfiscalconsolidationisexpansionary in the short run; rather, it tends to have negative effects oneconomic activity. This is the case even in the euro area, where fiscalconsolidationmaybeneededovertimetoavoidadebtcrisis.Infact,thefront-loadingof fiscal austerity in the euro areaperiphery is oneof the reasons thisregionfellintoadouble-diprecessionin2011-2012.

Second,theworkthattheIMFhasdoneisconsistentwiththeviewthatfiscalausterity iscontractionary,at least in theshort run. Ifyou raise taxesand thusreduce disposable income, or if you cut government spending, evenunproductive government spending, you are reducing aggregate demand.Therefore, reducing disposable income and aggregate demand will have anegative effect on economic activity in the short run. Also, when you havesynchronized fiscal austerity inmany parts of theworld, the fiscalmultiplierswillbelarger.Indeed,until2012,fiscalausteritywaslimitedtotheperipheryofthe euro area and the UK. But in 2013 even the United States will have asignificantfiscaldragand,giventheEuropeanFiscalCompact,eventhecoreoftheeuroarea (Germanyandothers)will implement fiscal austerity.Then, inasituationinwhichmanycountriesarepracticingausterityatthesametime,thosefiscalmultiplierscouldendupbeingactuallylargerthanwhenfiscalausterityislesssynchronizedglobally.

Third, thereareabout adozeneconometric/statistical studiesabout the2009USfiscalstimulus.MostofthesestudiesreportthatitwasexpansionaryonGDPandthattheresultswerelargeandsignificant.

Fourth,thereisincreasingevidencethatthefiscalmultipliersarelargerwhenyouareboundbyazerointerestratepolicyandwhenthereisameaningfulslackin the economy, that is, when you are in recession or you are growing veryslowly.

So, a fair reading of the empirical evidence suggests that fiscal stimulus iseffective in stimulating growth, especially after a financial crisis, when theeconomy has a large slack and is in a liquidity trap; conversely, excessivelyfront-loadedfiscalconsolidationhasanegativeeffectongrowth.

The empirical evidence and the conceptual arguments about how toappropriately respond to a balance sheet crisis with a fiscal stimulus explainsome of the severe economic contraction that the euro area economiesexperienced.For example, inSpain and Ireland, therewas clear evidenceof abalance sheet crisis driven by private-sector behavior. But how much of theseverity of the crisis in Spain, with an unemployment rate of 27 percent andrising(55percentandrisingamongtheyoung)wasduetothefactthattherewasinitially very limited fiscal expansion and then, when spreads rose, there wassignificantfront-loadedfiscalausterity?

RobertoPerotti inchapter16arguesthattheoptimalresponsetohighdeficitanddebtdependsinpartonwhetheracountryhas"fiscalspace"ornot,meaningwhether active bond market vigilantes have increased the country's sovereignspreadsand led toa lossofmarketaccessornot.Theargumenthasalsobeenmade that in the case of the euro area periphery therewas no fiscal stimulusoption:Ifthemarketsarepunishingacountryandspreadsarehighandrising,orifthecountryhaslostmarketaccess,theonlyoptionisfiscaladjustment.

This argument is only partially valid and includes at least three importantcaveats. First, whether a country has fiscal space or not depends in part onwhether it has a central bank that is willing to do quantitative easing andmonetizepublicdebt.Inthecaseoftheeuroarea,ifthebehavioroftheECBhadbeenmoredovish,theimplicationforfiscalspacewouldhavebeendifferent.

Thesecondpoint is theexistenceofacentralbank that iswilling toavoida

self-fulfillingbadequilibrium.Thisimpliesthatarunagainstthepublicdebtorthe widening of the sovereign spread can be avoided even if actual debtmonetizationdoesn't occur but is only available as an option.Think about theECB'sOMTprogram: In a sense, this has been themost successfulmonetarypolicytoolever,becausenotasingleeurohasbeenspentyettobackstopItaly,Spain, or anyother country.But the spreads in Italy andSpainhave fallenby250to300basispointsin2013,comparedtowhattheywereinthesummerof2012.

So,themereexistenceofapotentiallenderoflastresortcanleadtoabetterequilibrium, even if the lender doesn't act. That's an important factor indeterminingwhetheracountryhasmarketaccessornot.

Third, even if a country (its sovereign) doesn't have market access, eitherbecauseithaslostitorbecausetheprivatesectorisimposingmarketdiscipline,theexistenceofofficialcreditors(e.g.,theIMF,theEU,theEuropeanFinancialStabilization Mechanism [EFSM], the European Stability Mechanism [ESM])canprovide a sovereignwith some fiscal pace.Given the existenceof officialexternalcreditorsthatcansubstitutefortheprivateones,thequestionis,whatistheir optimal use? There has always been a debate about a country-say, anemerging market country that gets in trouble-and the existence of anorganizationthatprovidesitwithfinancialsupportconditionalonausterityandreforms.TheIMFallowsacountryunderfinancialstresstohaveabetterpathoffiscalconsolidationthanitwouldhaveifthatofficialfinancingdidn'texist.

In the case of the euro area, of course, the existence of official creditors (aeuro-area-wide lending and liquidity mechanism) also gives sovereigns underpressuresomedegreeof flexibility.Therefore, inevaluatingwhetheracountryhas fiscal spaceornot, the considerations abovematter.Countries such as theUnitedStates,theUK,andJapanstillhaveverylowinterestratesdespitealargefiscal deficit and debt in part because central banks have been willing to doquantitativeeasingandeffectivedebtmonetization.Meanwhile, incountries intheeuroareawheredebtratiosarenohigherthanthoseoftheUnitedStates,theUK, or Japan (and in some cases are lower), spreadswere high andwideningwhentheECBwasessentiallyrefusingtoprovidethattypeofmonetaryeasing.

Thenextissueworthdiscussingisfiscaldominanceandhowmuchoneshouldbeconcernedaboutit.Inasituationinwhichbudgetdeficitsarelargeandthereis apoliticalbias towarddeficits, there is always the riskof fiscaldominance.Theriskisthatacentralbankisgoingtobeforcedessentiallytomonetizethesedeficitstopreventadebtcrisis.Inagameofchickenbetweenafiscalauthorityandamonetaryauthority,itisthelatterthatblinksiffiscaldominancerules.

On the issueof fiscaldominance, there isadifferencebetween theviewsofthe ECB and the Bank of Japan (BOJ) under Masaaki Shirakawa (who bothworriedaboutthefiscaldominanceeffect)andtheviewsoftheFedandtheBOJunderHaruhikoKuroda(whodon'tseemtobeworryingaboutsuchrisk).

One can interpret the Fed views as follows. First, the central bank cannotreallybullyfiscalauthoritiesintofiscaldiscipline.Acentralbankcan'tthreatenthe fiscal authority and force it to make fiscal adjustments by denying anecessarymonetaryeasing.

Second,ifthecentralbanktriestobullythefiscalauthorities,itmightendupinapoliticalclashwith them,and theensuingbacklashcould lead toa formallossofcentralbankindependence.

Third, if the central bank withholds a necessary monetary policy stimulusbecauseitwantstoforcethefiscalauthoritytoadjust,itmightnotsucceedandmight actually cause a severe recession.So,monetarypolicy shoulddowhat'snecessaryfortheeconomyregardlessofwhat'shappeningonthefiscalside.

Thus, thebest thatacentralbankcando-as theFedchairmanhasdone-is tojawbonefiscalpolicymakerstomakethenecessaryfiscaladjustment.Usingthethreat ofwithholding necessarymonetary stimulus to induce fiscal adjustmenthasnegativeandperverseeffects.

However,theECB'sandtheGermanviewoftheissueoffiscaldominancehasbeenverydifferent.Forexample,inthecaseoftheeuroarea,theOMThasbeenmadeconditionalonstrictandeffectivefiscalandstructuralconditionalityasawayofactivelylimitingfiscaldominance.

Second,byreducingspread,theexistenceoftheOMThasledtoconcernsinGermanyandevenintheBundesbankthatthereispolicydelayandslackintheeuro area periphery. According to this view, complacency and moral hazardexist in the euro areaperiphery; austerity and reformsarenotoccurring at thepacedesiredby theECBand thecoreof theeuroarea.Therefore, theGermanviewandtheBundesbank'sviewisthatmarketdisciplineissometimesgoodandnecessarytoforcegovernmentstoimplementwithoutdelaythenecessarypolicyactions. On the other hand, excessivemarket discipline-in the form of higherspreadsis destabilizing attempts to reduce deficits andmake debts sustainable.Thus,marketdisciplineisadouble-edgedsword.

Certainly there are risksof fiscal dominance in thepresenceof the liquiditysupportof thecentralbank.And there isalsoa riskofmoralhazardwheneveryou have other official resources-such as the ESM, EFSM, and IMF loans-tosupportsovereignsunderpressure.Buttheargumentaboutmoralhazardisabitexcessive. Assume you are a government and you have to do painful fiscalausterity,butsupposethattheconfidencefairydoesnotcomealong(i.e.,spreadsdon't fall despite austerity and reforms) because there is uncertainty about thecredibility of the government actions and about how long the government isgoingtostickwiththepolicies.Thereisaseriousriskthatfiscaladjustmentandreformmay fail in a situation in whichmarkets do not yet find the country'spolicies fully credible. In this case, ifyoudon'thaveofficial support (whetherfromthecentralbankorfromofficialcreditors),theincentivetopracticepainfulausterityandmakereformsmightbelow,becauseevenifyouattemptthemyoumightfailandendupinadebtcrisis.Therefore,moralhazardisreducedratherthan increasedwith the presence of officialmoney because you have a carrotinducingthegovernment to implementpainfulpolicies thatare likely tofail inthe absence of official financing. Here, the carrot is that if the governmentimplements the necessary and painful efforts, liquidity will be provided thatreduces the risk of a bad equilibrium. Therefore, in the presence of officialfinancing, a self-fulfilling crisis is avoided andmoral hazard is reduced ratherthanincreased.

A few years ago, I wrote a paper with Giancarlo Corsetti and BernardoGuimaraes(Corsetti,Guimaraes,andRoubini2004)inwhichweshowed,using

theanalyticalframeworkofglobalgames,thattheexistenceofofficialfinancingcan reducemoral hazard rather than increase it. Official financing provides acarrotthatgivesanincentivetoagovernmentthatmightotherwisereason,"IfIam likely to fail,why should Imake the policy effort in the first place?"Thepresenceofofficialfinancingreducestheprobabilityofabadequilibrium,sotheargument thatofficialsupportcausesmoralhazardandfiscaldominancemightbeincorrect.

Thenextquestiontoaddressishowtoreduceadebtoverhang;thatis,whatisthe optimal approach to deleveraging from high levels of public and privatedebt?Thereare severaloptions.The firstone is fiscal austerity: agovernmentcancutspending,raisetaxes,andthusincreasepublicsavings.Butthatoption,iftoofront-loaded,leadstotheKeynesianparadoxofthrift:ifthefiscaladjustmentistoofast,theeconomymaycontractagain,andthegoalofreducingdeficitsanddebtmayfail.This ispartlywhathasbeenhappening in theeuroareaand theUK.The second option is a coercive debt restructuring/reduction. That optionmightbecomenecessaryandunavoidableifacountryhasanissueofsolvencyratherthanliquidity.

Another option is very aggressive monetary policy (zero policy rates andquantitativeeasing);thisiseffectivelyaformofdebtmonetizationthatleadstoalownominal interest rate and possibly negative real interest rates.This optionmightnotbeinflationaryifthecountryisinaliquiditytrapandthereisalargeamount of slack in the goods and labor markets. A variant of this option isfinancingfromofficialcreditors(theIMF,theEU),whichprovidesabreathingspaceforasloweradjustmentofspendingandsavings.

Another option is to cause expected or unexpected inflation towipe out thereal value of nominal public debt. Yet another optionmight be to use capitallevies on wealth or on creditors as a way of resolving the debt overhang. Avariant of that is to use financial repression and capital controls to keepgovernmentbondyieldslowerthantheyotherwisewouldbe.

All except the first option, adjustment and higher savings, imply someredistributionofwealthfromcreditorsandsaverstodebtorsandborrowers(e.g.,

an indebted government). But the adjustment from a debt overhang thatdepresses the spending of debtors necessarily implies that such a transfer ofwealthshouldoccurovertime.Thepoliticalquestionis,whoshouldbemakingsuch redistribution policy decisions? Should the decision be made by fiscalauthoritiesthroughacapital levyonwealthordebtrestructuring,orshouldthecentral bank have the power to accomplish the same result via debtmonetization?

Somearguethatsuchpowershouldnotbegiventothecentralbank,astheseare eminently fiscal and thus political decisions.But if lownominal rates andnegative real ones from debt monetization allow a smoother deleveragingprocess that reduces the risks of a recession deriving from excessive austerityand prevents amore disruptive debt restructuring or the distortionary costs offinancial repression, capital controls, andcapital levy,debtmonetizationcouldbetheleastcostlyoption.

The final issue worth addressing concerns the optimal pace of fiscalconsolidationfollowingafinancialcrisisthathasledtoalargeincreaseinpublicdebt.

Inmost cases inwhichmarket access has not been lost or inwhichofficialfinancing(fromthecentralbankorfromofficialexternalcreditors)isavailable,theoptimalpaceoffiscalconsolidationwouldentailafiscalstimulusintheshortrun, while the economy is weak and the private sector is deleveraging itsindebted balance sheets, plus a credible plan formedium - to long-term fiscalconsolidation,tobeimplementedwhentheeconomyhasrecoveredenoughandprivate balance sheets are mostly mended. This short-/medium-/long-termadjustment can be helped by effective debt monetization by a central bank,whichmakesthedeleveragingprocesssmootherandavoidstheriskofrunsonbanks and governments. Delaying fiscal adjustment forever can cause a"zombification"ofgovernmentsandprivateagentsandeventuallyadebtcrisis,eveniftheinitialsurgeinthepublicdeficitwasduetoaprivate-sectorbalancesheetcrisis.Conversely,excessivelyfront-loadingthefiscalconsolidationriskspushinga fragileeconomy intoadouble-dip recession thatwillonlymake thedeficitanddebtproblemsworse.

Compared to this optimal path of fiscal consolidation, the major advancedeconomies-the euro area, the UK, and the United States-have followed asuboptimal path. In the euro area and theUK, fiscal austerity has been front-loadedintheshortrun.Thisisoneofthereasons-alongwithamonetarypolicythat wasn't loose enough and zombie banks that were not appropriatelyrecapitalized, thus driving a credit crunch-why both regions have fallen into adouble-diprecession.

IntheUnitedStates,gridlockandlackofbipartisanshipinCongressmakeitlook as though the country has no credible path for medium - and long-termfiscalconsolidation,asRepublicansblockfurthertaxincreaseswhileDemocratsblockspendingandentitlementreforms.Meanwhile,intheshortrun(thankstothesamegridlock),thefiscaldragwillbeexcessivein2013becauseofafrontload of fiscal consolidation engendered by the sequestration deadlock. So theUnited States is taking the opposite of the optimal path: too much short-runfiscal drag in 2013 and 2014, with no credible plan for consolidation in themedium term.Still, compared to the euro area and theUK,whichhave front-loadedtheirfiscalausteritysince2011,theUnitedStatessuccessfullypostponedits central government austerityuntil 2013,which explainswhy the country isgrowing, however anemically, while the euro area and the UK have beencontracting.

The fiscal adjustment in the euro area has been an example of poor policyplanning;itexplainswhytheperipheryoftheregionisstillstuckinarecessionthat is becoming a near depression in some countries (Greece and Spain) andwhy the recession is even spreading to some parts of the core (France andBelgium). A few remarks are important concerning the euro area fiscaladjustment.

First, the austerity has been excessively front-loaded in the euro areaperiphery, with seriously damaging effects. It should be significantly back-loaded.

Second, the flexibility that theEUCommission allows in achieving cyclicaladjusted targets and the fiscal relief it offers to countries in recession are

palliatives that occur after the patient has nearly been killed. Setting veryaggressive structural fiscal targets for a country, then allowing some cyclicalslackonthose targetsonce theausteritymeasureshavecausedarecessionthatworsened the deficit is no rational solution. Structural fiscal targets should berealisticinthefirstplacetoavoidsuchausterity-inducedrecession.

Third, the adjustment in the euro area is asymmetric betweenperiphery andcore,and thusrecessionaryanddeflationary. Ifcountriesandgovernments thatoverspend and undersave should spend less and save more, countries (likeGermany) that oversave and underspend should spendmore and save less (inpart,throughfiscalpoliciesoftaxreductionandspendingincrease).Otherwise,the adjustment is asymmetric and recessionary, as it leads to a shortage ofaggregatedemandforthewholeregion.

Fourth,thereisfiscalspaceinthecoreoftheeuroarea,especiallyincountriessuchasGermany,whereinterestratesarelowandmarketaccessisample.Itistrue that the public debt in Germany is 80 percent of GDP and that implicitliabilitiesfromagingareadditionalfiscalburdens.Butsovereignspreadsaresolow inGermany that if thecountrywere to implement for a coupleofyearsafiscalstimulus in theformofareductionin taxesandincreases ingovernmentspendingasawayofboostingitsowneconomicgrowthandthatoftheoveralleuro area, this policy actionwouldnot lead to any loss of fiscal credibility aslongas thecountryhadaplan formedium- to long-term fiscal consolidation.What we are speaking about is a short-term program to try to jump-starteconomicgrowthinaeuroareathatisinadeeprecession.

Fifth, while the growth problems of the euro area are more structural thancyclical comparedwith thoseof theUnitedStates, in2012-2013 theeuroareadidn'tgrowevenat its lowpotentialgrowth rate, as itwas ina recession.Thefactthattheregionwasstuckinarecessionsuggeststhatthelackofaggregatedemand, not just supply-side constraints, explains this persistent economicdownturn.

Moreover, someof thenecessary structural reformsare, like fiscal austerity,contractionary in theshort run.Suppose, forexample, thatacountrymakes its

labormarketmoreflexibleandreduceshiringandfiringcosts.Thefirstimpactofsuchreformwillbearise intheunemploymentrate,as thefirmsthatcouldnot fire redundant workers will now be able to do so. That surge in theunemploymentrateisexactlywhathappenedinGermanywhenitimplementeditsstructuralreformsintheearly2000s.Theimplicationisthattherehastobeatrade-offbetweenstructuralreformsandfiscalausterity,ratherthanadamagingrecessionary front load of both. For example, if a country implements rapidstructural reforms that are recessionary in the short run, it should be givengreater fiscal flexibility, as the reformsmightmake the recessionworse in theshortrun.ThatisexactlywhatGermanywasclaimingwhenitimplementeditsAgenda2010anditsunemploymentratewasrising;theGermansaskedtheEUfor a fiscal break because their reformswere causing a rise in unemploymentthatwasleadingtoalargerfiscaldeficit.Theyargued,correctly,thatthereformswouldeventuallyincreasegrowth.Intheshortruntheywerecausinganincreaseinthedeficit,butthatdeficitwouldgoawayoncetheeffectsofthereformsongrowthmaterializedinthemediumterm.

So,therehastobeatrade-offbetweenausterityandreforms.Youcannotjustfront-loadboth the austerity and the structural reforms. If youdomoreon thestructuralside,youhavetoprovideforgreaterfiscalflexibilityintheshortrun;otherwise,therecessionislikelytobecomemoresevere.

A final observation about the euro area:There is absolutely no talk about agrowth agenda.There is talk about abankingunion, a fiscal union, a politicalunion. But if you're not going to have economic growth, and if the austeritymakes the recession worse, you will eventually have a social and politicalbacklashagainsttheausterity.Also,sovereignsaretryingtostabilizetheirpublicdebt,domesticandforeign,asashareofGDP.Butonecanworkasmuchasonewants on thenumerator of the debt ratio, but if thedenominator (GDP)keepsfalling,debtratioswillkeeprisingandwillbecomeunsustainable.Thatiswhatis happening right now in the euro area: Despite draconian fiscal adjustment,publicdebtratiosarestillrising,andtheymayeventuallybecomeunsustainable.

Excessively front-loaded fiscal consolidation in both the periphery and thecore of the euro area has been counterproductive. It is an important factor in

explaining why the region has been stuck in a deep double-dip recession in2012-2013.

References

Corsetti, Giancarlo, Bernardo Guimaraes, and Nouriel Roubini. 2004."International Lending of Last Resort and Moral Hazard: A Model of IMF'sCatalyticFinance."CEPRDiscussionPapers4383.London:CEPR.

Herndon, Thomas,Michael Ash, and Robert Pollin. 2013. "DoesHigh PublicDebtConsistentlyStifleEconomicGrowth?ACritiqueofReinhartandRogoff."University of Massachusetts Amherst, Working Paper #322, Aprilhttp://www.peri .umass.edu/fileadmin/pdf/ working_papers/working_papers_301-350/WP322.pdf.

Reinhart, Carmen, and Kenneth Rogoff. 2010. "Growth in a Time of Debt."AmericanEconomicReview(May):573-578.

AgustinCarstens

Thechoiceof theappropriateexchangerateregimeforanycountryisanissuethat has been extremely important in the past and still is today. It is a policydecision that, to a large extent, conditions themacroeconomic frameworkof acountry. This chapter discusses a number of issues related to this importantdecision.Firstitdiscussestheimportanceofchoosingtheexchangerateregime.It then analyzes the implications of the degree of exchange rate rigidity orflexibility for the domestic economy, particularly for other macroeconomicpolicies.Thenitdiscussesthespecificcaseoftheeuroarea.Finally,itexploresthesideeffectsassociatedwiththechoiceofanexchangerateregime.

There is hardly a more important economic decision that a country has tomakethanchoosingitsexchangerateregime.Thereareatleasttworeasonsforthat:

• First, such a decision conditions the scope and flexibility of the rest of themacro policies that the country can implement. Certainly this considerationalonemakeschoosingtheexchangerateregimeanextremelyrelevantdecision.

•Second,suchachoiceaffectsthecountry'seconomicrelationshipwiththerestoftheworld.

Theconditioningeffectsofthechoiceofanexchangerateregimeimplyverydifficulttrade-offsthatpolicymakershavetoevaluatewithutmostdeterminationand care when deciding on the exchange rate regime that best suits theircountry'seconomy.

Inparticular, thedecisionboilsdowntoestablishingthedegreeofflexibilitythat theexchangerateshouldhave, that is,howmuch therelativevalueof the

domesticcurrencycanvarywithrespecttoothernationalcurrencyorcurrencies.Usuallywhendefininganexchangerateregime,authoritiesrefertotherelativepriceof thedomesticcurrencywith respect to thecurrencyof itsmain tradingpartnerand/orthecurrencythatdominatesintheregion.

In principle, it can be said that the less flexibility that is allowed for theexchangerate, themorerigiditythecountryimposesontherestof itspolicies.For instance, when a country fixes its exchange rate, it sacrifices monetarypolicy as an independent policy instrument, because to a large extent, thecountryimportsthemonetarypolicyofthecountrythatitfixesthevalueofitsown currency against. Under these circumstances, the ability of the nationalcentralbanktoperformitsdutyaslenderoflastresorttothebankingsystemorthe sovereign is severely compromised. Furthermore, the sustainability of theregime also imposes restrictions on fiscal policy and makes it essential topreserveahealthyfinancialsystem.

Infact,insofarasfixingtheexchangerateimposesverystrongconditionsontherestofthemacroframework,choosingsucharegimeusuallyhasbeenusedasacommitmentdevice.Thatis,foracountrytomakeaspecificexchangeratesustainable, it has to observe all the constraints such a regime entails. Putdifferently,thestrongerthepolicycommitment,themorecredibletheexchangerateregimewillbe.

Nevertheless, for decades the real challenge formost countries has been toacquire and preserve the ability and political will to remain true to suchcommitmentsandkeeptheexchangerateregimecredible.Timeaftertimeithasbeenseenthateventually,aftersomeyears,sometimesevenaftermonths,thesecommitments tend to be ignored, and the exchange rate regime collapses.Theway that authoritieshandle a regime switch is alsovery important since if theswitchismishandled,ithasthepotentialtogeneratenegativeconsequencesforgrowthanddevelopmentforyearstocome.

It shouldbenoted that establishinga flexible exchange rate regime imposesothertypesofrestrictionsandexposesthecountrytoothertypesofrisks.Undersucharegime,themainchallengeistoconstructaneffectivenominalanchorfor

theeconomy.Thereisabroadrangeofoptionsinthisregard,rangingfromtheadoptionofmonetaryaggregateobjectives to inflation targeting.Recenteventshave shown the limitations of infla tion targeting, in particular the fact thatcentral banks should not ignore asset-price inflation so that they can preservefinancialstability.

Regarding fixing the exchange rate, there are many different ways toimplementit.Thesimplestalternativeistofixthevalueofagivencurrencywithrespect to another currency. The "ultimate" fix would entail establishing acurrencyunioninaregion,allowingnationalcurrenciestodisappearaltogetherin favor of adopting a single currency such as the euro. The sustainability ofsuchacomplexregimeimposesverystrictanddemandingconditionsonallthemember countries. With respect to the euro area, the real question is howeffectiveithasbeenasawholeinfulfillingthoseconditionsinacredibleway.In this setting, a lot has been left to be desired; therefore, important effortsthroughpolicyadjustmentshavetobemadenotonlytopreservethecredibilityoftheregimebutalsotoenhanceit,sothatthepossibilityofthebreakdownoftheeuroistakenoutofthepictureforgood.

Thereisanotherveryimportantdimensiontotoexchangerateregimes,relatedtothefactthatanexchangerateistherelativepriceoftwocurrencies.Therefore,fromanationalperspective,sucharelativepricecanbeaffectedbytheactions(monetaryexpansion,capitalcontrols,etc.) thatothercountriesmayundertake,suchactionsoftenhavingnegativeunintended(orintended)consequencesontheexchange rate. Thus, the domestic currency can appreciate or depreciate inresponsetoactionstakenbyanothercountry,givingrisetodebatesencapsulatedbyphrasessuchas"currencywars"or"beggarthyneighborpolicies."

Withrespecttothispoint,ithasrecentlybeenarguedthatmonetarypoliciesinsome advanced countries have generated huge capital inflows to othereconomies,particularlyemerginganddevelopingones,consequentlygeneratingsubstantial real exchange rate adjustments. Some of these adjustments arenonequilibrium appreciations, affecting the tradables sector and the growthpotentialofemergingeconomies.Thisisatypicalexampleofaspillovereffectofanadvancedeconomy'smonetarypolicyaction.Furthermore,suchaneffect

can trigger or lead to different types of reactions, including some form ofretaliation,leadingtonetdeadweightlossesfortheworldasawhole.

To deal with these issues, international coordination might be attemptedaroundasetofexchangerateobjectives.ThishasbeendonewiththehelpoftheIMFinthepast.Avalidquestioniswhethertheinternationalcommunityshouldtrytodothisagain.

Insum,thechoiceofanexchangerateregimeisafundamentaldecisionthatacountry has to make. Such a decision has important implications for theeconomy as awhole, as it affects the degrees of freedomof other policies, aswellastherelationshipofthedomesticeconomytothatoftherestoftheworld.Therefore, the huge political economy implications that the election of theexchangeratearrangemententailsmakesitahighlyrelevanttopic.Toillustratethis, Iwill finish on a lighter note,with an anecdote:At the beginning of the1990s, Mexico was facing a major crisis, and for practical purposes its onlyoptionwastogofromafixedtoaflexibleexchangerateregime.Societywasnotsupportiveatall.Thethencentralbankgovernorwassummonedbythesenatetoexplain the rationale for Mexico going from a very unsuccessful fix withrecurrent adjustments, with often tragic consequences, to a flexible exchangerate.Andhecouldn'tgetthepointacrossthatforMexico,thebestdecisionwastogotoaflexibleexchangerateregime.

Finally,thegovernorhadabrilliantideatoexplainsuchamovebyusingananalogy.He said to the senators, Imagine that youhave the taskof painting ahouse. Now, you can do it in two ways. One way would be to place thepaintbrush in a fixed position and move the house, which would be theequivalenttofixingtheexchangerate.Thealternativeistokeepthehouseinitspositionandmovethepaintbrush.Withasteadyhand, it isverylikelythat thetaskcouldbecompletedinacheaper,morebeneficial,moreefficientwaywithaflexiblebrush.

This simple analogy convinced the senate, and since then we have had inMexicoaflexibleexchangerate,andithasworkedverywell.

JayC.Shambaugh

Theexchangerateregimedecisionisoneofthemostimportantonesacountrycanmake in termsofmacroeconomicpolicy. It has important implications forhowacountrywillmanageitsfinancialaccountanditsmonetarypolicyoptions.Thisbriefchaptercannotrunthroughallaspectsofthedecisionbutwillinsteadfocusonone thing:whathavewe learnedfromthe2008-2009crisisabout theexperience of fixed versus floating exchange rates, aswell as the institutionaldesignofcurrencyunions.

My thesis is that we have relearned many things we should already haveknownbefore thecrisis. Inmanyareas, suchas fiscalpolicyat thezero lowerbound,monetarypolicyatthezerolowerbound,ormacroprudentialpolicy,thiscrisissurprisedmanyand,at thevery least,providedevidencefor issuesaboutwhichwedidnot have a clear understanding.Conversely, regarding exchangerate regimes, an undergraduate who had taken a basic internationalmacroeconomicsclasscouldprobablyhavedescribedaheadoftimemostofthekey events that took place. That student would have known that in general,floatingcanserveasashockabsorber,externaladjustmentmaybeeasierifyoufloat, and entering a currency union at the wrong price can be very painful.These are thingswe already knew, but they are also thingswe saw time andagainthroughoutthecrisis.

One surprisewas thatwhenpegsbrokeduring thecrisis, theydidnot spiralwildly out of control. This time around, when a fixed exchange rate broke,countriestendedtoloosenthebandstheykepttheexchangeratein,asopposedtohaving a sharpdepreciation.Thatwas a newexperience, but theother corestylized facts are things that academics and policymakers should haveunderstood.

Thesesamelessonscanserveasremindersforwhattheinstitutionalstructureshouldbeforcurrencyunions,and theyhighlight thecostof inadequateshockabsorberswithinacurrencyunion.

FloatingasaShockAbsorber

Apopularifratherdarkjokeofearly2009asked"What'sthedifferencebetweenIrelandandIceland?"Theanswer:"Oneletterandsixmonths."Asitturnedout,therewasanotherverystarkdifference(amongmanyothers):onehadafloatingexchangerateandtheotherdidnot.Comingintothecrisis,Icelandwaslikelyinmuchworseshape,withacurrentaccountdeficitgreaterthan20percentofGDPinoneyear,whereastheexternaldeficitpeakedatcloseto5percentinIreland.Both,though,hadlargecreditboomsandsomewhatspectacularfinancialbusts.Andyetthetwocountrieshadverydifferentexperiences.

In figures 19.1a-f, the solid line is Ireland and the dotted line is Iceland.Icelandexperiencedasharpdepreciationwhenthecrisishit(figure19.1a).Ithadahigher inflationratecomingoutof thecrisis (figure19.1b).But the inflationdid not offset all of the depreciation, so therewas a substantial real effectiveexchangeratedepreciation(figure19.1c).Intherealeconomy,realexportsgrewfaster (figure 19.1d). The figure for real GDP (figure 19.1e) shows a modestdifference, although, depending on the base year used, this picture could lookdifferent. The last figure (figure 19.1f), though, shows the starkest difference:thesharpjumpintheunemploymentrateinIrelandrelativetothatinIceland.

Therearemanyotherdifferencesbetweenthesetwocountries,andIcelandhasnotnecessarilybeenaparagonoffinancialpolicymanagement,butthecontrastisareminderthatwhentherearesevereshocks,itisusefultohavethepressurereleasevalveoftheexchangerate.

Ifonelooksmorebroadly, therewereanumberofcountries,whetherIsrael,Poland,Sweden,ortheUK,thatwhenthecrisishitbenefitedfromtheabilitytohavetheirexchangeratechange.Evenmorebroadly,inthetwodecadesbeforethecrisis,peggedandnonpeggedcountries tended togrowataround thesamerate.But during the crisis the pegged countries grew about a percentage point

more slowly. There may have been different shocks, as well as many otherpoliciesallmixed together in that result.So thepoint is innowayconclusive.And some countries that maintained a floating exchange rate in the past fiveyearsfelttheywereappreciatingtoomuch,andtheylikelywouldhavepreferredtohavebeenpegged rather than to experience the appreciation.But a floatingexchange rate, both in theory and from casual observations made during thecrisis,canserveasashockabsorber.

Figure19.1

a. Exchange rate versus the euro (Jan. 1999= 100). b.Consumer Price Index(2006=100).c.Realeffectiveexchangerateindex(2006=100).d.Realexportindex(2006=100).e.RealGDPindex(2006=100).f.Unemploymentrate.

ExchangeRateChangesandExternalAdjustment

Thebiggerquestioniswhathappenswhenadjustmentisneeded.Peggingversusfloating should not make a country grow faster or slower over time, butexchangerateflexibilityshouldhelpinmakinganexternaladjustmentifoneisneeded.

Looking now at external adjustment with fixed rates, one thing economistshaveknownfora long time is that it shouldbemoredifficult foracountry tomanageasubstantialchangeinitscurrentaccountiftheexchangerateisfixed.Prices are sticky. They are especially downward sticky. Thus, it is hard forsubstantial real depreciation to occur if the nominal exchange rate cannotchange.' This should make it more difficult to have a substantial externaladjustment.

Table19.1 looks at10countries thathadvery largecurrent accountdeficitsgoingintothecrisisandverylargeexternaladjustmentsduringthiscrisis.Theyentered the crisis with a current-account-deficit-to-GDP ratio of at least 10percent, and they cut the ratio by at least 10 percentage points. There is a

roughly even split between pegged and nonpegged economies. This is quiteimportant, as it highlights the fact that a fixed exchange rate does notmean acountrycannothaveexternaladjustment.

Table19.1

ExperienceofCountriesMakingMajorExternalAdjustmentintheCrisis

On the other hand, of the six countries that were pegged, none of themexperiencedrealGDPgrowthovertheperiodfrom2008to2011,whereasthreeof the four nonpegged countries did. On average, the pegged countriescontracted by about 10 percent, as opposed to an expansion in the nonpeggedcountries. The nonpegged countries experienced a relatively mild averagechange in the exchange rate against the base country; theywere not all largedepreciations, but there was a substantial difference in the real economyexperiencesofthecountries.Thesearesmallcountries,anditisasmallsample.This is not rigorous econometric evidence, but the experiencewewould havepredictedgoingintothecrisisdidbroadlyhappen.Whenfinancialflowsseizedupanditwasmoredifficulttofinanceexternally,thepeggedcountriesthatmadelargeadjustmentshadahardertimethandidthefloatingratecountries.

RelativePricesonEntryintoaCurrencyUnion

Economistshavealsoknownforalongtime,certainlysince1925,whenKeyneswaswriting critical things about Churchill (Keynes 1925), that if you enter acurrency union at thewrong price, it can be painful. Keynes pointed out thatrejoiningthegoldstandardatthewrongpricewasverypainfulforBritain,andthis is something that both theory and the experienceof countries afterWorld

WarIdemonstrate.

When one looks at the lead-in to the euro, the convergence criteria wereinflation, debt and deficits, and a stable exchange rate. In a recent discussionabout the euro area with my students, I asked, "What were the convergencecriteria?"Astudentraisedherhandandsaid,"Thebalanceofpaymentsdeficitcouldn't be too big." Someone tapped her on the shoulder and said, "No, youmean `fiscal."' She said, "Right, right, fiscal deficit can't be too big." And Ithoughttomyself,thatissomewhattelling.Whileitmayseemintuitivethattheywould, the rules were not putting limits on the balance of payments deficit.Therewasnorulethatpurchasingpowerparityhadtohold.Policymakerswereabout to fix the nominal exchange rate forever. But, confirming that relativepriceswereattherightlevelwasnotpartoftheentrycriteria.

Overtheperiodfromentryuntil2007,currentaccountdeficitsandsurpluseswerequitepersistentintheeuroarea.Figure19.2ashowsthecurrent-account-to-GDPratioatentry(x-axis)andthecurrent-accountto-GDPratioin2007(y-axis).

Therelationshipisquitestrongandtheslopeisnearly1.Thispersistencehasbeendocumentedbefore(see,e.g.,LaneandPels2012orKangandShambaugh2013fordiscussion).Thenexttwofigureslookacrossothergroupsofcountriestoseeifthepatternisanydifferent.

Figure19.2blooksatthemorerecententrantstotheeuroareaand/orcountriespeggedtotheeuro.

Thelinealsohasaslopecloseto1,butthefitaroundthelineisnotquiteastight (the R' is lower). Figure 19.2c takes the remaining countries from theEurostat database that are neither the original 12 entrants into the euro (thosethatenteredpriorto2003)norlaterentrantsandpegs.

There are roughly 15 countries floating in that data set. Relative prices canmove fluidly for these countries, as thenominal exchange rate is floating.Yetthereisalmostaperfectlystraightlineforthisgroup,too.Inthebuilduptothecrisis, current accounts, no matter the exchange rate regime, were verypersistent.Countries thatwererunningdeficitskept runningdeficits.Countries

thatwererunningsurpluseskeptrunningsurpluses.Theredidnotappear tobepressure topushcountriesback towardabalancedcurrentaccount. Itdoesnotseemthatfloatingexchangerateshelpedthesecountriesbacktobalance.

However, thequestionbecomes,onceadjustmenthadtohappen,wasthereadifference?Heretheanswerappearstobeyes.Figures19.3a-cshowthecurrentaccount in 1999 (or entry into the euro currency union) on the x-axis and thechangeintheunemploymentrateduringthecrisisonthey-axis.

Therelationshipappears tobesteeperfor theearlyeuromembers.Countriesthathadabigcurrentaccountdeficitatentry into theeurowounduphavingamuchhigherspikeintheunemploymentrateonceadjustmenthadtohappen.Forpeggedcountriesorlateentrants,therelationshipisnotassteep,meaningthatagiven current account deficit correlates with a smaller increase in theunemploymentrate.Forcountrieswithafloatingexchangerate,therelationshipis essentially flat. The current account in 1999 had almost no bearing on theunemployment response during the crisis. The current account deficit is acountry's net borrowing from the world-the country's borrowing, not thegovernment's. Those countries thatwere borrowing extensivelywhen the eurobeganappear tohavepaidapriceonceadjustmenthad tohappen.That is lesstrueforfloatingexchangeratecountries.Theyhadmoreflexibilityinhowtheyapproachedthecrisis.

Figure19.2

a. Early entrants, current-account-to-GDP ratio. b. Pegs and late entrants. c.Floats.

Figure19.3

a.Euro.b.Lateentrantsandpegs.c.Floats.

Itmaybeworthnotingthatmanyofthemorerecententrantsintotheeuroareahavehad fairly substantial current account deficits on entry.Themost notable

caseisthatofCyprus,whichhadacurrentaccountdeficitof16percentofGDPinitsfirstyearasamemberoftheeuroarea.Slovenia(5percentdeficit)andtheSlovak Republic (3 percent deficit) did not have extremely high deficits, butwerecertainlynotinexternalbalance.Estonia,whichhada2percentsurplusin2011, is the only recent entrant with a surplus. These recent entries raise thequestion of whether exchange rates are being fixed permanently at the rightprice.

TheNon-ExchangeRateCrisisofBreakingPegs

Oneexperienceinthecrisisthatissomewhatnewisthewayinwhichexchangeratepegsbroke.Itiswellestablishedthatexchangeratepegsarefragile.2Figure19.4 shows the percentage of countries that were pegged in any given yearbetween1973and2011thatbroketheirpeg.3Thereisasmallspikerightaround2008,somorecountriesthanusualbroketheirpegsduringthiscrisis,whichisnotsurprising,giventhenatureofthecrisis.

Figure19.4

Percentageofcountrieswithpeggedexchangeratethatbroketheirpegs,byyear,1973-2011.

Figure19.5

a.Median annual range of exchange rate change across countries in year thatthey broke a peg, post-2006 versus 1973-2006. b. Fraction that held tight torangeafterbreakingpeg,post-2006versus1973-2006.

Whatwasmoresurprisingiswhatisseeninfigures19.5aandb.Figure19.5ashowsthatofthosecountriesthatbroketheirpegafter2006,therangeoftheirexchangerateagainstthecountrytheywerepeggingtowasnotparticularlylargeaftertheybrokethepeg.

Themedianrangeisjustover6percent,relativetoamedianrangeofover10percentinthepost-1973era.Inasense,manyofthesepegsdidnotreallybreak.The countries simply loosened the band somewhat. Figure 19.5b makes thisclearer.Itshowsthepercentageofthecountriesthatbroketheirpeg,butmerelywent froma2percentband toa5percentband,abroader range,butnothighexchangeratevolatility.

Almosttwo-thirdsofthecountriesinthecrisisthatbroketheirpreviouslytightpeg simply loosened thebands and stayed somewhat pegged to thebase.Thiscontraststoroughly40percentinthe1973-2006timeperiod.

The above findings suggest a strong preference in some countries forexchangeratefixity.Eventhoughthecrisishighlightedsomeofthechallengesthatcomewithhavingafixedexchangerate-specifically,thatadjustmentcanbedifficult with a fixed exchange rate-plenty of countries want to stay close topegged,eveniftheycannotmaintainatightpeg.Inanygivenyear,aroundhalfthecountries in theworldhavea fixedexchange rate. Ifwedropanumberofvery small countries from the sample, it is close to 40 percent pegging.Also,though, which countries peg changes from year to year. The debate overexchangerateregimesdoesnothaveacornersolutionthateconomistsoftenlike.Noteverycountryisgoingtofloat,andnoteverycountryisgoingtopeg.Thisremindsusoftheimportanceoffiguringoutthecostsandbenefitsofpegging.

ImplicationsforCurrencyUnions

Anumberof thesestylizedfactshaveimplicationsforcurrencyunions.Again,manyofthethingswehaveobservedincurrencyunions-particularlyintheeuroarea-are things thatwerewell understood in the literaturemore than a decadeago.

Certainlyfromamacroeconomicperspective,theeuroareahasnotdonewell.GDPgrowthhasbeenslower than thatofa typicalcountryandevena typicaladvanced country. The unemployment rate is much higher than in otheradvanced economies aswell. These facts, though, should not necessarily be areflection on thewisdom of currency unions. The euro area floats against the

world,sotheaverageperformanceoftheeuroareaisnotareflectionofthefactthat it has a currency union internally. Weak economic performance couldsimplybeareflectionofworseshocks.Alternatively,itcouldbeareflectionofworsemacroeconomicpolicymanagement,ofmonetarypolicythathasbeentootightorfiscalpolicythathasbeentootight.

What tells us more about the wisdom of the euro as a currency and theinstitutional structure that supports it is what happens when the need foradjustment differs across countrieswithin a currencyunion. In that sense, thiscrisishasrevealedsomeofthetrueproblemswithintheeuroarea.Figures19.6aandbshowtherangeandstandarddeviationofunemploymentratesacrossUSstatesandacrosstheeuroarea(thegraybarsaretheUnitedStates,theblackaretheeuroarea).

Figure19.6

a. Range of unemployment rates, United States versus euro area. b. Standarddeviationofunemploymentrate,UnitedStatesversuseuroarea.

Inbothareas,thereisastarkjumpwhenthecrisishitsfrom2007to2010.Therangeand standarddeviationofunemployment rates arehigher.From2010 to2012,though,thedispersionstartscomingdownintheUnitedStates.Thestateswithhighunemploymentratesarehavingtheirunemploymentratefallfaster.Inthe euro area, the dispersion has kept going up as unemployment rates havefalleninsomecountriesthatwerealreadydoingwellbuthavecontinuedtoriseintheworst-hitcountries.

This sequence of events suggests a lack of adequate shock absorberswithinthe euro area. Differential monetary policy within the currency union andexchangerateadjustmentsarenolongerpossible,andlabormobilityandfiscalfederalismarenotsufficienttooffsetshocksinoneregionandsmooththeminrelationtoanother.Again,thisisnotasurpriseoranewobservation.Theentireoptimalcurrencyareatheorywasfoundedbyapapernotingtheimportanceofashockabsorber(labormobility)foracurrencyuniontobeanoptimalcurrencyarea(Mundell1961).Itwaswellknownatthefoundingoftheeurothatformingacurrencyunionwithouttheseshockabsorberswaslikelygoingtobepainfulifasubstantialasymmetricshocktookplace.4

Summary

A number of important observations seem to have come out of this crisis.Countrieswithfixedexchangeratesneedtorecognizethatexternaladjustmentisgoing tobevery challenging.This suggests two things: one, it is important toavoid excessive borrowing if there is a fixed exchange rate, again, notnecessarilybythegovernmentorthestatebutfortheeconomyoverall,becauseif the country needs to rebalance the external account rapidly, the adjustmentwilllikelybeverypainful.Andtwo,itsuggeststhatifexchangeratesarefixed,thereneedstobesomeotherwaytocushionbigshocks.

The contributions to this volume emphasize the way fiscal policy andmonetarypolicyoperateatthezerolowerbound,andthisworksuggeststhatitisimportant to have fiscal stabilizers within a currency union, because if themonetary policy is going to run into the zero lower bound, not having anyagencythatisactingtostabilizetheeconomywithfiscalpolicyisgoingtobeaproblem. In addition, states or countries within a currency union should notradicallycutbackontheirfiscalpolicyinarecession,becauseweknowthey'regoing tobedoing so in an environmentwithveryhighmultipliers.Therefore,currencyunionsalsoneedtohavesomewaytoavoidthatsituation.

Also, economists and policymakers have learnedmore about themacro andfinance linkagesandwhatsomehave referred toas"doomloops"between thebanksandthesovereigns,highlightingthefactthatitiscrucialthatweaknessesinbanksnot tobecomeanasymmetricshockwithinacurrencyunion.IfwhenWashington Mutual got into trouble the state of Washington had beenresponsibleasafiscalbackstop, thestateofWashingtonwouldhavebeeninaverychallengingfiscalposition.InsteaditwasanFDICproblem,notaproblemforthestateofWashington.Thishighlightstheneedfordepositinsurance,banksupervision,alenderoflastresort,andbankresolution,allatthecurrencyunionlevel,becauseiftheydonotexist,itsuggeststhatthefinancialsystemcouldbetheoriginatorofasymmetricshocks.

Euro area institutions are changing, in some cases far more rapidly thananyonecouldhaveimaginedfiveyearsago.Ontheotherhand,oneofthethings

the crisis seems to highlight is that we cannot ignore what we know. Theimplications of exchange rate regimes are an area of economics in which thestandard literature could have predictedmost of what took place. It was wellknownprior to the launchof theeuro that labormobilitywasweakand fiscalfederalismnonexistent.The fact that this lackof shock absorberswouldmaketheeuroariskybetwaswellestablished.Likewise,thelackofanofficiallenderof last resort and the lack of a prudential supervisory role for the ECBwerewidelydiscussed.'Asaprofession,wehavelearnedmoreaboutmacrofinanciallinkages,buttherisksofacurrencyunionwithfragmentedbackingsupervisionwereunderstood.

Whenone steps back and considers the exchange rate regime decision, it isclearthatverysmallcountriesaregoingtopeg,especiallyiftheyaretiedtooneeconomy.Also,itseemsthatverylargeeconomieswithopenfinancialmarketsarenotwillingtosubjugatemonetarypolicytotheexchangerate.Thequestioniswhat countries inbetween shoulddo.Thecountriesof the euroareaclearlyhavedecidedtosacrificeotherpolicyoptionsforexchangeratestability.Doingso, though, without the necessary institutional structure was a gamble. Thegamble was that the needed change in the institutional structure and in theeconomywouldhappenbefore abig crisis.Thequestionnow is, now that thecostsareapparent,canpolicymakersfillintheneededinstitutionalstructurefastenough?

Notes

1.SeeShambaugh(2012)fordiscussion.

2. See Obstfeld and Rogoff (1995) and Klein and Marion (1997) for earlierdiscussions. Klein and Shambaugh (2008) explore the duration as well as thepropensityofbrokenpegstoreform.

3.Apegisdefinedhereasacountrythatstayswithina2percentagepointupordownrangeoverthecourseofayear(excludingsingle-yearpegs,whichmaybeduetoalackofvolatility)orthosethatareperfectlyflat,withtheexceptionofa

one-time realignment. See Klein and Shambaugh (2010) for an extensivediscussionofexchangerateregimeclassifications.

4.SeeObstfeld(1997)andObstfeldandPeri(1998)fordiscussion.

5.SeeObstfeld(1997)fordiscussion.

References

Kang, Joong Shik, and Jay C.Shambaugh. 2013. "The Evolution of CurrentAccount Deficits in the GIPS and the Baltics: Many Paths to the SameEndpoint."IMFWorkingPaper,InternationalMonetaryFund,Washington,DC.

Keynes,JohnM.1925.TheEconomicConsequencesofMr.Churchill.London:HogarthPress.

Klein, Michael, and Nancy P.Marion. 1997. "Explaining the Duration ofExchange-RatePegs."JournalofDevelopmentEconomics54(2):387-404.

Klein,MichaelW.,andJayC.Shambaugh.2008. "TheDynamicsofExchangeRateRegimes:Fixes,Floats,andFlips."JournalofInternationalEconomics75(1):70-92.

Klein,MichaelW.,andJayC.Shambaugh.2010.ExchangeRateRegimesintheModernEra.Cambridge,MA:MITPress.

Lane,Philip,andBarbaraPels.2012."CurrentAccountImbalancesinEurope."Centre for Economic Policy Research Discussion Paper DP8958, Centre forEconomic Policy Research, London. http://papers.ssrn.com/ sol3/papers.cfm?abstractid=2066331.

Mundell,RobertA.1961. "ATheoryofOptimumCurrencyAreas."AmericanEconomicReview51(4):657-665.

Obstfeld,Maurice. 1997. "Europe's Gamble." Brookings Papers on EconomicActivity2:241-317.

Obstfeld, Maurice, and Giovanni Peri. 1998. "Regional Non-adjustment andFiscalPolicy."EconomicPolicy13(26):205-248.

Obstfeld,Maurice,andKennethRogoff.1995."TheMirageofFixedExchangeRates."JournalofEconomicPerspectives9(Fall):73-96.

Shambaugh, Jay. 2012. "The Euro's Three Crises." Brookings Papers onEconomicActivity,Spring,157-211.

MartinWolf

The effort to bind states together may lead, instead, to a huge increase infrictions among them. If so, the event would meet the classical definition oftragedy:hubris(arrogance);ate(folly);nemesis(destruction).

-MartinWolf,FinancialTimes,December1991

The creation of the euro was among the most revolutionary developments inmonetary history. Advanced European economies agreed to replace theirnational monies with a shared fiat currency, managed by a jointly ownedinstitution,theEuropeanCentralBank(ECB).Theydidthis,moreover,withoutagreeing to anyof theother featuresof contemporarymonetary areas, notablymechanisms for fiscal transfers or for common regulation and support of thebanking system. In all these respects, the governments of member statesremainedsovereign, ifnotionallyconstrainedbyasetof rulesgoverningfiscaldeficitsanddebt.

A possible justification for this extremely limited institutional infrastructurewasthattherulesconcerningfiscalpolicyandthecentralbank'sabilitytoactaslender of last resort in a crisis would together be enough to ensure adequatestability.Theywouldeitherpreventcrisesor,iftheyfailedtodoso,makethematleastreasonablymanageable.Anotherpossiblejustificationwasthebeliefthatitwasessentialatleasttostart.Oncetheeuroareawaslaunched,anyfailuretopreventormanageseverecriseswouldmotivatepolicymakerstocreatemissinginstitutionsorimproveexistingones.

Byearly2013,afteratleastthreeyearsofcrisis,theinstitutionalframeworkof

the euro areahadbeen shown tobe inadequate.The fact that the crisis forcedrapid institutional andpolicy innovationsproves that.Whathadexistedbeforethe crisis proved inadequate, but whether a severe crisis would produce thereformsneededtomaketheeuroareabetterabletocoperemainsunclear.

To work out what reforms are needed, wemust start by asking what wentwrong.ThisisatopiconwhichPaulDeGrauwe,formerlyattheUniversityofLeuvenandnowattheLondonSchoolofEconomics,hasshedmuchlightinanumberof important and illuminatingnotes andpapers.His conclusion is thattheeuroareasimplyneedsagreatdealofreform,particularlyinthepoliciesofthe central bank. My conclusion is that the euro was a bad idea. Bothconclusionsmightbecorrect.

Certainly, the loss of sovereignty for the governments ofmember states hasimposedlargecostsonthemandtheircitizens.Anexcellentwaytoshowthisisvia thecontrastingexperiencesofSpainand theUnitedKingdomin thecrisis.Spainlacksthetoolstohandlesuchabigfinancialcrisiswithanyease.TheUKdoesnot lack those tools, though it has failed touse themas fully as itmighthave.

TheContrastingCasesofSpainandtheUnitedKingdom

SpainandtheUKarebothcrisis-hitcountries.Sincethecrisis,bothhavebeeninpoor fiscal positions.They alsohavebigproblems in their banking industries,thoughtheyarenotquitethesameproblem:Spain'sislargelydebtcreatedbyahugedomesticpropertyboom; theUKbankshave thatproblem,but theyhavealsobeendamagedby their globaloperations.Surprisingly, perhaps, the fiscalconsequencesoftheirdistinctcrisesareremarkablysimilar,asshowninfigure20.1. The expected path of the ratio of net public debt to GDP in these twocountriesisalmostidentical.

Forthosewhothinkthemaindeterminantof theinterestrateongovernmentdebt is the public debt burden, the implication is clear: the interest rates onSpanishandUKpublicdebtshouldbequitesimilar.Buttheyarenot.Theyieldon10-yearUKpublicdebtisfarmorelikethatonGerman10-yearpublicdebt

thanlikethatonSpanishdebt,asfigure20.2shows,eventhoughGermany'sdebtisexpectedtobeunderfarbettercontrolthantheUK's.ThedivergencebetweentheyieldsonSpanishandUKdebthasbeenverylargeindeed.Thishasmadeitfar more difficult for the Spanish government to manage its debt and hasadversely affected broader credit andmonetary conditions in Spain relative tothoseintheUK.

Figure20.1

RatioofNetPublicDebttoGDP,Spain,UK,andGermany.

Source:InternationalMonetaryFund,WorldEconomicOutlookdatabase,April2013.

Figure20.2

Yieldson10-YearGovernmentBonds,Spain,UK,andGermany.

Source:ThomsonReutersDatastream.

Whyshould the interestratedifferencebetweenthe twocountrieshavebeenso very large? The answer liesmainly in the fact that the UK is a sovereigncountry,withitsownfinanceministry,centralbank,andfloatingcurrency,whileSpain is a subordinate government inside a currency union that has no sharedtreasuryandasupranationalcentralbank(theECB).

Supposeholdersofagovernment'sdebtbelievethatitmightbeunabletorollthe debt over on reasonable terms. They would rationally fear an outright-possibly sudden-default. Creditors cannot seize the assets of bankruptgovernmentsas theycan takeholdof theassetsofcompanies.This isbecausenational governments are sovereign in their own jurisdictions. So lenderswilldemand an interest rate that protects them against default risk.But at such anelevatedinterestrate,thegovernmentmaybedrivenintothedefaultthatlendersfear,makingtheirprophecyofdoomself-fulfilling.

This is the danger ofmultiple equilibria. Olivier Blanchard, IMF economiccounselor,putsitthisway:

Athighlevelsofdebt,theremaywellbetwoequilibria,a"goodequilibrium"atwhich rates are lowanddebt is sustainable, and a "bad equilibrium" inwhichrates are high, and, as a result, the interest burden is higher, and, in turn, theprobabilityofdefaultishigher.Whendebtisveryhigh,itmaynottakemuchofachangeofheartbyinvestorstomovefromthegoodtothebadequilibrium.2

Preventingsuchashiftisoneofthejobsofcentralbanks.Thus,acentralbankguarantees liquidity in themarket for sovereign debt.That hugely reduces theriskofasuddendefault.Thisreductionofliquidityriskincreasesconfidenceinlenders.Asusual,liquidityandsolvencyrisksarecloselyrelated.

The principal reasonwhy interest rates in Spain have been somuch higherthanthoseoftheUKisthatSpainhadnosuchlenderoflastresort.Spanishdebtwas subject to liquidity risk, and so, when liquidity risk looked significant,marketspricedthedebtaccordingly,pushingSpainintoabadequilibrium.TheECBwasnotbelievedtobewillingorabletoensureliquidityinthemarketsforsovereign debt of the euro area. In a panic, then, everybody fled to the safestdebt,Germany's,causingabigcrisisincountrieswithworsedebtpositions.

LessonsfromECBInterventions

TheplausibilityoftheviewthatthebiggestproblemfacingSpainhasbeenthelack of a central bank of its own is strengthened bywhat happenedwhen theECBdidfinallyindicateitswillingnesstointerveneinthemarketforpublicdebtofcountriesindifficulty.ThedeclineinyieldsonSpanishdebtshowninfigure20.2datesalmostprecisely toJuly26,2012.This iswhen thepresidentof theECB,MarioDraghi, said to an audience inLondon, "Withinourmandate, theECBisreadytodowhateverittakestopreservetheeuro.Andbelieveme,itwillbe enough."3 This statement in turn led to the announcement by the ECB onAugust 2 of its Outright Monetary Transactions (OMT) program aimed "atsafeguardinganappropriatemonetarypolicytransmissionandthesinglenessofthemonetarypolicy."4Rightlyorwrongly,marketsconcludedthattheriskofasuddenoutright default onSpanishbondshadgreatlydiminished.This in turnpushedthepriceofbondsfromabadequilibriumtoabetterone.Asyieldsfell,the government did indeed start to look more solvent, thereby justifying the

markets'renewedoptimism.

Not coincidentally, the decline from the previous interest rate peak, in late2011, dates from the announcement by the ECB of its three-year Long-TermRefinancing Operation in early December 2011.5 But that operation provedunsuccessfulinkeepinginterestratesdown.ThatiswhytheECBwasdriventoadopt the OMT program in the teeth of opposition from Jens Weidmann,presidentoftheBundesbank.'

Moreover,thesamedeclinesoccurredinItalyasinSpain,stronglysupportingthe argument that it is ECB policy rather than actions by governments thatexplainsthesharpdeclineininterestratesonthelong-termgovernmentbondsofvulnerablecountries.7TheabilityoftheECBtotriggersuchadeclineinyieldsis exactly what De Grauwe predicted. Now that it has become fact, he hasanalyzedthisadjustmentinanotherimportantarticle.Thepointhemakesisthatthese were largely self-fulfilling panics that the ECB has, for the moment,ended.'

The crises, then,were in large part the result of allowing government bondmarkets to operate without grown-up supervision. Fortunately, the grown-upsareback.Thatisgoodnewsfortheeuroareaandtheworld.Ahugeamountofunnecessary and ill-timed fiscal austerity has been imposed, just because theeuroareadidnothaveapropercentralbank.Itnowhassomethingthatisatleastabitmorelikeapropercentralbank.Itshows.

WhyECBInterventionDidNotEliminatetheSpanishRiskPremium

However,theinterventionoftheECB,pleasinglyeffectivethoughithasbeeninloweringrates,hasnotreducedSpanishinterestratestoUKlevels,atleastatthetime of this writing. Why should that be? One can see two possible sets ofexplanations.

First,theECB'sOMTprogramoperatesunderimportantlimitations.Themostimportantoneisthat theprogramisnotunconditional, thoughinprincipleit ispotentiallyunlimited.Withoutconditionality, theECBcouldnothaveobtainedinternalapprovalorexternalconsent,aboveallfromtheGermangovernment,to

intervene.Thus,theECBstatedonSeptember2,2012:

AnecessaryconditionforOutrightMonetaryTransactionsisstrictandeffectiveconditionality attached to an appropriateEuropeanFinancial Stability Facility/EuropeanStabilityMechanism(EFSF/ESM)programme.SuchprogrammescantaketheformofafullEFSF/ESMmacroeconomicadjustmentprogrammeoraprecautionaryprogramme(EnhancedConditionsCreditLine),providedthattheyinclude the possibility of EFSF/ESM primary market purchases. Theinvolvement of the IMF shall also be sought for the design of the country-specificconditionalityandthemonitoringofsuchaprogramme.9

Another restriction is that the explicit rationale of the program is not tosupport government debt markets but rather to make monetary policy workeffectively. This rationale is ingenious, since it allows the ECB to claim,imaginatively, that theaimof theprogramismonetaryandsowithin itsbroadremit,ratherthanfiscalandsooutsidethatremit.Thus,theECBannounced:

The Governing Council will consider Outright Monetary Transactions to theextent that they are warranted from amonetary policy perspective as long asprogramme conditionality is fully respected, and terminate them once theirobjectives are achieved or when there is non-compliance with themacroeconomicadjustmentorprecautionaryprogramme.

Asaresult,theECBstated,"Transactionswillbefocusedontheshorterpartof the yield curve, and in particular on sovereign bonds with a maturity ofbetweenoneandthreeyears."Thismakessensefornormalmonetarypolicy,butthe restriction limits the commitment of the ECB to support the market insovereigndebt.

Asecondsetofexplanations for the failure toachieveconvergenceof long-terminterest ratesbetweenSpainand theUKis that the formersuffers fromanumberofhandicapsthatthelatterdoesnot.Theseincludetheriskofexitfromtheeuroareaorofbreakupoftheeuroarea,deflationrisk,andothereconomicdifferences.

ItisimpossiblefortheECB,oranyinstitution,toeliminatetheriskthatSpainmightleavetheeuroareaorthattheeuroareamightitselfbreakup.Solongas

thatriskcontinuestoexist,investorsinSpanishbondsneedtotakeoutinsuranceagainst the possibility of a sudden and costly redenomination into bonds in anew currency thatwould then swiftly depreciate. Furthermore, in the event ofsucharedenomination,itislikelythatexchangecontrolswouldalsobeimposed,whichcreatesanotherriskforinvestorsinSpanishgovernmentbonds.

Furthermore,iftheeuroareawerenottobreakuporifSpainwerenottoexit,Spainwouldremainvulnerabletodeflationrisk.Indeed,outrightdeflationisthemechanism through which external competitiveness is restored inside thecurrencyunion.ButdeflationwouldraisetherealvalueofSpanishdebt,makingitsdebtlesssustainable.Ifthedeflationwerebigenough,thetimeprofileofdebtmight end up substantially worse than shown in figure 20.1. This would beparticularlytrueifthedeflationaryprocessalsoinflictedadeeperthanexpecteddepression,depressingthedenominatorstillfurther.

Finally,itseemsthatSpain'sinitialdisequilibriumwasratherlargerthanthatof the UK. Its current account deficit was 10 percent of GDP in 2007, forexample, against 2 percent in the UK. Consequently, the adjustment Spainneeded would have seemed far bigger. Again, Spain's net internationalinvestmentpositionwassubstantiallymorenegativethanthatoftheUK,makingthe country more dependent on foreign investors, who are usually, for goodreason,more fearfulofdefault thandomesticonesare.Spainalsohada largerboom in construction than theUK. For all these reasons, Spainwas likely tosuffer a longer and deeper recession than the UK, as indeed has happened.Investorsmightreasonablysupposethatthegovernmentofacountryundergoingsuchadeepandintractableslumpmightnotmakemeetingitsdebtobligationsapriority. Inall, investorsmight reasonablyreach theconclusion thatSpainwasnotasgoodaninvestmentriskastheUK.

Theratingagenciesseemtohavereachedthatconclusion.Asofthiswriting,in June 2013, Standard&Poor's rated theUK's sovereign bonds atAAA andSpain'satBBB-.Moody'sratedthematAalandBaa3,respectively.Fitch'sratedthematAA+andBBB.Thishugegapbetweenthetwocountries'bondratingsmaypartlyreflecttheadditionalandimportantfactthattheUKhashadalongerrecordofmanagingitsdebtwellthanhasSpain.Itmayalsoreflectthenormal

behavior of the rating agencies: "I am your rating agency, therefore I followyou" seems tobe their long-standingmottowith regard to themarket.But thebig fact is that the rating agencies downgraded Spain's sovereign bondsmassivelycomparedwiththeUK's.

Meanwhile, theUKpossessed fundamental countervailing advantages. First,adjustment to a shift in the desire to hold sterling-denominated liabilities hasworked, at least in part, via the price of the currency rather than the price ofbonds. Such price flexibility reduces the need for quantity adjustments inresponsetoshiftsinthedesiretoholdacountry'sliabilities.InSpain,instead,alargerquantitativeadjustmenthasbeen required.This is shown in the scaleofthe recession, revealed in figure 20.3, and the size of the current accountadjustment,showninfigure20.4.Fromonepointofview,Spain'sadjustmentisimpressive.ButitisalsoapowerfulindicatorofthecollapseinSpain'sdomesticabsorptioncomparedwiththeUK's.

Figure20.3

GDPintheCrisis,SpainandUK.

Source:InternationalMonetaryFund,WorldEconomicOutlookDatabaseApril2013.

Figure20.4

CurrentAccountBalances(asashareofGDP).

Source:InternationalMonetaryFund,WorldEconomicOutlookDatabaseApril2013.

Inaddition,theUKhascaptivesaverswhoneedtomatchthecurrencyoftheirassets with those of their liabilities. Thus, in a crisis, the UK government'sliabilitiesprovideasafehaventosuchinvestors,amongwhichwillbeinsurersandpensionfunds.Intheeuroarea,however,therelevantsafehavenisthedebtof the German government in particular, and to a lesser extent that of otherstable creditor countries, such as the Netherlands. Thus, in a crisis, fearfuldomestic saverswill fly towardUK government debt and away fromSpanishgovernmentdebt.

WhytheDisadvantagesofaCurrencyAreaArePartlyUnavoidable

NearlyallthesedifferencesbetweentheUKandSpain,fromthelackofapropercentralbanktothescaleoftheeconomicdisequilibria,derivefromthefactthatSpain is a member of the euro area while the UK is not.Membership in thecurrencyareahasprovedtobeabigdisadvantageinhandlingaseverefinancialcrisis.Acrucialquestioniswhether,withadifferentinstitutionalstructure,these

disadvantagesmighthavebeenorcanbeavoided.

To some extent, the answer must be no. The advantages possessed by acountrywithitsownfloatingcurrencyandcentralbankarelarge,atleastwhenhandlingthecontractionaryconsequencesofahugefinancialcrisis.

Remember,too,thattheaimofcreatingtheeuroareawastopromoteinternalcapitalflows.Thus,thehugecross-borderflowsofprivatefinancethatpreceded(and then triggered) the crisiswere hardly some accidental bug in the system.They were a deliberate feature. Similarly, the euro area was designed, quitedeliberately, around the idea of a goal-independent central bank required tofocus on price stability and expected not to finance governments directly,whateverhappened.Moreover,inamulticurrencyunion,theonlyalternativetosuch a highly independent central bank would have been subordination to acommittee of financeministers.While such an arrangement is conceivable, itwould have been extraordinarily ponderous and, in any case, unacceptable toGermany and a number of othermember countries. Sowe have to regard theindependenceof the central bankand thenatureof itsmandate as a feature atleastofthiscurrencyunion.Asaresult,theECBcannotmakeanopenendedandunconditional commitment to purchase public debt. A limited or conditionalcommitment cannot be completely credible, and an incompletely crediblecommitmentby thecentralbankwillnot shift adverseexpectationsdurablyorcompletely.

Of course, some of these features of the euro area were not logicallynecessary. Germany might, for example, have had a different philosophy ofmonetarypolicyandmacroeconomicmanagement.Itmighthavebeenmorelikethat of the United States. This would have made it easier for the ECB tointerveneinsovereigndebtmarkets,asDeGrauwewanted,onalargescale.Thedifficultiesexperiencedinthecurrencyunionwouldthenhavebeensmaller.Butthe facts that creditor countrieswould like to restrict the support theyoffer todebtor countries and that countries in trouble face exceptionally difficultproblems in managing the crisis and subsequent adjustment are inherent inalmostanyconceivablecurrencyunion.

WhytheEuroWasaBadIdea

What is the conclusion? It is that therewere huge risks in the creation of theeuro,someinescapableandsomeinherentinitsdesign.Thenatureandextentoftheseriskshavebeenrevealedinthecrisis.

Whywerepeopleunconcernedabouttheseriskspriortothecrisis?Abigpartof the reason is thatmanypeoplebelieved thatcurrencyriskwas theprincipalsourceofcrises.Thisviewwascertainlyconsistentwithexperienceinthe1960s,1970s,1980s,and1990s.Consequently,proponentsoftheeurothoughtthattheeliminationofseparatecurrencieswouldeliminatemostoftherisksofcrises.

Events have proved this view false. Indeed, since 2010, euro areapolicymakers and economists have discovered that the opposite is true. First,currency risk cannot be eliminated, since there is always the possibility thatcurrencies might be recreated. Second, currency risk will reemerge in otherways,particularlyintheformofshockstothesupplyofcreditandfinancialandfiscal crises. Finally, when such risks become real, they will inflict hugefinancial,economic,social,andpoliticalpain.

Thebiglessonthecrisishastaughtusisthathigh-incomecountriesembeddedinacurrencyunionaremorevulnerable tobalance-of-paymentscumfinancialcrises thansimilarcountrieswithfloatingexchangeratesand theirowncentralbanks. The currency union has, in fact, replaced the brief currency crises andexchangeraterealignmentsoftheoldexchangeratemechanismwithwhatnowappear to be long-running solvency, employment, and political crises.Amoreactivecentralbank,onewilling topush sovereigndebt towardgoodequilibriaand away frombadones,would be a big help. In the course of the crisis, theECBhasbecomemoreof suchabank.But for reasons largely inherent in thecreationofanycurrencyunionandcertainlyinherentintheeuroarea,theECBis not going to act like a national central bank. The question formembers todecideiswhetherthestressestheysufferasaresultareworthit.

Notes

1.SeePaulDeGrauwe,"TheGovernanceofaFragileEurozone,"Universityof

Leuven, April 2011, http://www.econ.kuleuven. be/ew/academic/intecon/Degrauwe/PDG-papers/Discussion_papers/ Governance-fragile-eurozone_s.pdf;idem, "Managing a FragileEurozone,"Vox,May 10, 2011, http://www.voxeu.org/article/ managing-fragile-eurozone; Paul De Grauwe and Yuemei Ji,"MispricingofSovereignRiskandMultipleEquilibriaintheEurozone,"CEPSWorking Document, January 20, 2012, http://www.ceps.eu/ book/mispricing-sovereignrisk-and-multiple-equilibria-eurozone; and idem, "Mispricing ofSovereign Risk and Multiple Equilibria in the Eurozone," Vox, January 23,2012, http://www .voxeu.org/article/ mispricing-sovereign-risk-and-multiple-equilibria-eurozone.

2. Olivier Blanchard, "Rethinking Macroeconomic Policy," blogpost, IMFDirect, April 29, 2013, http://blog-imfdirect.imf.org/ 2013/04/29/rethinking -macroeconomic-policy.

3.MarioDraghi,presidentoftheEuropeanCentralBank,speechdeliveredattheGlobal Investment Conference, London, July 26, 2012, http://www.ecb.int/press/key/date/2012/html/spl2O726.en.html.

4. European Central Bank, "Technical Features of Outright MonetaryTransactions," press release, European Central Bank, Brussels, September 6,2012,http://www.ecb.int/press/pr/date/2012/html/prl2O9O6-l.en.html.

5.EuropeanCentralBank,"ECBAnnouncesMeasurestoSupportBankLendingandMoneyMarketActivity," press release, EuropeanCentral Bank,Brussels,December 8, 2011, http://www.ecb.europa.eu/press/ pr/date/2011/html/prl11208_1.en.html.

6. Michael Steen, "Weidmann Isolated as ECB Plan Approved," FinancialTimes,September7,2012.

7. See Joe Wiesenthal, "In One Chart, Here's Why Roger Altman Is WrongaboutHowtheMarketsForcedAusterityonEurope,"Businesslnsider.com,May2013, http://www.businessinsider.com/ in-one-chart-heres-why-roger-altman-is-wrong -about-how-the-markets-forced-austerity-on-europe-2013-5; and PaulKrugman,"Allabout theECB,"blogpost,NewYorkTimes.com,May10,2013,http://krugman.blogs.nytimes.com/2013/05/10/all-about-the-ecb.

8.SeePaulDeGrauweandYuemeiJi,"MoreEvidenceThatFinancialMarketsImposedExcessiveAusterityintheEurozone,"CEPSCommentary,February5,2013, http://www.ceps.eu/book/ more-evidence-financial-markets-imposed -excessive-austerity-eurozone; idem, "Panic-Driven Austerity in the Eurozoneand Its Implications," Vox, February 21, 2013, http://www.voxeu.org/article/panic-driven-austerity-eurozone-and-its-implications.

9. European Central Bank, "Technical Features of Outright MonetaryTransactions," press release, European Central Bank, Brussels, September 6,2012;http://www.ecb.int/press/pr/date/2012/htmt/prl2O9O6-l.en.html.

GangYi

ChinaasanOptimalCurrencyArea

China is a large country with tremendous differences across its land. Thediversitiesbetweeneasternandwesternregionstosomeextentexceedthoseoftheeurozone.Forinstance,intheeasternpartofChina(suchasthePearlRiverDelta),thepercapitaincomeisaboutfivetimesthatofthewesternpart(suchasGuizhou Province and Qinghai Province), whereas inside the euro area,Germany'spercapita income isonlyabout1.5 times thatofGreece.However,even with these differences, China is still better positioned as an optimalcurrency area than the euro area and can issue a commoncurrency, forChinasatisfiesseveralkeyelementsthatarerequiredbyanoptimalcurrencyarea.

Thefirstoneconcernsfiscalsetting.China,asasovereigncountry,possessesa centralized fiscal transfer payment system that enables the transfer of fundsfromtherelativelyrichregionsinthecoastalareatothepoorregionsinitsless-developedmiddleandwestern regionseveryyear.Second,Chinahasahighlymobilelabormarket.Manypeoplefromthelow-incomeregionsseekbetterjobopportunities in high-income regions; consequently, at least 160 million ruralworkers are now working in China's cities. Third, China has a very unifieddomesticmarket thatsharesan integratedsystemof tradingand transportation.Domesticmarkets,suchasthecommoditymarketandtheproductmarket,areallhomogeneous. Fourth, China's macroeconomic policies are very effective andcoordinated. Those are some of the reasons why China can be viewed as anoptimalcurrencyarea,eventhoughtheregionaldifferencesbetweentherichandthe poor, in terms of per capita income, are still very large.As long as fiscaltransfer andmacroeconomic policies workwell and the labormarket remainshighlymobile,Chinaasanoptimalcurrencyareawilloperatesmoothlyandstay

sustainable,thoughthediversitywithinChinawilllikelyremain.

TheotherreasonI'dliketomentionistheabilityoftheChinesegovernmenttoaddress financial systemic risks. For example, I was involved in tacklingfinancial crises as a key representative of the People's Bank of China. Localfinancialcriseswerebreakingoutinprovincesorregions,includingGuangdongProvince and then in Hunan Province in the late 1990s, as well as XinjiangUygurAutonomousRegionandNingxiaHuiAutonomousRegion.

Guangdong isadevelopedareawithmanypeoplewhoarevery rich.But inthe late 1990s, Guangdong experienced major fiscal insolvency and financialinstitutionproblems.Otherregions,suchasNingxiaandXinjiang,althoughlessdeveloped,alsoexperiencedfinancialdifficulties.Whenaddressingthesekindsof problems, we usually have to decide first whether the central bank shouldintervene by providing liquidity as the lender of last resort.We also need tofigureouthow tohandle localgovernmentdebt anddealwith the relationshipbetween local governments and the central government.We have to design aframework that not only can avoidmoral hazard but also can calm the crisissituation. In thisway, thecapabilitiesof thePeople'sBankofChinaandotherfinancial supervisors to identify problems and address potential systemic riskshave been tested and proven, and they solidified the ground on which thecommoncurrencyisestablished.

China'sPolicyChoicesoverExchangeRateRegimes

There has been an ongoing theoretical debate in the history of exchange rateregimes since the beginning of the 1970s, when the Bretton Woods systemcollapsed and the exchange rate regimes of many countries kept evolving. Inreality,exchangeratearrangementsdonotfallintoasimpledichotomyoffixedorfloatingexchangerates.

According to the IMF, there are eight categories of exchange ratearrangements, and the IMF's view regarding exchange rate arrangements hasalso evolved over time. During the Bretton Woods era, the IMF was of theopinionthatfixedexchangeratearrangementswerebetter.BeforetheArgentine

crisis, it argued for the bipolar view (either fixed or floating), while after thecrisisitcalledforfloatingexchangerates.After2009theIMFtooktheviewthatallexchangeratearrangementshadmeritsandshortcomings.

The floating exchange rate arrangement has several merits. First, using afloating exchange rate makes it easy to adjust to economic shocks, and thearrangementitselfislessvulnerabletospeculativeattacks.Italsohelpstoavoidthe resourcemisallocation caused by exchange rate distortions. But a floatingexchange rate also has its problems.For example, itmay cause uncertainty orexcess fluctuation or overshooting, and it leads to a larger exchange rate risk.Fixedexchangeratearrangementsarefavorablebecausetheygiverisetolowerinflation (for developing countries) and facilitate more investment and trade.However, they also result in exchange rate distortions and are vulnerable tocurrencycrisisandspeculativecapitalflows.

HereIwillbrieflydiscussChina'sprogressinexchangerateregimereforms.In1994,Chinaunifiedthedual-trackexchangerateregime,whichusedtooffertwo different foreign exchange prices of the renminbi. Since then, China hasbeenimplementingamanagedfloatingexchangerateregime,andreformswerepursued in a "minibang" approach rather than a "big bang," which perfectlysuited China's specific circumstances. From 1994 through 2002, the realeffectiveexchangerate(REER)oftherenminbi,whichisregularlycompiledbytheBankof InternationalSettlements, increasedvery rapidly togetherwith theUSdollar,towhichtherenminbiwaspegged.Between2002and2005,whentheUSdollarweakenedinvalue, therenminbistartedtodepreciatealongwith thedollar,whichcausedsomeproblemsanddisputes.

In 2005, the preconditions for foreign exchange mechanism reform werebroadly inplace.Forexample, thefinancial reformswerewellunderway,andmajorstate-ownedbankswereoverhauledandlisted(orweretobelisted).Theforeign exchange market was developed to a better level, and China's capitalaccountwasgraduallyopened.Mostimportantly,Chinawasexperiencingstablegrowthandlowinflation,whichprovidedarelativelyfavorablebackgroundforreform.Therefore,whentheopportunetimearrived,Chinadecidedtodepegtherenminbiagainst theUSdollar,and themore flexible renminbiexchange rates

followedshortlyafter.From the2005 reform to theendofFebruary2013, therenminbinominalexchangeratetotheUSdollarappreciatedabout32percent,andtheREERoftherenminbiappreciatedbymorethan36percent.Thetrendoftherenminbiexchangeratehaschangedfromunilateralappreciationtotwo-wayfloating,andthepricenowfallsintoabroadequilibriumrange.

However, the reform process was interrupted by the global financial crisis,whenwehad tonarrow thedaily floatingbandof the renminbibecauseof thesituation in the post-LehmanBrothers period. That said, the renminbi did notdepreciateasmuchas thecurrenciesofotheremergingmarketeconomiesdid.But the reformprocesswas not suspended for long.On June 19, 2010,Chinadecided to further promote themarket-based exchange ratemechanism reformwithreferencetoacurrencybasket,emphasizingthefundamentalroleofmarketsupplyanddemandandtheflexibilityoftherenminbi.OnApril16,2012,Chinawidenedtherenminbidailyfloatingbandto±1percentfrom±0.5percent.Alsosince 2012, the People's Bank of China has significantly reduced marketintervention.Todate,marketforcesareplayingadominantroleintheformationoftherenminbiexchangerate.China'sexchangerateregimewithoutadoubtisbecomingmoreandmoremarket-oriented.

Nevertheless, a floating exchange rate regime does not rule out foreignexchangemarket intervention. In somecircumstances, intervention is stillwelljustified.Forexample,whentheexchangerateexceedsthepredeterminedband,orwhenthecapitalaccountexperienceslargeimbalancesandthereareexcessivetrades in the foreign exchangemarket, orwhen the financialmarket falls intocrisis-scale turmoil,market intervention can be restarted to the extent that theintervention is done in a two-way manner, with the intent of preventing orcorrectingalarge,short-termfluctuationoftheexchangerate.

Asa resultofmore flexibleexchange rates, inbothnominal and real terms,China'scurrentaccountsurplusasapercentageofGDPexperiencedasignificantdownward adjustment. The ratio of China's current account surplus to GDPpeaked in 2007 and 2008, but declined rapidly thereafter. By 2011 it haddeclinedto1.9percent,andin2012theratiowas2.3percent.

The direction of renminbi exchange rate reform is irreversible. China willcontinuetoenhanceexchangerateflexibility,andinthenearfuture,Chinawillincrease the floating band of the renminbi even further, lettingmarket supplyanddemandplayafundamentalroleinexchangerateformation.Chinawillalsoreduce central bank intervention, enhance the self-rebalancing capability offoreign exchange market, and improve the self-pricing and risk-managementcapabilitiesofthefinancialinstitutions.

ExchangeRateandRealEconomyAdjustments

The exchange rate and the real economy are interlinked naturally. When theexchangerateadjusts,therealeconomywillbeinfluencedbytheexchangerateandreflecttheexchangeratechangesinitsownadjustment.Chinahasachievedgoodperformance in thefirst threeof fourmacroeconomicobjectives,namely,growth, employment, and inflation.Regarding the fourth objective, balance ofpayments, Chinawent through a periodwhen its current account surplus wasrelativelylarge,butinrecentyearsthesurplushasstartedtoconvergetonormallevels.

Thequestionhereis,howcanweputthisadjustmentonasustainabletrack?Itisveryimportanttoemphasizethatpursuingamoreorlessbalancedbalanceofpayments account is our goal. And this process of adjustment, which startedyears ago, will continue in the future. The adjustment first started in salary;therefore,wecanfindasteadyincreaseinlaborcosts.Recently,especiallyinthepastfiveyears,laborcostshaveincreasedevenmorerapidly.

Since the 1990s, labor costs in China have increased more slowly thanproductivityincreases.Overthepastfiveyears, laborcostshaveincreasedataslightly greater rate than productivity, which implies that China'scompetitivenesshasbeendecreasing. Inmyview, this isahealthyadjustment,forithelpsbringupconsumption.Itcanalsobeviewedasacatchupeffect.Thisprocesswillcontinueaslongasfirmsremainprofitable,andIamoftheopinionthattheywill.

Thesecondadjustmentalsoconcernslaborcosts,butnotintermsofsalary.I

refertothesocialsecuritysystem:pensions,medicalinsurance,otherinsurancemechanisms,andsoforth.Inrecentyearsthesocialsecuritysystemforworkershas been improved further, andwe have paid special attention to designing anationwidesocialsecuritysystemthatcoversnotonlycityworkersbutalsoruralworkers, that is,migrantworkers from the rural areas.At themoment,wearestriving to furtherenhance thenationwidesocial securitysystemand toensurethat migrant workers can enjoy their rights as citizens with regard to socialsecuritybenefits.Ifwecanmakethissystemwork,thisportionofthecostswillsurelyincrease,leadingtohigherlaborcosts.

Thethirdsourceofrapidcostincreasesisenvironmentalcosts,whichreferstoairpollution,soilpollution,andwaterpollution,allofwhichhavebeenaseriousprobleminChinarecently.TheChinesegovernmentandthepeoplerealizethatthe current situation is not sustainable, and their investment in environmentalprotection (which, in other words, relates to cost) has been increasingtremendously.Improvedenvironmentalpracticesinindustryandmanufacturingare also expected to decrease the competitiveness of China's products. In thefuture,itispossiblethatChina'senergypricesandthepricesofotherresources(such as water and utilities) will continue to increase until they reacheconomicallysustainablelevels.AlloftheseadjustmentsthatIhavementionedarehealthy,buttheywillundoubtedlyaffectChina'scompetitiveness.

Whatisthebottomline?Inmyopinion,aslongaswecanmaintainamoreorless balanced balance of payments account, this adjustment process cancontinue,whichishealthyforChinaandwillalsocontributetotherebalancingof the world economy. As far as I can see, this process will continue goingforward, and through channels such as prices it will adjust the economy andachieve what we have been pursuing for years, that is, a relatively balancedbalance of payments account. China has become a major importing country.Right now, the annual import volumeofChina is aboutUS$2 trillion. If thistrend continues, a modest estimate suggests that the lower bound of importgrowthrateforChinawillbeabout6percentannually.Andby theyear2020,China's imports will be around US $3 trillion, which is helpful to balancingChina'scurrentaccountposition.

DuvvuriSubbarao

IntellectualShiftonCapitalControls

Thechangeinourworldviewoncapitalaccountmanagementisbyfaroneofthemost remarkable intellectual shifts brought on by the crisis. In her openingremarksat theconference, IMFManagingDirectorChristineLagardesaid thatthe crisis shattered the consensus on many macroeconomic issues andshibboleths.Nowhere is thismore true than in thebroadpolicyareaofcapitalaccountmanagement. Inmy view, the three big issues onwhich the precrisisconsensushasdissolvedarethefollowing:first,movementtowardafullyopencapital account; second, the use of capital controls as short-run stabilizationtools;andthird,thedesirabilityofforeignexchangeintervention.Iwillcommentbrieflyoneachofthese.

MovementtowardaFullyOpenCapitalAccount

The first issue on which consensus was broken is the need for a fully opencapitalaccount.Before thecrisis, theconsensuswas thateverycountryshouldeventuallymovetowardafullyopencapitalaccount.Thedebatewasonlyaboutthe appropriate strategy-in particular, about sequencing and timing-fortransitioningtofullcapitalaccountconvertibility.

ChinaandIndia

Let me invoke the example of India. Moving toward full capital accountconvertibilityhasalwaysbeenourpolicygoal.Theonlyvariablewas the roadmap for getting there,which, itwas agreed, should be redefined from time totime, consistentwith the evolving situation.Therewas alsogeneral agreementthatweshouldstartbyfloatingtheexchangerateanddecontrollinginterestrates

and finishwith the capital account,with the rationale that this strategywouldbestpreservemacrostability.

There has been a long and vigorous debate inChina too on opening up thecapitalaccount,witharoughlysimilarconsensusasinIndiaaboutsequencing.Overthepastfewyears,though,Chinahasapparentlychangeditsstrategy,asisevident from the country's policy direction. If you accept that measures tointernationalizetherenminbiareabigsteptowardcapitalaccountconvertibility,thenthisinitiativebyChinahasbeenmuchbolderthanitsactionsonfreeingupexchangeandinterestrates.

ControlsandFinancialStability

Thecrisishas,however,changedallthis.Itshiftedthedebatefromthestrategyandtimingforcapitalaccountconvertibility toquestioningtheveryimperativefor capital account convertibility. In other words, the consensus that everycountry should eventually move toward a fully free capital account is nowbroken.

The main argument in support of the new view-that full capital accountconvertibilityneednotbeaneventualgoal-is thatcontrolspreventedemergingmarkets from adopting some of the financial products that proved toxic inadvancedcountries.So,thereismerit,itisargued,inretainingcapitalcontrols.Againstthisistheoldargument,whichisstillquitepersuasive,thatascountriesbecomemoreintegratedeconomically,theywillneedtobecomemoreintegratedfinancially.

With thatbackground, thequestionsconcerningmoving towarda fullyopencapitalaccountarethefollowing:

1. Although there is virtual consensus that free trade in goods is welfareenhancing, opinion is divided on the virtues of financial openness. Whatexplains thisdifference? Inwhatways is financial liberalizationdifferent fromtradeliberalization?

2. Is full capital account convertibility still an appropriate objective for every

country?

3.Ifso,whatisthebeststrategyforachievingit?Shoulditbefestinalente,or"makehasteslowly"?

CapitalControlsasaStabilizationTool

Thesecondissueonwhichtheprecrisisconsensushasbeenbrokenistheuseofcapitalcontrolsasastabilizationtool.Beforethecrisis, theconsensuswasthatcapital controls are bad, always and everywhere. That consensus no longerholds. The received wisdom today is that capital controls are not onlyappropriatebutevendesirableincertaincircumstances.Evenso,therearemanyunsettleddebates.

EffectivenessofCapitalControls

The first big debate is about the effectiveness of capital controls. People havequestioned effectiveness on the basis of mainly two arguments: first, capitalcontrolsdonotalterthevolumeofflowsbutalteronlytheirtenor,andsecond,capital controls can easily be circumvented by disguising short-term flows aslong-termflows.

PriceversusQuantityControls

Thenthereisthedebateoverwhattypesofcontrolareeffective.Countrieshaveusedbothprice-basedcontrolssuchastaxes,aswellasquantitybasedcontrols.However, evidence on which of them has been effective, and under whatcircumstances, is not conclusive.Two contributors to this volume address thisissuefromaLatinAmericanperspectiveinchapters23and24.

InIndia,forexample,wedeploybothprice-basedandquantitybasedcontrols.Ourexperiencehasbeenthatalthoughquantitycontrolsaremoreeffectiveintheshortterm,theycanalsobedistorting,inefficient,andinequitable.

CapitalControlsversusPrudentialMeasures

There isalsoanargumentaboutwhethercapitalcontrolscanbesubstitutedbyprudential measures. It is not clear that they are always exact substitutes. Ifcapital inflows are intermediated through the banking system, then prudentialmeasurescanbeapplieddirectlytodomesticbanks,circumventingtheneedforcontrols.Butwhatiftheinflowsaredirect?Thatistosay,loansarechanneleddirectly from foreign entities to domestic companies. In that case, the onlymechanismtopreventexcessiveleverage,andforeignexchangeexposure,maybebyimposingcontrols.

Againstthatbackdrop,thequestionsregardingcapitalcontrolsasashort-runstabilizationtoolarethefollowing:

1.Canwe define the distortion that capital controls aremeant to correct? Forexample, how do we determine whether capital flows are excessive ordangerous?

2. What have we learned about the effectiveness of capital controls as astabilizationtool?

3.Whencanprudentialmeasuresbesubstitutedforcapitalcontrols?

4. What criteria should we adopt to choose between price-based andquantitybasedcontrols?

5. Are capital controls symmetric as between inflows and outflows? In otherwords,shouldweuseonetypeofcontrolstocontrolinflowsandanothertypetolimitoutflows?

ForeignExchangeIntervention

The third important issue on which the precrisis consensus has dissolved isforeignexchangeintervention.Theprecrisisconsensus,atleastamongadvancedeconomies,wasthatinterventionintheforeignexchangemarketissuboptimal.Thatconsensusnolongerholds,withevensomeadvancedeconomiesdefendingtheir currencies from the safe haven impact. Emergingmarkets, for their part,have had a long and varied experience struggling with foreign exchange

intervention.Thepolicydilemmaintheeventofreceivingcapitalflowsbeyondthecountry'sabsorptivecapacitycanbequitecomplex.

Ifyoudidn't intervene in theforeignexchangemarket, thenyouwouldhavecurrencyappreciationquiteunrelatedtofundamentals.Ifyouintervenedbutdidnotsterilizetheresultantliquidity,youbecomevulnerabletoinflationpressuresand asset-price bubbles. If you intervened in the foreign exchangemarket andsterilized the resultant liquidity, youmay find interest rates firming up,whichwouldattractevenmoreflows-aclassiccaseofDutchdisease.Whatallthissaysisthatthereisreallynobenignoptionfordealingwithvolatilecapitalflows.

There isoneother important issue relating to foreignexchange intervention.Both currency appreciation and currency depreciation, quite unrelated tofundamentals, are complex problems. But there is a significant asymmetrybetween intervention for fighting appreciation and intervention for fightingdepreciation.

Whenyouarefightingcurrencyappreciation,youareinterveninginyourowncurrency.Your capacity todo so is, at least in theory, unlimited, quite simplybecauseyoucanprintyourowncurrency.Butwhenyouare fightingcurrencydepreciation,youareinterveninginahardcurrency.Yourcapacitytointerveneis therefore limited by the size of your foreign exchange reserves. Whatcomplicatesthedilemmaisthatthemarketisawareofthis.

So,thereistherealdangerthatbyinterveningintheforeignexchangemarket,you could end up losing foreign exchange reserves and not gaining on thecurrency.The lower your reserves dip, themore vulnerable you become.Andthevulnerability canbecomequite serious if your reserves gobelow the levelthat markets perceive as necessary to regain market access. It should also beclear thatafaileddefenseof theexchangerate isworse thannodefenseatall.Whenyoudecidetointerveneintheforeignexchangemarket,itisimportanttomakesurethatyourinterventionissuccessful.

Inthatcontext,thequestionsonthistopicofforeignexchangeinterventionarethefollowing:

1. Under what conditions is it appropriate for countries to intervene in theforeignexchangemarket?

2.Underwhatconditions is foreignexchange interventionpreferable tocapitalcontrols?

3. In most cases, countries claim that they are intervening in the foreignexchangemarketnottotargetanyparticularratebuttomanagethevolatilityinthe exchange rate. Is it necessary, then, to define up front your measure ofvolatilitythatwilltriggerintervention?

IhaveraisedverydifficultquestionsforwhichIhavenoanswers.Theexpertcontributors of the upcoming chapters shed light on the situation in their owncountries.

Note

IthanktheIMF,ProfessorOlivierBlanchard,andManagingDirectorChristineLagardeforinvitingmetotheconferenceatwhichIdeliveredthispaper,andfortheprivilegeofchairingthesessiononwhichthissectionofthebookisbased.

JoseDeGregorio

International financial integration and capital account management have beencentralissuesinthepolicydiscussioninrecentyears.However,theseissuesarenot new in emerging market economies. Some of these economies have haddisastrous experiences with financial crisis, most of the time caused bymishandled financial integration and weak macroeconomic policies. Theresilience of emergingmarket economies, in particular their financial systems,duringtherecentglobalfinancialcrisisshowsthatsomekeylessonshavebeenlearned.

Theexternalbalancehasusuallybeenatthecenteroffinancialandcurrencycrises. Periods of exuberance, capital account liberalization, rigidities in theexchangerate,andweakfinancialsystemscreateperiodsofoverheating,whichare followed by costly adjustments. Domestically, these episodes have beeninduced either by fiscal profligacy or by unsustainable private sector booms.Howtotakeadvantageofforeignfinancingwhilemakingtheeconomyresilientto changes in international conditions has become an important question forpolicymakersandresearchers.

Beforeproceedingwiththediscussion,itisusefultoclarifysomeideas.Often,there is no clear distinction between net and gross capital inflows, and littleunderstandingregardinghowtotacklethemandthepotentialconsequencesandrisks.

Netcapitalinflowsarethecounterpartofcurrentaccountdeficits.'Excessivenet inflowsmaybean indicationthat theeconomyisrunninganunsustainablecurrentaccountdeficit.Domesticexpenditurescouldbeatlevelsthatcannotbepermanently financedand thuswillbe followedbya sharpcorrection.At firstglance,thecurrentaccount-ornetinflows-iswhatmattersforexchangerates,in

particularfortherealexchangerate,whichistherelativepricebetweendomesticand foreign goods that gives the signal for resource allocation and demandpatternsconsistentwithsavingsandinvestmentdecisions.

Grossinflows,inturn,aretheresponsetoportfolioallocation.Grossflowsarecentraltofinancialstability.Theformandvolumethatgrossflowstakehaveadirect impact on the vulnerability of the financial system. It has long beenargued,rightly,thatforeigninvestmentandequityflowsaremorestable,whilebankingflowsaremorelikelytobesubjecttosharpreversals.

In this regard,aseparationbetweennetandgross inflowsbecomes relevant.Net inflows have to do with real exchange rates and competitiveness, whilegross flows have to dowith financial stability.There are interactions betweennet and gross flows as well as exchange rate developments and financialstability,butasastartingandorganizingdistinctionitisausefulone.

InthischapterIdiscussthreerelevantissuesonfinancialintegration,aswellas the challenges capital flows impose on policymaking. First I review theevidence on capital flows, then I discuss the benefits of financial integration.Finally,Igoovertheissueofcapitalaccountmanagementandpoliciestolimitthevulnerabilitiescomingfromfinancialopenness.

EvidenceonCapitalInflowsintoEmergingMarkets

After running significant current account deficits before the debt crisis, LatinAmericahadnoaccesstovoluntaryinternationalcapitalmarkets.Capitalflowsresumedintheearly1990sasresultoflowworldinterestratesandtheresolutionof the debt crisis. These developments raised several policy concerns (Calvo,Leiderman, and Reinhart 1994), and the expression "the problem of capitalinflows"wascoined.ThispreoccupationwasintensifiedbytheMexicancrisisofthemid-1990sandlaterbytheAsiancrisis.

Capital inflows were financing increasing current account deficits. Thesedeficits could become unsustainable and force a severe adjustment.Unsustainability canbedrivenby thecurrent accountor the financial (capital)account. In the first case,mounting artificial appreciationof the currency as a

resultofexchangeraterigiditieswouldbeexpected tobefollowedbymassivedepreciation and a currency crisis. In the second case,when the source is thecapitalaccount,evenanapparentlysustainablecurrentaccountdeficitcouldbereversed by a sudden halt in capital inflows as a result of changes in foreigninvestors' risk appetite, fear of insolvency, or simply contagion after a generalwithdrawal of investors from emerging markets. Of course, making thedistinction between a capital - and a current-account-driven reversal is quitedifficult,sincetheyareexpostthesame.Itissurprisingthatthecross-referencesbetweencurrentaccountreversalsandsuddenstopsareratherscarce.

As figure 23.1 shows, in the mid-1990s there was indeed a deficit in thecurrentaccountinemergingmarkets.ItstartedearlierinAsiaandlasteduntiltheAsiancrisis.InLatinAmerica,itstartedintheearly1990sandlasteduntil1998.On average, it was not massive, but there were disparities across countries.Mexico had an average deficit of 6.2 percent from 1992 to 1994. Somethingsimilaroccurred in someAsiancountries thatwerehitduring theAsiancrisis,such as Malaysia and Thailand. However, it was not the case in Korea andIndonesia. The reversal in Asia was sharp, whereas in Latin America it tookplacemoregraduallyafterthelate1990sandwasfollowedbyseveralyearsoflowgrowth.

Figure23.1

CurrentAccountBalance(%ofGDP).

Note: The graphs for Latin America and developing Asia represent simpleaverages across countries. Latin America includes Argentina, Brazil, Chile,Colombia, Mexico, Peru, and Venezuela. Developing Asia includes China,India,Indonesia,Korea,Malaysia,thePhilippines,andThailand.TheemergingmarketscategorycorrespondstotheIMF'sweightedaveragedefinition.

Source:InternationalMonetaryFund,WorldEconomicOutlook.

Things have been rather different recently. During the 2000s, emergingmarketswere net exporters of capital. Emergingmarket economies have beenrunning,onaverage,currentaccountsurpluses;hence, innet terms,capitalhasbeenflowingoutof thesemarkets.Onlyrecently,LatinAmericahadacurrentaccountdeficit.

In recent years, capital has been flowing "uphill" (Prasad, Rajan, andSubramanian 2007) from developing countries to advanced economies. Thisphenomenonhasbeendominatedby the largedeficits in theUnitedStatesandthe large surpluses inoil-exportingcountries.Chinahasalsoplayeda relevantrole in financing theUScurrent accountdeficit, as shown in figure23.2.Thispatternisevidentsincethemid-1990s,butitwasmuchmorepronouncedintheyearsbeforethecrisis.Thelineinthefigureshowsthecurrentaccountbalanceof Latin America, the newly industrialized Asian economies, and developingAsia.2Theyhaveclearlybeennetexportersofcapitalsincethelate1990s.

Figure23.2

WorldCurrentAccount(billionsofUSdollars).

Source:InternationalMonetaryFund,WorldEconomicOutlook(WEO);2013istheWEOforecast.NIAE,newlyindustrializedAsianeconomies.

What is the basis for concerns about capital inflows to emerging marketeconomies?Therearetworeasons.Thefirstisthatgrossinflowshaveincreasedovertime,despiteoutwardnetflows.Figures23.3aandbshowgrossinflowsforthesampleofAsianandLatinAmericancountries.

The increase ingross flows isvery significant.The figures show, consistentwith the usual narrative, that the most important and stable component ofinflowsinLatinAmericaisforeigndirectinvestment(FDI).Bankingdebtflows,which make up the bulk of the "other investment" category, are much lessimportantandalsomorevolatile.Incontrast,inAsia,theroleofportfolioflows

and banking flows is muchmore important; indeed, the retrenchment of debtflowsduring theglobal financial crisiswasmuchmore severe inAsia than inLatinAmerica.'Thissuggestsseveralpolicy issuesregardingfinancialstabilityandthevulnerabilityofemergingmarketstoexternalfinancialturmoil.

However,despiteacontainedcurrentaccountdeficit,itispossibletoobservenet(unofficial)capitalinflowsifthereisaccumulationofinternationalreserves.Under no foreign reserves accumulation, net capital flows equal the currentaccount. Since emerging markets have been accumulating large amounts ofinternational reserves, capital flows could be flowing into emerging marketsdespite a surplus in the current account. Figure 23.4 replicates figure 23.2,subtracting from the current account balance the accumulation of internationalreserves. The latter is a capital outflow, so the difference is total outflowsincludingreserveaccumulation,andsonetinflowsisanegativenumber.

It is clear that despite no demand to finance excess domestic expenditure,capitalhasbeenflowingtoemergingmarketsbecauseoftheadditionaldemandfor reserves. Indeed, surges in capital inflows during recent years have cometogether with large accumulations of reserves and moderate current accountdeficits, even surpluses in some countries. This is very different from theexperience of the 1990s, when the incidence of current account deficits wasmuchmorerelevant(DeGregorio2014).

Inrecentyears,emergingmarketeconomieshavenotbeenfloodedbycapitalflows, andnet flowshave come togetherwith reserve accumulation.Causalityamongreserves,capitalflows,andcurrentaccountbalanceisadifficultissue.Intheaccountingdefinitions,accumulationofreserves(AR)isequaltothebalanceinthecurrentaccount(C)plusthebalanceinthefinancialaccount(F).Ifthereisanincreaseinreserves,AR>0,aARwillresultinanimprovementinthecurrentaccount balance, while the remaining (1 - a)AR will result in an increase incapitalinflows.

Figure23.3

GrossCapitalInflows(billionsofUSdollars).

a.LatinAmerica.b.Asia.

Source:InternationalMonetaryFund,InternationalFinancialStatistics.

Figure23.4

WorldCurrentAccountMinusReservesAccumulation(billionsofUSdollars).

Source:InternationalMonetaryFund,WorldEconomicOutlook.

If theaccumulationofreservesresultsonly inan increase incapital flows,awillbezero.Incontrast,ifalltheaccumulationofreservesabsorbscapitalthatisflowingin,withoutfurtherflows,thecurrentaccountshouldbeaffectedwithavalueofaequaltoone.Thereislittleevidenceonthis,andestimatesarewide,ranging from0.4 (IMF2012) to0.8 (BergstenandGagnon2012).4Ofcourse,thevalueofadependsonthecharacteristicsofthecountries,butinorderforthis

parameter to be large, one needs to show that the effects of sterilizedinterventionon theexchange ratearesizable;otherwise, it isdifficult toaffectthecurrentaccountthroughreserveaccumulation.Butevidenceontheimpactofinterventiononexchangerateiselusiveandtheeffectsareatmostarelimited.Therefore,accordingtothisindirectevidence,thevalueofU.islikelytobelow.However,muchmoreresearchisneeded,sincethisissueisatthecoreofotherissuessuchascurrencymanipulation,capitalflowsabsorption,andtheimpactofreserveaccumulationintheglobaladjustment.

FinancialIntegration

The evidence for the positive effects of financial integration on economicperformanceiselusive.Mostsurveysandrecentresearchhavefoundsmallornosignificanteffects,andthepolicyconclusionisgenerallythatinordertoreapthebenefits fromfinancial integration, itmustbedonewithinahealthy regulatoryand supervisory framework. Certainly, unfettered financial integration hasproved tobe riskyand,mostof the time,hashadverynegativeconsequences.However,theevidencedoesnotsupportfinancialautarky.Indeed,theevidencealsoshowsthatascountriesgrow,theirleveloffinancialintegrationincreases.

Several recent papers survey and provide additional evidence on financialintegrationandgrowth.Forexample,Obstfeld(2009)concludesthat"despiteanabundance of cross-section, panel, and event studies, there is strikingly littleconvincingdocumentationofdirectpositiveimpactsoffinancialopeningontheeconomicwelfare levelsorgrowth ratesofdevelopingcountries."And fromapolicy point of view, "This survey discusses the policy framework in whichfinancialglobalizationismost likelytoprovebeneficial."Obstfeldalsoreportsthathighlevelsofincomearecorrelatedwithhighlevelsoffinancialintegration.Ofcourse,causalitydoesnotgo fromfinancial integration todevelopmentbutfromhighlevelsofincometomorefinancialintegration.

Similarly,Koseandothers(2009)findthat"overall,ourcriticalreadingoftherecentempiricalliteratureisthatitlendssomequalifiedsupporttotheviewthatdeveloping countries can benefit from financial globalization, but with manynuances.Ontheotherhand,thereislittlesystematicevidencetosupportwidely

citedclaimsthatfinancialglobalizationbyitselfleadstodeeperandmorecostlydeveloping country growth crises." They also find that financial integrationmight have collateral effects that may induce productivity growth, such asimprovedinstitutionalqualityandbettermacroeconomicpolicies.

In a recent meta-regression analysis, based on 2,340 regressions, Jeanne,Subramanian,andWilliamson(2012) fail[ed]"toproducerobustevidenceofapositive relationship between financial globalization and growth, raisingquestionsabout thepursuitofall formsof internationalfinancial integrationasanurgentpolicygoal."

However,theevidenceshowsthatthereareimportantdifferencesaccordingtothetypeofcapitalflows.Borensztein,DeGregorio,andLee(1998)foundthatfor countries with a minimum level of human capital, FDI spurs economicgrowth. This evidence is confirmed by Jeanne, Subramanian, andWilliamson(2012),whofound"somewhatreassuringly,portfolioequityandFDIflowsaremorelikelytogeneratepositiveandsignificanteffectsongrowthcomparedwithbankingorportfoliodebtflows."

The evidence on the weak link between financial integration and economicgrowthdoesnotcomefromtheimpactoffinancialintegrationontheincidenceof financial crisis.As reported byKose and others (2009), based on evidencefromEdwards (2005),countrieswithhighercapitalmobilitydonothavemoreexternalcrises,andthecostofcrisisisnogreaterincountriesthatrestrictcapitalinflows.

Themostsupportiveevidenceonthepotentialbenefitsoffinancialintegrationcomes from lookingat thresholdeffects.Theconclusion from this literature isthateconomiesneedaminimumlevelofgovernance,institutionaldevelopment,qualityofmacroeconomicpolicies,andothercharacteristicstobeabletoabsorbcapitalflowswithoutdetrimentaleffectsongrowth.ThispointwasfirstraisedinPrasad and others (2003) and recently revisited by Chen and Quang (2012).Thesefindingsmayberelatedtoindirecteffectsofopeninguponproductivitygrowth.Still,theevidenceisnotstrongenoughtoprovidedefiniteconclusions.

Two additional findings have raised doubts about the benefits of financialintegration.First,countriesthathavegrownthemostarethosethatrelyless,notmore,onforeignsavings(Prasadetal.2007).However,thisisprobablybecausecountries that have grown fast, especially inEastAsia, have reliedmore on avery high savings rate and capital accumulation, so their need for net foreigncapital are relatively small.Weknow there is a two-way relationship betweensavingsandgrowth.High-savingseconomies,inpartbecauseofhighergrowth,havelessneedofforeignfinance.Asecondandrelatedfindingisthe"allocationpuzzle" of Gourinchas and Jeanne (2011), in which capital flows to low, nothigh, total factor productivity growth countries. However, as the authorsemphasize, this is also related to the linksbetween savings andgrowth, ratherthan a direct consequence of financial integration. Therefore, these additionalfindings are not necessarily related to the effects of integration on economicgrowth, but they point toward more fundamental determinants of economicgrowththatalsohaveanimpactonthedegreeoffinancialintegration.

Summingup,theevidenceshowsthefollowing:

• There is not a clear link from financial integration to economic growth.Financial integrationbyitself isnotanengineofgrowth.However, thereisnoevidencethatitisharmful.

• The type of capital flows matters for economic growth. FDI and portfolioequities tend to bemore supportive of economic growth,while this is not thecaseforbankingflows.Thiscouldbebecausefinancialcrisescomemostlyfromdistortionsinthebankingsector.

• High income is correlated with high financial integration. As economiesdevelop, their financial integration with the global economy increases.Therefore, financial integration is a result of economicgrowth, andwedonotknowwhatwouldhappenifeconomiesavoidedintegrationwhiletheygrew.Isitpossibletokeepgrowingwithaclosedcapitalaccount?Theevidenceindicatesthatthisisunlikely.

•Someevidenceshows that therearesome thresholdeffects; that is,countries

need to have some minimum institutional standards to benefit from financialglobalization.

The main policy implication is that opening up requires a regulatory andsupervisory framework that allows a country to reap benefits from integrationwhile preserving financial stability and avoiding costly financial crisis.Economiesmustfacethechallengeofintegrationasgrowthproceeds.

The Latin American experience regarding financial integration and theincidenceof crisis is quite informative.As figure23.5 shows,LatinAmericancountrieshavebecomemoreintegratedbutalsomoreresilient.

Figure23.5

InternationalFinancialIntegrationinLatinAmerica(%ofGDP).

Note:TheindexcorrespondstointernationalassetsplusliabilitiesoverGDP.

Source:LaneandMilesi-Ferrettidatabase.

Duringthedebtcrisis,therewaslessfinancialintegration,andintegrationwasmoretiltedtodebtflows.LatinAmericancountriesfinancedrapidcreditbooms,andcountriesthatincreasedcreditthemostsuffereddeepercrises(DeGregorio

andGuidotti1995).Beingfinanciallyopen,inacontextofappropriatedomesticregulation,doesnotnecessarilyresultingreatervulnerability.TheAchilles'heelhasbeenthecreditboom.

Thisdiscussionhassomerelevance to tradeopenness.Acursory lookat theglobalevidence indicates thatmoreopeneconomiesdidnothaveworsecyclesduring the crisis thanmore closed economies. Economiesmore open to tradecould have sufferedmuchmore at the beginning of the crisis, but theirwholecyclewasnotnecessarilyworse.Beingopentotradedoesnotmakeaneconomymorevulnerable.

ManagementoftheCapitalAccount

The first line of defense against massive capital flows is exchange rateflexibility.Unsustainableexchangeratemanagementandone-sidedbetsareanincentiveforcapitalflowvolatility.Inaddition,aninflation-targetingregimeandsoundfiscalpoliciesshouldhelppreventexcessivecapitalinflows,whichisthesameasexcessivecurrentaccountdeficits.

However,thisisnotenough.First,thevalueofthecurrencyneededtoreduceincentivesforcapitalinflowsmightbesufficientlyhighthatpolicymakersfinditinconvenient.Thereisawell-groundedbiastohaverelativelyweakcurrenciesinordertofosterexport-ledgrowth.Inthiscase,capitalcontrolswouldbeservinga competitiveness purpose. Second, the nature of flows might be such thatauthorities find it prudent to change the composition of flows or reduce somespecificinflows,suchasexcessiverelianceonshort-termbankingflows.Inthiscase, the control would be serving a financial stability purpose and could beconsideredamacroprudentialtool.

Before we discuss policies oriented toward the short-term management ofcapital flows, it is important to comment on long-term financial integration.Manyyearsago therewasa lotofdiscussionof sequencing.Whatmustcomefirst,financialopeningorfinancialliberalization?Thisisnolongeranissue.Thefirst task is to develop the domestic financial system. Setting a strongsupervisory and regulatory framework is crucial to ensure that the capital

account has a sound domestic financial system. Foreign financial institutionsmight help with the development of the domestic financial system, but theymightalsobeacauseofconcernifregulationisweak.

The form inwhich internationalbanksoperate indifferent countries isquiteimportanttoring-fencethedomesticfinancialsystemfromproblemsoriginatinginthehomecountriesoftheforeignbanks.Afirstimportantstepistoencourageforeignbankstohavethesamerulesandregulationsasdomesticones.Thiscallsfor the establishment of subsidiaries of foreign banks rather than branches.Subsidiarieshavetheirownboards,whichareresponsibleforbankoperationsinthe host country, and they have strong limits on operations with the parentcompany. Branches can more easily transmit turbulence to the host country.Subsidiarization is not a panacea, but it has worked reasonably well in LatinAmerica.

Ahighlydebatedissueistheuseandeffectivenessofcapitalcontrols.Whencapitalcontrolsareusedforfinancialstabilitypurposes,itispossibletorelabelthemmacroprudential tools.When their purpose is to affect the exchange rateand the current account balance, they are capital controls, although somemaycallthemmacroprudentialtoolsasacommunicationdevice.

The empirical evidence on effectiveness is varied, since capital controls areused for several goals and effectiveness is countryspecific. They are used tocontrol the volume of flows, to change their composition, to ensuremonetaryindependence, and to depreciate the exchange rate. These objectives combinesome financial stability concerns with macroeconomic stability concerns. Thelatterrefertolimitingexchangeratepressuresandreducingnetflows,whichisthesameasreducingthecurrentaccountbalance.

Regardingpurely financial stability concerns, themain riskofgross inflowsstemsfromcross-borderbankingflows.Anumberofmacroprudentialtoolscanbeusedtopreservefinancialstability,andrestrictionsoncross-borderflowscanbe one of them. In Korea, a tax levy on banks' noncore liabilities wasimplemented to curb the increasing importance (deemed to be a source ofvulnerability)ofcross-borderflows(BrunoandShin2013).

Recent work (Magud, Reinhart, and Rogoff 2011; Ostry et al. 2011;Habermeier,Kokenyne,andBaba2011)hasreviewedtheexistingevidenceoncapital controls.Broadly, the evidencehasnot found significant effectson theexchange rate.Some small effects havebeen foundon thevolumeof inflows.The most frequent finding is that capital controls affect the composition ofinflows,increasingmaturity.

LetmeusetheevidencefromChile,theposterchildformarket-basedcapitalcontrols,toclarifysomepoints.'Mostoftheclaimsabouteffectivenesslookatthe statistical significance without looking seriously at the economicsignificance.Aneffectcouldbe significantlydifferent fromzerobutofaverysmall, and therefore irrelevant, magnitude. In the case of Chile, the paper byGallego, Hernandez, and Schmidt-Hebbel (1999) is the only one that foundsignificanteffectsonthevolumeofflows.Theyestimatedthatthetotalimpactofcapital controls inChilewas to reduce inflowsbyabout2percentofGDP,whiletotalcapitalinflowsamountedtonearly27percentofGDP.Certainlyitisaverysmalleffect,andnotrobustacrossstudies.

Only some small short-run effects on the real exchange rate have beenidentified. Only Edwards and Rigobon (2009) estimate statistically significanteffectsontheextentoftheappreciationofthepeso.However,themagnitudeofsucheffectiseconomicallysmall.Accordingtotheirestimates, theeliminationof thecontrol,whichconsistedofanunremuneratedreserve requirement, fromitsmaximumwouldhaveappreciated theexchangeratebetween2percentand2.5percent.

Themost frequent finding-andnot only inChile - has been a change in thecomposition of inflows. The evidence for Chile is that short-term debt wouldhavedeclinedby0.5percentto1percentofGDPasaresultofcapitalcontrols(CowanandDeGregorio2007).Again, this isnotaneconomicallysignificanteffect.

Tobeconsistentwiththediscussionoftheevidenceonfinancialintegration,one could argue that capital controls do no harm. However, there are twoconcerns, supported by some evidence, regarding negative effects of capital

controls.

Aslongascapitalcontrolsareabletochangethecompositionofdebtflowsbyincreasingthecostofshort-termrelativetolong-termborrowing,firmsthatrelyonshort-termdebt(mostlysmallandmedium-sizeenterprisesandfirmswithashortcredithistory)willbenegativelyaffected.There is someevidence in theChileancaseofachangeinthestructureoffinancing,whichcouldhaveinduceddistortions (Forbes 2007). However, this is a characteristic of mostmacroprudential tools aimed at tapering credit expansion: They have theunavoidable cost of making credit more expensive; otherwise they would beineffective.

AlthoughIdonotthinkthiseffectcouldhavebeentoosignificantbecausethequantity effect is not so large-themain risk of capital controls is to create thefalse idea of insulation. Policymakers may think they have gained monetaryindependence to set the interest rate at any levelwithout repercussions on theexchange rate. Indeed, the most famous Latin American cases of capitalcontrols-Chileinthe1990sandBrazilinthelate2000s-tookplaceinthecontextofveryhighinterestrates,whichcouldhavebeenpartlyresponsibleforthelargeappreciationstheircurrencieswentthrough.Indeed,bylate1996,atthepeakofthecapitalinflowsurgeinChile,themonetarypolicyratewasabout15percent,'whilethefederalfundsratewasat5.25percent.Brazilhadasimilarexperience:Bymid-2008,whentherealreacheditsmaximum,themonetarypolicyratewasat12percentandrising to13.75percent,while thefederal fundsratewasat2percent.7

Because of concerns about potential costs, some countries might find itworthwhiletoapplycapitalcontrols,astheeffectivenessiscountryspecific.Forcontrols tobeeffectiveandminimizecostsanddistortions, it is important thatmacroeconomics policies be well aligned with macroeconomic and financialstability.Controls could serve as a complement andnot a substitute for soundmacroeconomic and financial policies. But having strong macroeconomicpolicies and a strong financial system could make it unnecessary to considercapitalcontrols,aswas theexperience inmanyemergingmarkets thatmade itsuccessfullythroughtheglobalfinancialcrisis.

Notes

1.Thisignorestheaccumulationofreserveshere;thatisdiscussedbelow.

2. The countries in each category are those defined by the IMF in theWorldEconomicOutlook.

3.Forfurtherdiscussiononcross-borderbankingflows,seeCIEPR(2012).

4.TheelasticitycomputedinIMF(2012)isinteractedwithcapitalcontrols,andthe value ranges from zero for no capital controls to 0.4 with the strongestcapitalcontrolsinthesample.

5. For details, see Cowan andDeGregorio (2007). For a discussion of LatinAmerica,seeDeGregorio(2014).

6.Bythattime,monetarypolicywassetinUF(unidaddefomento),anindexedunitofaccount,sotohavethenominalequivalent,whichistheoneusedinthetext,Iusetheyearlyinflationrateatthattime.

7.InthecaseofBrazil,ChamonandGarcia(2013)foundnosignificanteffectsontheexchangerate,concludingthattheIOF(PortugueseinitialismforTaxonFinancialTransactions)didnotpreventappreciation.Theyarguedthatthe"realgame-changer" for the appreciating trend of the real (which has reverted inrecentyears)wasthecutinthemonetarypolicyinterestrate.

References

Bergsten,C.Fred,andJosephE.Gagnon.2012."CurrencyManipulation,theUSEconomy, and the Global Economic Order." Policy Brief PB12-25, PetersonInstituteofInternationalEconomics,Washington,DC.

Borensztein,Eduardo,JoseDeGregorio,andJong-WhaLee.1998."HowDoesForeignDirect InvestmentAffectEconomicGrowth?" Journal of InternationalEconomics45(1):115-135.

Bruno, Valentina, and Hyun Song Shin. 2013. "Assessing MacroprudentialPolicies:CaseofKorea."Facultypaper,PrincetonUniversity,Princeton,NJ.

Calvo,GuillermoA.,LeonardoLeiderman,andCarmenM.Reinhart.1994."TheCapitalInflowsProblem:ConceptsandIssues."ContemporaryEconomicPolicy12:54-66.

Chamon, Marcos, and Marcio Garcia. 2013. "Capital Controls in Brazil:Effective?Efficient?"InternationalMonetaryFund,Washington,DC.

Chen, Jinzhao, andThereseQuang. 2012. "International Financial integrationsand Economic Growth: New Evidence on Threshold Effects."Working Paper2012-06,UniversiteParisOuest,Paris.

CommitteeonInternationalEconomicPolicyReform(CIEPR).2012.BanksandCross-Border Capital Flows: Policy Challenges and Regulatory Responses.Washington,DC:BrookingsInstitution.

Cowan,Kevin, and JoseDeGregorio.2007. "InternationalBorrowing,CapitalControlsandtheExchangeRate:LessonsfromChile."InCapitalControlsandCapital Flows in Emerging Economies: Policies, Practices and Consequences,ed.S.Edwards.Chicago:UniversityofChicagoPress.

DeGregorio, Jose. 2014.HowLatinAmericaWeathered theGlobalFinancialCrisis.Washington,DC:PetersonInstituteofInternationalEconomics.

De Gregorio, Jose, and Pablo Guidotti. 1995. "Financial Development andEconomicGrowth."WorldDevelopment23(3):433-448.

Edwards, Sebastian. 2005. "Capital Controls, Sudden Stops, and CurrentAccount Reversals." NBER Working Paper 11170, National Bureau ofEconomicResearch,Cambridge,MA.http://www.nber.org/papers/w11170.

Edwards, Sebastian, andRobertoRigobon. 2009. "CapitalControls, ExchangeRateVolatilityandExternalVulnerability."JournalofInternationalEconomics78(2):257-267.

Forbes,Kristin J. 2007. "OneCost of theChileanCapital Controls: IncreasedFinancial Constraints for Smaller Traded Firms." Journal of InternationalEconomics71(2):294-323.

Gallego, Francisco, Leonardo Hernandez, and Klaus Schmidt-Hebbel. 1999.

"Capital Controls in Chile: Effective, Efficient?" Working Paper 59, CentralBankofChile,Santiago.

Gourinchas, Pierre-Olivier, and Olivier Jeanne. 2011. "Capital Flows toDeveloping Countries: The Allocation Puzzle." Faculty paper, University ofCaliforniaatBerkeley,Berkeley,CA.

Habermeier, Karl, Annamaria Kokenyne, and Chikako Baba. 2011. "TheEffectiveness of Capital Controls and Prudential Policies in Managing LargeInflows." IMF Staff Discussion Note 11/14, Monetary and Capital MarketsDepartment, International Monetary Fund, Washington, DC, August 5.http://www.imf.org/external/pubs/ft/sdn/2011/sdn1114.pdf.

International Monetary Fund (IMF). 2012. "External Balance Assessment(EBA): Technical Background of the Pilot Methodology." InternationalMonetary Fund, Washington, DC, August 3. http://www.imf.org/external/np/res/eba/pdf/080312.pdf.

Jeanne,Olivier,ArvindSubramanian,andJohnWilliamson.2012.WhoNeedsto Open the Capital Account? Washington, DC: Peterson Institute ofInternationalEconomics.

Kose,M.Ayhan, Eswar S.Prasad, Kenneth Rogoff, and Shang-JinWei. 2009."FinancialGlobalization:AReappraisal."IMFStaffPapers56(1):8-62.

Magud, Nicolas E., Carmen Reinhart, and Kenneth S.Rogoff. 2011. "CapitalControls: Myth and Reality-A Portfolio Balance Approach." NBERWorkingPaper 16805,NationalBureau ofEconomicResearch,Cambridge,MA. http://www.nber.org/papers/w16805.

Obstfeld, Maurice. 2009. "International Finance and Growth in DevelopingCountries:WhatHaveWeLearned?"IMFStaffPapers56(1):63-111.

Ostry, Jonathan D., Atish R.Gosh, Karl Habermeier, Luc Laeven, MarcosChamon, Mahvash Qureshi, and Annamaria Kokenyne. 2011. "ManagingCapital Inflows: What Tools to Use?" IMF Staff Discussion Note 11/06,International Monetary Fund, Washington, DC, April 5. http://www.imf.org/external/pubs/ft/sdn/2011/sdn1106.pdf.

Prasad,EswarS.,RaghuramG.Rajan,andArvindSubramanian.2007."ForeignCapitalandEconomicGrowth."BrookingsPapersonEconomicActivity38(1):153-230.

Prasad,Eswar S.,KennethRogoff, Shang-JinWei, andM.AyhanKose. 2003."Effects of Financial Globalization onDevelopingCountries: Some EmpiricalEvidence." IMF Occasional Paper 220, International Monetary Fund,Washington,DC.http://www.imf.org/external/pubs/nft/op/220.

MarcioHolland

This chapter presents the recent Brazilian experience of dealing with capitalinflows associated with domestic currency appreciation, and the use ofmacroprudentialmeasures to copewith the capital surges. Restrictions on thefinancial account in Brazil are only some of the ingredients of the country'seconomicpolicy,which includescontrolling inflationalongwithmaintainingaconventional monetary policy, as well as macroprudential measures, a fiscalconsolidation program, a solid financial system, a focus on investment andinfrastructure,andaverycomprehensiveincomeinequality-reductionpolicy.

The consequences of the 2008-2009 international financial turmoil have notyet come to an end, as the world is still waiting to see advanced economiesaddressing important financialandpoliticalproblems. In theUnitedStates, theFederal Reserve has become the main source of economic stimulus, havingimplemented the third round of quantitative easing,with only partial and lessthansatisfactoryresultsfortheUnitedStatesbutwithnegativeconsequencesforemergingeconomies.

InEurope,thecrisisstillpersists,mainlyineurozone'speriphery,withsevereeconomic and social consequences. Therefore, it is important that euro areacountries come up with rapid and durable solutions, especially in terms ofbankingsupervisionandfiscalconsolidation,sothateconomicgrowthpicksupin the region. Since the fourth quarter of 2011, quarter-over-quarter economicgrowth in the euro area has been zero or below. European leaders are stillstrugglingtofindasolutionthatwillputtheregionbackontrack.

In this particular circumstance, after establishing themonetary policy rates,central bankers in advanced economies began proposing a zero lower boundmonetary policy that included aggressive quantitative easing. By 2013, morethanUS$9trillionhadbeeninjectedintotheworld'sliquidity,withpartofthis

huge amount of money searching for very rare positive returns, mainly inemergingmarket(EM)economies.Theunconventionalmonetary-easingpoliciesconsist of central bank purchases of domestic government bonds or even theprovisionofexplicitguidanceonthefuturepathofinterestrates,withmedium-terminflationornominaloutputandunemploymentratesastargets.

As a rational response,Brazilianmonetary authoritieshaveput intopracticemacroprudential measures that include restrictions on the capital accounts.Roughlyspeaking,thesemeasuresconsistofincreasesinMF(ataxonfinancialtransactions)ratesonshort-termfinancialoperations,includingexternalloansuptooneyear,with themain aimbeing to reducegains in carry-trade strategies.Foreigndirectinvestmentsandevenlongtermfinancialoperationsareleftaside.

Thischapterhighlights the fact thatmanagingcapital inflows isonlypartofthepolicymix.Suchastrategyhasbeeneffectiveindealingwiththeincreasinginternational liquidity and in preventing inflows of very short-term foreigncapital,aswellasinchangingthecompositionofcapitalinflowstowardbetter-qualitycapital inflows.Therefore, theexchange rate staunched itsappreciationmovementandsincethenhasbeenlessvolatile.Itisalsoworthnotingthattheinterestratemayhaveincreasingpowertoaffecttheexchangerateundersuchapolicy.

Weknowhowcontroversialtheroleandtheeffectivenessofcapitalcontrolsare during crises. However, there has been a distinct scenario since 2008suggestingthatcountriessuchasBrazilshouldcareaboutitsexchangerateandqualityofcapitalsurges.Acombinationofazerolowerboundmonetarypolicywithquantitativeeasing,provokingexchangeraterealignmentsacrosstheworld,istreatedaspartofthesolutiontothefragilitiesinadvancedeconomies,butatthe same time, it represents sizable constraints on EM economies. We aredefinitelyinanoncooperative,non-zero-sumgame,andthusEMeconomiesarebeingharmedratherthanbenefited.Capitalaccountmanagementmeasuresareatechnical rather than an ideological issue. They join the policy toolkit, withsuccessfulresults.

In thenextsectionIexamine theeconomic literatureoncapitalcontrolsand

thendescribetherecentBrazilianexperience.

TheLiterature

Several recent studieshave reviewed the roleplayedbycapital controls in thecontext of a world economy emerging from a financial crisis. Even the IMF(2012) in its "institutional view" on capital controls suggests some signs ofprogress on the matter. The analysis of Ostry and others (2010) can beconsideredoneofthefirstinthenewcontext.Theydiscussnotonlythebenefitof the capital inflows into emerging markets but also the appropriate policyresponses. Baba and Kokenyne (2011) estimate the effectiveness of capitalcontrol in response to inflow surges inEMs such asBrazil,Colombia,Korea,andThailandinthe2000s.

It is fair to say that the globalization of capitalmarkets has been beneficialwhenitallowscapitalflowstomovetowardtheirmostattractivedestination,butat the same time this process has been associated with episodes of dramaticfinancialcrises.Inthisscenario,thereisanincipientdebateregardingtheroleofinternationalcapitalflowsintriggeringsuchcrises,andifthatisthecase,capitalcontrols become an important policy tool to be used by EM countries, ashappenedquiteoftenduringthe1990s.

Recently, a number of studies have argued that free capital mobility hascreated a highly unstable international financial system and that developingcountriesneed tomanagecapital flows. It is important tonote that this idea isnotanewoneanddatesbacktoJamesTobin(1978),whoarguedthatreducingmacroeconomicinstabilitywouldrequiretheadoptionofaglobaltaxonforeignexchangetransactionstoreducespeculationininternationalfinancialmarkets.

The rationale for imposing restrictions (capital controls) on internationalcapital flows is associated with the belief that capital markets are usuallycharacterized bymarket failures and distortions (information asymmetry), andthat such imperfections are magnified by difficulties in enforcing contractsacrossbordersandbyakindofherdbehavior,suchaswheninvestorsoverreacttoexternalshocks.

Oneofthereasonsmostoftenvoicedinthedefenseofusingcapitalcontrolsduring periods of crises is that it allows the central bank to stem the drain onforeign exchange reserves and that monetary authorities could initially raiseinterest rates; once capital controls are inplace, it gives room for a lower andmorestableinterestrate,whichactsinaprocyclicalway.Itisalsoimportanttonote that capital controls introduce a wedge between domestic and foreigninterest rates, and the domestic interest rate policy does not need to followinternational interestrateswhenfacingtheconsequencesof internationalcrisesandthebreakdownofuncoveredinterestparity.

Thediscussionofsomepolicyissuesregardingtheeffectivenessofimposingcapital controls should be carried out with the understanding of the requiredsteps (sequencing reform) toward the liberalizationof the capital account.Themainissueisnotwhetherornotcapitalcontrolsshouldbeeliminatedbutunderwhat conditions (when and how fast) they would be effective in achievingdesirableeconomicoutcomes.Mostcountries'experienceswithcapitalcontrolshaveshownthattheprivatesectorfoundwaysofgettingaroundcapitalcontrols,usually adopting strategies based on overinvoicing (underinvoicing) imports(exports)andmislabelingthenatureofcapitalmovements(short-termportfolioflowslabeledastradecredit).

The majority of the studies have argued that before liberalizing the capitalaccounts, it is necessary to reverse major fiscal imbalances and achievemacroeconomic stability. The past experience of many developing countries,including Brazil in the 1990s, has shown that although price stability wasobtained, it was still necessary to implement fiscal reforms to improve theoverall macroeconomic fundamentals of the economy. Other than this,establishing a sound banking system is also necessary before developingcountriescanliftrestrictionsoncapitalmobility,asbankswillintermediatetheinflowsofcapital,whichshouldnothappeninaninefficientway.

Previousworks, such as thosebyReinhart andSmith (2002) andKaminskyand Schmukler (2001), examined the role of temporary controls on capitalinflows, emphasizing that capital controls have two crucial features: they areasymmetric (the target is capital inflows rather than capital outflows) and

temporary. The authors examined possible reasons for policymakers to adoptcontrolsoncapitalinflowsandtwotypesofshocksthatcanresultinexcessivecapital inflows (temporarychanges in the foreign interest rate and indomesticmonetarypolicy).Themajor empirical findingsare that the tax rateoncapitalinflows should be very high to affect the capital account balance, that theeconomicbenefitoftaxingcapitalinflowsisnotsignificant,andthattakingtoolongtoremovecapitalcontrolscanreversewelfarebenefits.

MalaysiaandThailandare twoEMcountriesmakinguseofcapital controlsduring episodes of financial crises in the recent past; the experiences of thosecountrieswereexaminedbyEdisonandReinhart (2001).Theirmainempiricalfindingssuggestthatcapitalcontrolshelpreduceinterestratevolatilitybutthereis mixed evidence for avoiding exchange rate volatility. Another importantfinding refers to a wider and more variable bid-ask spread during controlperiods, and little evidence that capital controls were effective in reducingvolatilityspillovers.

Kaminsky and Schmukler (2001) dealt with the question of whether or notcapitalcontrolsaffectthelinkbetweendomesticandforeignstockmarketpricesand interest rates-in otherwords,whether theymatter for internationalmarketintegration.Theauthors found littleevidence thatcapital controlscansegmentdomestic and foreignmarkets, and evenwhen they do, the effects do not lastlong.Finally,theyfoundthatitisdifficulttodistinguishtheeffectsofcontrolsoninflowsandoutflows.

TheChileanexperienceduringthe1990shasbeenexaminedindetailbyDeGregorio,Edwards,andValdes(2000),whodevelopedworkthataddressedtheissueofwhetherornotcontrolsoncapitalinflowsareefficientthroughtheuseof unremunerated reserve requirement. They also examined the effects oninterest rates, the volume and composition of capital inflows, and the realexchange rate.Theirmain empirical findings suggest that it is difficult to findlong-runeffects,andthatcapitalcontrolsgenerateanincreaseintheinterestratedifferential only in the short run, no effects on the real exchange rate, and asignificant effect in the composition of capital inflows in favor of a longermaturity.

Anotherstudylinkedto theChileanexperiencewithcontrolsoninflowsandoutflowsofcapitalduring the1990swas reportedbyEdwards (1999),and theempirical results suggest that controls on outflows are not effective, whereascontrolsoninflowshavetheadvantageofaffectingthematurityofforeigndebt,whichisadesirableoutcomeforthemonetaryauthorities.ThethreemaingoalsofChile'scapitalcontrolsweretoslowdowntheinflowofcapitalandchangeitscomposition toward capital of longer maturities, to reduce and postpone realexchange rate appreciation, and to help the monetary authorities adopt anindependent monetary policy (maintaining an interest rate differential). Theauthoralsofoundthatcontrolsoncapitalinflowsarenotsufficienttoeliminatefinancial instability. The GARCH-generalized autoregressive conditionalheteroskedasticity-estimationrevealsthattherestrictionsoncapitalinflowsweresuccessful in reducing stock market instability, but not short-run interest ratevolatility.

Generally, the discussion about the effects of restrictions on capital inflowshasshownthatcontrolsareimportanttoexplainingchangesinthecompositionofcapital flows in thedirection intended(reducing theshareofshort-termandportfolio flows, and increasing foreign direct investment). The literatureaddresses the issue ofwhether external factors (international interest rates andliquidity) or internal factors (domestic fundamentals) are more important toexplainingtheincreaseinthefinancialflowstoemergingeconomies,andrelatethis to the question of how these countries respond to an increase in capitalflows.

The empirical evidence has indicated that capital inflow ismore volatile inLatinAmericathaninAsia,andshort-termcapitalismorevolatilethanallothertypes of capital flows. The adoption of sterilized intervention increases thevolume of total capital flows through short-term capital, and capital controlshave no significant effect on reducing the overall volume of flows, but theyaffect the composition of capital flows in favor of foreign direct investments.Finally, short-term flows are not sensitive to changes in international interestrates,althoughthecompositionofcapitalflowsdoesrespondtosuchchanges.

After the 2008 financial crisis, as some important economies reestablished

restrictions on their financial capital account as part of the policy toolkit,academics,policymakers,andinternationalinstitutionsalikehavebeentryingtoshed new light on such controversial issues. Thus, we next present Brazilianrecentpracticesandpolicyresponsesinthecontextofthecrisis.

TheBrazilianExperience

First,itisimportanttodistinguishthecurrentglobalscenariofromthatfacedbyLatinAmericaneconomies in the1990s.At that time, itwasobserved that theliquiditycurbwasmorerelatedtoweakdomesticmacroeconomicfundamentsinsuch economies than to international pressures caused by central banks inadvancedeconomies.Financialandcurrencycrisesassociatedwithdebtdefaultsused to be commonplace in the developing world. Those crises were usuallyexplained either bywrong domestic economic policies (crazy policymakers inthe developing economies) or bywrong economic agents (crazy agents underself-fulfilling features). Asian criseswere explained bywrongmarkets, to theresultofcontagiousandherdbehaviors(seeFrankelandWei2004).

Ontheotherhand,accordingtotheIMF(2012,6),"Capitalflowliberalizationhasbeenpartofthedevelopmentstrategyinseveralcountries,inrecognitionofthebenefitsthatsuchflowscanbring."Actually,itisanovelstrategyassociatedwith unconventional monetary policies developed by the central economiestrying to resume growth. That denotes a very different nature of capital flowsurgesexperiencednowadaysincomparisonwithpastpractices.It is"liquidityinjection" rather than "capital flows liberalization,"usuallyparsedas "removalofrestrictions."

This distinction makes quite a difference to policy recommendations,especiallywithrespecttohowEMeconomiesshoulddealwithit.Itisalsonotamatterof"unrestrictedconvertibilityoflocalcurrencyininternationalfinancialtransactions." (IMF 2012, 10) It seems that the recent capital surges aremoreassociated with the less convertible currencies combating devaluation ininternationalcurrencies.

The relationship between growth and the exchange rate (devaluations,

misalignments,andvolatilities)hasbeenapproachedin theeconomic literaturefromdifferent perspectives.' It is still a controversial topic, though it is fair tostatethatdevaluationsfostergrowththroughdifferentchannels,mainlythetradeone. Weak domestic markets in advanced economies result in an excessiveinventory of manufactured goods searching for international markets; thus,devaluationwouldbeveryhelpful.

Undersuchspecificcircumstance,thecapitalflowmanagementmeasuresputforwardinEMeconomiesarearationalresponse,andasbeneficialastheyareprudential.

The current context for implementing capital controls is quite different.Currently,problemsarecausedbycrazypolicymakers inadvancedeconomies.Unconventional monetary policy-including rude quantitative easing programsunderazero lowerboundmonetarypolicy-hasbeenurgedbycentralbanks inadvancedeconomies.Figures24.1and24.2showhowsizabletheexpansionofinternationalliquidityisasaresultofthispractice.

Figure24.1

ExpansionofInternationalLiquidity(US$trillion).

Notes:EuropeanCentralBank;BankofJapan;BankofEngland.

Source:Bloomberg,http://www.bloomberg.com.

Figure24.2

Expansion of the Monetary Base in Advanced Economies, 2007-2014 (US $billion).

Note:".BankofJapanestimatedinApril2013.

Source:Bloomberg.

As can be seen from these figures, central bankers in advanced economies,suchas theUnitedStates,Europe, and Japan,have introducedveryaggressiveexpansionarymonetarypolicies.Notwithstandingthebenefitsofsuchmonetarystances,EMcountriesareconcernedthatthesurgeincapitalinflowscouldcauseproblemsfortheireconomies.Exchangerateappreciation,reserveaccumulationwith some fiscal costs, and incentives to excessively borrow abroad, risking adomestic credit boom, are only some consequences that have recently beenobserved. It shouldbeadded that the international interest rates arepretty low

andhaveremainedsofora longperiodof time,especially incomparisonwithnormal interest rate levels inEMeconomies,whichconstitutesan incentive tomake loans abroad in international currencies. It can also amplify currencymismatches, with the well-known propensity for instability in EMs during asudden cessation of capital flows, sometimes leading to unexpected exchangeratedevaluations.

In theBrazilian case, an increase in international liquidity created excessivepressureontheexchangerate.Consequently,theeconomicauthoritiescreatedatransaction tax (IOF)onnewcapital flows.Table24.1 shows the evolutionoftheIOFonportfolioinvestmentsandexternal loans,whichcanbesummarizedasfollows(observationstoJune2013):

1. Portfolio investments include fixed income and derivatives. They were alltaxedata6percentrate,withexceptionofcapexandinfrastructurebonds(therehasneverbeenanIOFonexternalflowstothesebonds).Also,anIOFhasnotbeenappliedtoequities.

2.Short-term(uptoone-year)inboundloansandoffshorebondissues(overseasdebt)aresubjecttoanIOFatarateof6percent.

3.Therehasnotbeena1percent IOF taxonforeignexchangeshortpositionsheldbybanks,funds,andcompanies.

It is worth noting that when it became necessary to do so, the Brazilianauthoritypromptlywithdrewsuchmeasures, indicating that theyareadditionaltools tomanagecapital flows.Thatwas thecaseobserved inJune2013,whentheIOFappliedtoportfolioinvestments,includingfixedincomeinstrumentsandderivatives,wasreducedtozero.

Figure24.3

NominalExchangeRate(realperdollar),2008-2013.

Source:CentralBankofBrazil.

The appreciation path of theBrazilian exchange ratewas interrupted by thefinancial crisis of September 2008, which was reversed with the quantitativeeasing monetary policies adopted by developed economies. To reduce thevulnerabilityandprocyclicalityof capital flows,Brazil introducedmeasures tomanage its capital account through prudential regulation. These measures arealsoillustratedoverthetimeinfigure24.3.Fromthisfigureitispossibletoinferthataftercontrolsonsurgesofcapital inflowswereinitiated, theBrazilianrealrealizedanappreciationandmovedtoanewstableequilibrium.

Ascanbeseen in figure24.3, theeffectivenessof the restrictionsoncapitalcontrolhasbeenlongerthanexpectedbytheeconomicliterature.Resultsincludeamorestableexchangerateandlessvulnerablebalancesheets,bothindomesticcompaniesandinthefinancialsystem.Therearesomeplausibleexplanationsforsuchsuccesses.

First,theBrazilianeconomyismuchlessdollarizedthanitwasadecadeago.

The dollarization of an economy is strongly associated with both currentinflation and inflation risk.2 It is a rational response of agents to deal withinflation.Asinflationandtherisksofhighinflationhavedecreasedsubstantiallyin Brazil, along with sovereign risks, currency mismatches in balance sheetshave shown a sizable reduction. As a result, capital account managementmeasuresdonotchangeforeigninvestors'perceptionsaboutthecountry.

Second, the composition of the capital inflows changed very clearly afterrestrictions were placed on short-term financial inflows. As described in theliterature, restrictions on capital flows play an important role in capitalallocation,changingthecompositionofcapitalinflowsinfavorofforeigndirectinvestments. In Brazil, foreign direct investment has been as high as it wasbeforetherestrictionswereplacedoncapitalinflows.Meanwhile,asthereturnsonthecarry-tradestrategyturnedouttobenegative(seefigures24.4and24.5),short-termcapitalinflowshavebeenreduced.

Asshowninfigure24.6,longtermcapitalinflowsremainstableataveryhighlevel, while portfolio investment has decreased toward a level not seen sincebefore the 2008-2009 financial crisis. Neither intercompany transactions norequityhavebeenaffectedbysuchmeasures.

Third, Brazil has put forward a set of macroprudential policies, includingcapital account management, along with conventional monetary policy.Consistentfiscalresultshavebeenaccomplished;thus,theratioofpublicdebttoGDP has decreased very quickly. In addition, Brazil has left behind anyprobability of fiscal insolvency. The international investor's confidence in theBraziliansovereignbondshasincreased,asshowninfigure24.8.Accordingtothisfigure,thedifferencebetweentheyieldson10-yearBrazilianandUSbondshas narrowed consistently since 2011. A commitment to fiscal responsibilitythroughout the years, combined with economic growth, has contributed to areductionincreditriskinBrazil.

Finally, the medium-term growth prospect also plays an important role,encouraging the foreign direct investments to search for opportunities, andBrazil today is one of the best countries in theworld inwhich to invest. The

Brazilian government has just launched a comprehensive program ofconcessionsintheinfrastructuresector,includingairports,ports,railways,high-speedtrains,oil,gas,andelectricity,foratotalamountofUS$235billion.Itisjust the first step toward addressing the entire spectrum of the country'sinfrastructure needs.Many investment projects in different sectors, such as inautomobile, chemistry, health care, and others, are being set up.As amiddle-classsociety,Brazilhasadynamicdomesticmarketwithalowunemploymentrateandlowerincomeinequalities.

Figure24.4

Returns on Carry-Trade Strategies in Selected Countries, 2008 and 2013 (%,12monthaccumulated).a.June2008.b.June2013.

Source:Bloomberg.

Figure24.5

Brazil:ReturnsonCarry-TradeStrategy(2005-2013)(%perannum).

Source:Bloomberg.

Figure24.6

Brazil:FinancialAccounts,2008-2013(US$billion).


Figure24.7

Brazil:ForeignDirectInvestment,2008-2013.


Figure24.8

Yield"on10-yearBrazilianandUSBonds,1999-2013(%).

Notes:"Measuredbythedifferenceinyieldsonissuedateon10-yearBrazilianbonds denominated in US dollars and yields on US Treasury bonds (samematurity) traded on the secondary market on the same date. ""From October2012,yieldson10-yearBrazilianandUSTreasurybonds(samematurity)tradedonthesecondarymarketonthesamedate.

Source:MinistryofFinance,Brazil,andBloomberg.

It seems that growth prospects and investment opportunitiesmatter to foreigninvestorsmorethanrestrictionsinshort-termcapitalinflowsdo.

Summary

Excessive international liquidity has provoked side effects in EM economies,andBrazilismorearulethananexceptionamongthem.Tocopewiththeeffectof unconventional monetary policies implemented by advanced economies'central banks, Brazil has introduced a set of macroprudential policies that

includescapitalaccountmanagement,alongwithconventionalmonetarypolicy.Capitalaccountmanagementismuchmoreatechnicalissuethananideologicalone. Successful results have been obtained under the circumstances. Theeconomicpolicystanceincludesmeasurestofosterinvestmentsinthecontextofinflation under control, a growth-friendly fiscal consolidation program, a solidfinancialsystem,andstrongincome-inequalityreduction.

After restrictions on the country's financial account were implemented, theBrazilian real appreciated,moving upward until it stabilized at the parity of 2realsperUSdollar.Itisfairtosaythattheeffectivenessofthecontrols,begunin2011,haslastedlongerthanexpected.Thebenefitsofthecontrolsaregenerallygreaterthantheeventualcosts.Assuchmeasuresarealsoprudential,borrowingabroadinothercurrencyhasbeenlessleveraged.

Notes

1.SeeHollandandothers(2013)fornewempiricalfindingsontherelationshipbetween growth and exchange rate volatility. According to these authors, itseems that exchange rate volatilitymatters to growthmore thanmisalignmentdoes.

2.Financialdollarization is a topicassociatedwithhigh inflationand inflationrisk in Latin American economies, and Vieira, Holland, and Resende (2012)have associated such phenomenawith sovereign risks as dollarization remainshigh,evenafterinflationandinflationriskdecrease.

References

Baba,C.,andA.Kokenyne.2011."EffectivenessofCapitalControlsinSelectedEmerging Markets in the 2000s." IMF Working Paper 11/281, InternationalMonetaryFund,Washington,DC.

De Gregorio, J., S.Edwards, and R.O.Valdes. 2000. "Controls on CapitalInflows: Do They Work?" NBER Working Paper 7645, National Bureau ofEconomicResearch,Cambridge,MA,April.

Edison, H., and C.M.Reinhart. 2001. "Stopping Hot Money." Journal of

DevelopmentEconomics66(2):533-553.

Edwards,S.1999."HowEffectiveAreCapitalControls?"JournalofEconomicPerspectives13(4):65-84.

Frankel,J.,andShang-JinWei.2004."ManagingMacroeconomicCrises:PolicyLessons."Manuscript,HarvardUniversity,Cambridge,MA.

Holland,M.,F.V.Vieira,C.GomesdaSilva,andL.C.Bottecchia.2013."GrowthandExchangeRateVolatility:APanelDataAnalysis."AppliedEconomics45(26):3733-3741.

InternationalMonetaryFund(IMF).2012."TheLiberalizationandManagementof Capital Flows: An Institutional View." International Monetary Fund,Washington, DC, November 14. http://www.imf.org/external/np/pp/eng/2012/111412.pdf.

Kaminsky,G.L., andS.Schmukler. 2001. "Short - andLong-Run Integration:Do Capital Controls Matter?" Policy Research Working Paper 2660, WorldBank,Washington,DC.

Ostry,JonathanD.,AtishR.Ghosh,KarlHabermeier,MarcosChamon,MahvashS.Qureshi, and Denis B.S.Reinhardt. 2010. "Capital Inflows: The Role ofControls." IMF Staff Position Note 10/04, International Monetary Fund,Washington, DC, February 19. http://www.imf.org/external/pubs/ft/spn/2010/spn1004.pdf.

Reinhart,C.M.,andR.T.Smith.2002."TemporaryControlsonCapitalInflows."JournalofInternationalEconomics57(2):327-351.

Tobin, J. 1978. A Proposal for International Monetary Reform. EasternEconomicjournal4(3/4):153-159.

Vieira,FA.C.,M.Holland,andM.F.Resende.2012.FinancialDollarizationandSystemicRisks:NewEmpiricalEvidence. Journal of InternationalMoney andFinance31(6):1695-1714.

HeleneRey

BenefitsofFinancialIntegration:TheoryandEvidence

I will start by asking basic questions about international capital flows.Manypolicydiscussionsarebasedonthepremisethatinternationalcapitalflowsbringsome important benefits to countries' economies.When askedmore precisely,policymakers identify two main benefits of international capital flows:improvements inallocativeefficiencyand in risksharing.Becauseof financialintegration,capitalcanflowtoplaceswhereitisputtoitsmostproductiveuse,thatis,placeswherethemarginalproductofcapitalishighest.Thisviewcomes,ofcourse,straightfromtheneoclassicalgrowthmodel.

Policymakers would add that international capital flows are also beneficialbecausetheyenablebetterrisksharing,whichisagainastatementconveyedbymanyeconomicmodels.

EmpiricalEvidence

Numerous studies exist that actually try to look at the data for the effects ofinternationalcapitalflowsongrowthoronconsumptionvolatility,tryingtotestfor these two types of gains. Surprisingly, these effects are hard to find inmacroeconomic data: the benefits of capital flows are remarkably elusive. Asattestedbythemostrecentsurveysreviewingalonglistofempiricalpapers,itishardtofindrobustevidenceofanimpactoffinancialopennessongrowthoronimprovedrisksharing(e.g.,Eichengreen2002;Koseetal.2006;Obstfeld2009;Jeanne,Subramanian,andWilliamson2012).

Tobefair,somepaperspointtowardtheexistenceofthresholdeffects:Capitalflowswouldbebeneficialonlyafteracountryhasreachedacertainamountofinstitutionalorfinancialsectordevelopment.Therearealsosomedifferencesifonelooksacrossdifferenttypesofcapitalflows:Somecapitalflowsseembetter

atdeliveringgrowthand risk-sharingbenefits thanothers.But thisevidence isnotveryconclusivebecause,often,thesampleusedseemstomakeadifference.Somepapersusingmicroeconomicdatafindadecreaseinthecostofcapitalatthetimeoffinancialintegration.Sothequestionis,whydon'tweseemoreofaneffectattheaggregatelevel?

There is also some recent research analyzing the role of global banks andlooking atwhether the large international capital flows thatwe seewithin theinternal realm of global banks have had any effect on the real side of theeconomy. Cetorelli and Goldberg (2012) point toward a better allocation ofliquiditywithinglobalbanks.Butonemightwonderwhetherthiscomestogetherwith a weakening of the monetary policy transmission, as global banks canreshuffle liquidity across borders to offset the effect of national monetarypolicies. Ifglobalbankscanallocate liquidityamong theirvarioussubsidiariesand branches, that may have benefits, but it might also be a way ofcircumventingtheeffectsofmonetarypolicy.

So,fromthepointofviewoftheempiricalevidence, thejuryisstillout.Sofar, however, the evidence seems surprisingly less conclusive than what onemighthavethought,givenbothourstrongtheoreticalpriorsandthesheersizeofinternationalcapitalflowsintheworldeconomy.

CalibratedModels

The neoclassical growth model is behind many of our economic intuitionsregardingwhy the free flow of capital could be beneficial. Interestingly, evenwithin thatparadigm, realisticcalibrations indicate thatgains tend tobe small.Gourinchas and Jeanne (2006) have shown, in the context of small openeconomiesandinadeterministicsetting,thatgainsweresecondorder.Allthatinternational financial integrationdoes in that context is to speedup transitiontoward the steady stateof the economy.Coeurdacier,Rey, andWinant (2013)allowforuncertaintyandestimatewelfaregainsfromallocativeefficiencyandrisk sharing together, within the context of a general equilibrium neoclassicalgrowth model. They find that even in such a world, where the interactionbetween the precautionary savingsmotives and allocative efficiency effects is

modeled explicitly, welfare gains are small. Such a model can, in particular,generatetherealisticoutcomethatavolatileemergingmarketendsupexportingcapitalwhenitopensupitsfinancialaccount(unlessitisanextremelycapital-scarce country, far away from its steady state at the outset). So both on theempirical side and on the theoretical side it is hard to find support, at thisjuncture,forlarge,quantifiablebenefitsofinternationalfinancialintegration.

Iamnotnecessarilyclaimingthatbenefitstointernationalfinancialintegrationdonot exist,only that theyhavebeenelusive so far. In that light, itwouldbeusefultoidentifymorepreciselythechannelsthroughwhichcapitalflowsmaybebeneficial.Weshouldlookatmorespecifictypesofflows,andmorecloselyatpotentialeffectsontotalfactorproductivity.

CostsofFinancialFlows

Onthecostside,havinggone throughanumberofcrises inemergingmarketsand in advanced economies, we have some ideas about costs to internationalfinancialintegrationandcapitalflows.ReinhartandReinhart(2008)wroteaboutcapital flow bonanzas, that is, periods in which international liquidity isabundantandtherearelargecapitalflowsintoemergingmarkets,whichmaybesubjecttosharpreversal.

These largecapital flows tend tobecorrelatedwith inflation inassetprices.Thereisasurprisinglylargecommoncomponentinriskyassetprices(Miranda-Agrippino andRey 2012). In otherwords, althoughwemight think that riskyasset prices around the world are largely determined by specific countrymacroeconomicconditions, local conditions, thatwouldbewrong.There is animportantglobalfactor.

Associatedwith thesecapital flowbonanzashasbeenexcessiveappreciationofcurrencies,whichstrainedthecompetitivenessofthetradablessector.Withinthe euro area the loss of competitiveness of the periphery has been to someextentcausedbymassiveinflowsofcapital,whichhavebidupthepriceofrealestate.Thebankingsystemhaschanneledmassivecapitalflowsintoanumberofcountries, suchasSpainand Ireland, fueling realestate investmentbooms that

haveincreasednontradables'pricesandunitlaborcosts.

ChangesintheInternationalFinancialLandscape

Themainchangeintheinternationalfinanciallandscapeinthepast20yearshasbeenthetremendousincreaseincrossbordergrossassetflowsandpositions(seeLane and Milesi-Ferretti 2007; Gourinchas and Rey 2013). We need newframeworks,newwaysofthinkingthroughthebenefitsandcostsofintegrationthat take intoaccount the importanceofgross flows (inaddition tonet flows).That is to say, thinking about current account sustainability is not enough-weneedtoworryalsoaboutgrossflows.

For financial stabilitypurposes,gross flowsmatter, and theymattermoreastheexternalbalancesheetsofcountriesexpand.Intermsofthetransmissionofthe2008financialcrisistoEurope,forexample,thepositionoftheeuroareavis-a-vis the United States was roughly balanced; there was no current accountissue.ButthereweremassiveexposuresthroughthegrosspositionsofEuropeancountries. Their financial systemswere exposed to US toxic assets, exchangerate movements, and funding risk. This illustrates that there are potentiallymassivevaluationeffects(capitalgainsandlosses)whenexternalbalancesheetsarelarge.Thisisreallywhathaschangedinthepasttwodecadesandwhatweneed to take into account. It is no longer only about current account and netflows;itisalsoaboutgrossflowsandlargegrosspositionsatthefinancialsectorand country level. Risk transmission can be heightened through variouschannels, including a currency mismatch between assets and liabilities or amaturitymismatch.

ProcyclicalityofCreditFlows

Credit flowsareprocyclical (Committeeon InternationalEconomicPolicyandReform2012;BrunoandShin2013).Theygrewatafastrateinthe2003-2007precrisisperiod.Thereisapositivefeedbackloopbetweengreatercreditsupply,asset price inflation, and a compression of spreads. Smaller risk premiumsamplify the credit boom; asmeasured, the risk is lowandbalance sheets lookhealthier as asset prices go up. By relaxing constraints, this creates additional

spaceforlendingandforcredit,andsoon.

Thismechanismoccurswhenvalue-at-riskconstraintsoperateinthebankingsectors(AdrianandShin2012).Thisisamajorpositivefeedbackloopbetweencreditsupplyandriskspreads,onethatcontributestotheprocyclicalityofcreditflowsandtheirimportanceintherun-uptothecrisis.

ManagingBalanceSheets

In the presence of positive feedback loops, we need circuit breakers. Thetraditional feedback loop that has confronted policymakers is the following.Largecapitalinflowsintoagrowingeconomytendtocreateinflation,exchangerate appreciation, and expectation of inflation. In such a situation, the centralbank response is often to increase the interest rate to keep inflation undercontrol, but because yields are now higher, capital keeps flowing in and theexchangeratekeepsappreciating.Thispositivefeedbackloopjustifiedtheuseofcapital controls or, more broadly, capital flow management. Beside thistraditionalfeedbackloopisthisnewfeedbackloopdescribedabove,whichhastodowithcreditflowsandtheprocyclicalityofleverage.Highcreditflowsbidup asset prices, improve balance sheets, and lead to more flows and creditcreation.Thelargebalancesheetsbeingbuiltuphavetobemonitoredcarefully.Thisisallthemoreimportantbecausevaluationeffectscanbeofthesameorderofmagnitude as current accountmovements (Gourinchas,Rey, andTruempler2012).

Whenshouldoneintervene?Whenshouldoneactivatecircuitbreakerstocutthosepositivefeedbackloops?

Itisimportant,inmyview,nottowaittoolong;nottowait,forexample,forthequasi-certaintythatthereisabubbleinassetpricesorrealestatetointervene.Rather, one should continuously stress-test the balance sheet of the financialsectorandofthecountryandjudgewhetherlargebutrealisticchangesinassetpricescouldjeopardizefinancialstability.Ifso,macroprudentialinterventionorsome type of capital flow management intervention should take place. Iunderstandthedifficultyofdoingstresstestsingeneralandestimatingsecond-

roundeffectsinparticular,butdoingstresstestsonacontinuousbasis,evenifitisanimperfectprocess,isanecessarymonitoringtool.

What are the tools available for intervention? The choice betweenmacroprudential tools and capital management tools has to be somewhatpragmatic, depending on where the problems are and on the differentinstitutionalsettings.Macroprudentialtoolstendtobemoretargeted.Butcapitalcontrolsmaybeappropriateifthereisalotofdirectcrossborderlendingandthebankingsystemiscircumvented.

Conclusions

Weshouldnotforget in thiswholediscussionofcircuitbreakers that thereareusually important domestic distortions that interact with capital flows. Inpractice,forpoliticalreasons,weseemanysubsidiestoinvestmentinrealestate.Thesesubsidiesareinstrumentalincreatingtheinitialbubbleinrealestatepricesandinvestment.Byallmeans,thefirstthingtodoistoremovethesedistortions.It is also important to remember that excessiveborrowingbyacountrymeansthat someone else is lending excessively: Macroprudential policies apply tolendersjustaswellastheyapplytoborrowers.

I have discussed the use of capital flow management and macroprudentialtoolsfromanexantepointofview(topreventcrises),buttheremaybealso,insomecases, an important role for capital accountmanagementexpost (after acrisis).Forexample, capital controlscanbeused toavoidmajorcapital lossesforhouseholdsandcompaniesthatborrowedinforeigncurrencyandareheavilyexposed to further exchange rate depreciation. This type of ex post policyinterventionmayhavebeenusefulinacountrysuchasIcelandwheretherearelargeamountsofkrona-denominatedassetsinportfoliosofforeigninvestorsandwheremassive capital flight and large, ensuing depreciationwould have beenlikelyintheabsenceofcontrols(seeBaldurssonandPortes2013).

Butwehave tokeepinmindthat in thiscrisis,wehave todealwithclearlysubpar preventive policies, which have left us with a very difficult situation.Meanwhile,wemustreallythinkhardaboutbettergovernancelookingforward.

References

Adrian, T., and H.S.Shin. 2012. "Procyclical Leverage and Value at Risk."FederalReserveBankofNewYorkStaffReport338,FederalReserveBankofNewYork,NewYork.http://app.ny.frb.org/research/staff_reports/sr338.pdf.

Baldursson, E, andR.Portes. 2013. "Gambling for resurrection in Iceland: theriseandfallofthebanks."CEPRDP9664.

Bruno, V., and H.S.Shin. 2013. "Capital Flows, CrossBorder Banking andGlobalLiquidity."NBERWorkingPaper19038,NationalBureauofEconomicResearch,Cambridge,MA,May.http://www.nber.org/papers/w19038.

Cetorelli, N., and L.Goldberg. 2012. "Liquidity Management of US GlobalBanks:InternalCapitalMarketsintheGreatRecession."JournalofInternationalEconomics88(2):299-311.

Coeurdacier,N.,H.Rey,andP.Winant.2013."FinancialIntegrationandGrowthin a Risky World." Manuscript, London Business School, London; NBER,Cambridge,MA;andDepartmentofEconomics,SciencesPo,Paris.http://www.helenerey.eu.

Committee on International Economic Policy and Reform. 2012. "Banks andCrossBorder Capital Flows: Policy Challenges and Regulatory Responses."Brookings Institution, Washington, DC. http://www. brookings.edu/research/reports/2012/09/ciepr-banks-capital-flows.

Eichengreen, B. 2002. "Capital Account Liberalization: What Do theCrossCountryStudiesTellUs?"WorldBankEconomicReview15:341-366.

Gourinchas,P.-O., andO.Jeanne.2006. "TheElusiveGains from InternationalFinancialIntegration."ReviewofEconomicStudies73:715-741.

Gourinchas,P.-O.,andH.Rey.2013."ExternalAdjustment,GlobalImbalances,and Valuation Effects." In Handbook of International Economics, ed. GitaGopinath,ElhananHelpman,andKenRogoff.Elsevier:NorthHolland.

Gourinchas,P.-O.,H.Rey,andK.Truempler.2012."TheFinancialCrisisandtheGeography of Wealth Transfers." Journal of International Economics 88

(2):266283.

Jeanne, 0., A.Subramanian, and J.Williamson. 2012.Who Needs to Open theCapitalAccount?Washington,DC:PetersonInstitute.

Kose,M.A.,E.Prasad,K.Rogoff,andS.J.Wei.2009."FinancialGlobalization:AReappraisal."IMFStaffPapers56:8-62.

Lane,P.,andG.M.Milesi-Ferretti.2007."TheExternalWealthofNationsMarkII: Revised and Extended Estimates of Foreign Assets and Liabilities, 1970-2004."JournalofInternationalEconomics73(2):223-250.

Miranda-Agrippino, S., and H.Rey. 2012. "World Asset Markets and GlobalLiquidity."Manuscript,LondonBusinessSchool,London.http://www.helenerey.eu.

Obstfeld,M.2009."InternationalFinanceandGrowthinDevelopingCountries:WhatHaveWeLearned?"IMFStaffPapers56(1).

Reinhart,C.,andV.Reinhart.2008."CapitalFlowBonanzas:AnEncompassingView of the Past and Present." In NBER International Seminar inMacroeconomics, ed. Jeffrey Frankel and Francesco Giavazzi. Chicago:UniversityofChicagoPress.

GeorgeA.Akerlof

I learnedalotfromtheconference,andI'mverythankfultoall thespeakers.Ihavebeenaskedtogivemyoverallviewoftheconference.DoIhaveanimageofthewholething?Idon'tknowwhethermyimagewillhelpanybodyatall,butmyviewisthatit'sasifacathasclimbedahugetree.It'supthereand,oh,myGod,wehavethiscatupthere.Thecat,ofcourse,isthishugeeconomiccrisisthathasbeenuponussince2008.

Everyoneattheconferencehadsomeideasaboutwhatweshoulddotogetthepoor cat out of the tree.What I found sowonderful about the conferencewasthat the speakers all had their own respective images of the cat, with no twodiscussants of the same opinion. But occasionally those opinions meshed, inunexpectedandproductiveways.That'smyviewofwhatwasaccomplished. Iwill provide my own thoughts on the crisis and how well we've been doingrelative to the cat. My thoughts are a slightly different angle from whateverybodyelsehasbeensaying,fromtheirpervasivelydifferentvantagepoints.Iwillconcentrateon thepostcrisisUnitedStates,but theanalysisalsopertainsinternationally.

There is an excellent paper, by Oscar Jorda, Morris Schularick, and AlanTaylor(2011),whichdividedupdownturnsfrom1870to2008for14advancedcountries into financial recessions and normal recessions.They looked at howGDPrecoveryvariedinseverityaccordingtocreditoutstandingrelativetoGDPin theprecedingboom.And their conjecturewas stronglyconfirmed:notonlyare financial recessionsdeeper and slower in recovery thannormal recessions,theyalsohaveslowerrecoverythegreaterthecredit-to-GDPratiois.

Thatisthehistory.

Howdo their findings reflecton thecurrent crisis?Curiously, itdependsonthemeasurementofcreditoutstanding.Withbankloanstotheprivatesectorasthemeasure of credit, the United States' recovery is about 1 percent of GDPbetter than the mean recovery for financial recessions. When in addition themeasureofcreditalsoincludescreditgrantedbytheshadowbankingsystem,theUnitedStates' recovery is about 4 percent better than themedian recovery forfinancial recessions.Thegraphs in thepaperbyJordaandothers illustrate thispoint.

Butwiththeonsetoffinancialderivatives,wehavenowayofknowinghowtomeasure "credit." If derivatives are used to hedge risk, thenwewould expectderivativestosoftenthecrash.

Forexample, if thebuyer-rather than theseller-ofacreditdefaultswapgoesbankrupt in the event of a defaultwewould expect the credit default swap tosoften the crash. On the other hand, if we think that derivatives escalategambling,thenwewouldexpectthemtoexacerbatethecrash.Theconventionalinterpretationofthe2007-2008crashintheUnitedStatesholdsthatderivativesenhanced gambling in a differentway. In parable, derivatives allowed a daisychainofescalatingvaluationofmortgages-theyweremade,forexample,intheCentral Valley on the shadiest of bases, then passed through into derivativepackages,whichwere ratedA and higher. Thiswas an environment inwhichjunkdidnotaffectratings.Somortgageoriginatorshadnoincentivetorequiredownpayments or borrower credibility.To agreat extent, theydidn't. In theircreation and ratings of derivatives, the investment houses and the ratingsagencieswereminingtheirreputationsasfiduciaries.Thisadditionalroleofthederivatives suggests that ameasure of credit basedon loans outstanding, eventakingintoaccounttheroleoftheshadowbanks,yieldsaconservativemeasureofthebenchmarkforwheretheUnitedStatesshouldnowbe.

Thatviewalsoconformswiththecommonperceptionsfromthefallof2008.At that time, theGreatDepressionwas thebenchmark forwhatwouldhappenwithout government intervention. From that vantage point, macro policy has

beennot justgood,but trulyexcellent.AlanBlinder's fantasticbook,After theMusicStopped(2013),saystheexactsamething.

Almost every program has been close to what the doctor called for. Thosemeasuresincludethefollowing:

EconomicStimulusActof2008

BailoutofAIG

RescueofWashingtonMutual,Wachovia,andCountryWidebyadoption

TroubledAssetReliefProgram(TARP)

StresstestsrunbytheUSTreasuryandtheFed

Declinesininterestratestoclosetozero

AmericanRecoveryandReinvestmentActof2009

Bailoutoftheautoindustry

InternationalcooperationinthespiritoftheG20meetinginPittsburgh,atwhichtheIMFplayedaleadingrole

Thereisonlyonemajorcriticismofthepoliciesputinplace.Weshouldhaveledthepublictounderstandthatweshouldmeasuresuccessnotbythelevelofthecurrentunemployment ratebutbyabenchmark that takes intoaccount thefinancial vulnerability that hadbeen set in thepreviousboom.Weeconomistshave not done a good job of explaining that our macrostability policies havebeeneffective.Thereis,ofcourse,agoodreasonwhythepublichasahardtimelistening.Theyhaveotherthingstodobesidesbecomingmacroeconomistsandmacroeconomichistorians.

But just a bit of common sense indicates why the policies have been sosuccessful.IfLehmanBrothershadbeen$1intheredanditneededtobeonly$1intheblacktostayoutofbankruptcycourt,theexpenditureofonly$2atjust

therightcrisismomentcouldhavesavedus fromaGreatDepression.That$2fingerinthedikewouldhavebeenallthatwasneeded.

The expenditures for the bailout were, of course, more than $2; they willprobablybepositiveandruntoafewbilliondollars.Buttheydidliterallystopthe financialmeltdown thatwas inprogress.Relative to the tensof trillionsofGDP that would have been lost with a repeat of the Great Depression, thesavingsfromTARPareoftheorderofmagnitudeof1,000to1.Figuratively,itisfairtocallthisafingerinthedike.

Theexpendituresbyboth theBushand theObamaadministrationson fiscalstimulushaveachievedlessbangforthebuck,butalmostsurelytheyhavebeeneffective. Current estimates of government expenditure multipliers aresomething like 2. That number also makes intuitive sense. Liquidity-trapestimates of a balanced budgetmultiplier are approximately 1, both in theoryandinestimation,andthetaxmultiplierisrobustlymeasuredasapproximately1. The government expendituremultiplierswill be the sum of the two, so thestimulusbillshavealmostsurelyalsohadsignificantpayoff.

Insum,weeconomistsdidabadjobofpredictingthecrisis,butthepostcrisiseconomic policies have been close to what a sensible economistdoctor wouldhave ordered. Those policies have come directly from the Bush and Obamaadministrations, and from their appointees. They have also been supported byCongress.

The lesson for the future is that good economics and common sense haveworked well-we have had trial and success.Wemust keep this inmind withpolicygoingforward.

References

Blinder,AlanS.2013.AftertheMusicStopped.NewYork:PenguinPress.

Jorda,Oscar,MoritzH.P.Schularick,andAlanTaylor.2011."WhenCreditBitesBack: Leverage, Business Cycles and Crises." NBERWorking Paper 17621,NationalBureauofEconomicResearch,Cambridge,MA,November.

OlivierBlanchard

TheIMF'ssecondconferenceonrethinkingmacroeconomicpolicyinthewakeofthe2008-2009economiccrisishasunderscoredthemanychallengesthat lieahead forpolicymakers.Rethinkingand reformsareboth takingplace.Butwestill do not know the final destination, be it for the redefinition of monetarypolicy, or the contours of financial regulation, or the role of macroprudentialtools.We have a general sense of direction, butwe are largely navigating bysight.

In this chapter I review six examples raised at the conference to underscoreourlackofknowledgeconcerningthe"correct"trajectorytotakeinthecomingyears,andthepossibleoutcomesofvariousregulatorymoves.(Moredeveloped,though preconference, thoughts are given in the introductory chapter, writtenwithGiovanniDell'AricciaandPaoloMauro.)

FinancialRegulation

There isnoagreed-uponvisionofwhat thefuturefinancialarchitectureshouldlook like and, by implication, no agreed-upon vision of what the appropriatefinancial regulation should be. Here I am reminded of Paul Volcker's famousobservationthattheonlyusefulfinancialinnovationofthelast40yearshasbeentheATM.Thisissurelyanexaggeration.Butwearestillunsureabouttherightroleofsecuritization,therightscopeforderivatives,theroleofmarketsversusbanks,andtheroleofshadowbankingversusbanking.

Still,itseemsclearthatsomethingsshouldchange,andindeed,policymakersare putting in place measures in the context of international or nationalinitiatives.Oneexampleistheincreaseinrequiredcapitalratios.Itmaynotbeapanacea, but it surely can make the financial system more robust. Even so,however, I am struck by the level of uncertainty and disagreement about the

effectsofcapitalratiosonfundingcostsandthusonlending.Reasonablepeople,asamong themMartinHellwigandAnatAdmati,argue thatwearenot so farfrom the Modigliani-Miller world, and banks can afford substantially highercapitalratios.Others(andnotonlybankers)arguethatsuchratioswouldinsteaddestroythebankingindustry.

Another example is capital flows and, by implication, the role of capitalcontrols. Iwas struck byHeleneRey's discussion in chapter 25, inwhich sheshowshowsurprisinglymeagerthehardeconometricevidenceisforthebenefitsof portfolio flows. I was also struck by Stanley Fischer's rhetorical question,whatistheusefulnessofshort-termcapitalinflows?Clearly,howwethinkaboutthe scope of capital controls depends verymuch on the answer to these basicquestions.

TheRoleoftheFinancialSector

Ithasbecomeaclichetosaythatmacroeconomicthinkingunderstatedtheroleof financial factors in economic fluctuations.Much analytical work has takenplaceoverthepastfiveyearstoreintroducethefinancialsystemintoourmodels.But we are not there yet. For example, is there a credit and financial cycleseparatefromthebusinesscycle,asClaudioBoriosuggests?Orshouldwethinkoffinancialshocksasanothersourceofdisturbanceandthefinancialsystemasjustanothersourceofamplification?

Was Stephan Gerlach right when he asked whether we should reallyreconsider all of macroeconomics for an event that may happen once everyhundred years? Or, instead, are financial shocks and the financial system socentral to macroeconomic fluctuations that the IS-LM modelwhich does notinclude an explicit financial systemis not an acceptable port of entry intomacroeconomics?

By implication, there is no agreement on how or evenwhether to integratefinancialstabilityandmacrostabilityintothemandateofcentralbanks.Doesitrequire a tweak to inflation targeting or much more radical rethinking? Theintellectually pleasant position is to argue thatmacroprudential toolswill take

care of financial stability, so monetary policy can still focus on its usualbusiness: inflation targeting. I read, perhaps unfairly, Michael Woodford'sdiscussioninchapter4assuggestingthatthecrisisshouldleadustoshiftfrominflation targeting to nominal income targeting, without a major emphasis onfinancialstability.Iamskepticalthatthisistherightanswer.Ithinkwehavetoberealisticabouttherolethatmacroprudentialtoolscanplayandaboutthefactthatmonetarypolicycannotignorefinancialstability.Thisbringsmetomythirdpoint.

MacroprudentialTools

At our first conference on rethinking macroeconomic policy, in 2011,macroprudentialtoolswere,touseAndrewHaldane'sphrase,verymuchthenewkid on the block. It was clear that the two standard tools, fiscal policy andmonetarypolicy,werenot therightones todealwithfinancial imbalancesandrisks.Thequestionthenwaswhethermacroprudentialpolicywasgoingtobethethirdlegofmacroeconomicpolicyorjustacrutchtohelpthefirsttwo.

We do not have the answer yet. But asmore andmore countries are usingthosetools,wearelearning.Idrawtwolessonsfromtheevidenceandfromtheworkpresentedinthisvolume.

First, these toolswork, but their effects are still hard to calibrate, andwhenused,theyseemtohavemoderatedratherthanstoppedunhealthybooms.ThisisalsomyreadingofBankofKorea'sGovernorKim'spresentationinchapter8.

Second,by theirnature, theyaffect specific sectorsandspecificgroups, andthis raises political economy issues. This is clear from Stanley Fischer'sdiscussioninchapter7oftheuseofloan-to-value(LTV)ratiosinIsrael.

GovernanceandAllocationofTasksamongMicroprudential,Macroprudential,andMonetaryPolicy(or,asAvinashDixitcalledthem,MIP,MAP,andMOP)

Howshouldmicroprudentialandmacroprudentialregulationbecoordinated?Itissometimessaidthattheyarelikelytoconflict.Conceptually,Idonotseewhythey should: I see macroprudential regulation as simply taking into account

systemiceffectsandthestateoftheeconomyinthinkingaboutbankregulationandthesituationofeachfinancialinstitution.

Forexample, I seemacroprudential regulation requiringhighercapital ratiosfrom more systemically important banks or when aggregate credit growthappearstoohigh.Thequestionishowtoworkoutthedivisionoflaborandtheinteractionsbetweenthetwosothatthisisindeedwhathappens.

If not done right, it might mean that as a bust starts, the microprudentialsupervisorignoressystemicaspectsandothereventsandasksforhighercapitalratios, while the macroprudential supervisor rightly believes the opposite isneeded. The United Kingdom's approach-the creation of a Financial PolicyCommitteethatcanimposecapitalratiosthatvaryovertimeandacrosssectorsto maintain financial stability-seems like a good way to proceed. AndrewHaldanediscussestheissuesinchapter5.

How macroprudential regulation and monetary policy should be combinedraisesmorecomplex issues.There is littlequestion thateachaffects theother:monetarypolicyaffects risk taking,andmacroprudential toolsaffectaggregatedemand.Sopolicymakersneedtocoordinate.

Given that monetary policy surely must stay with the central bank, thissuggestsputtingbothofthemunderoneroofatthecentralbank.Butthisinturnraises the issue of central bank independence. It is one thing to give the bankindependencewithrespecttothepolicyrate;itisanothertoletitsetmaximumLTV ratios and debt-to-income ratios.At some point, the issue of democraticdeficitarises.

Maybe the solution is not so hard, namely, to give various degrees ofindependencetothecentralbank.StanleyFischergaveusamarvelousanalogyandpointedus toward the solutionwhenhe said that anybodywho ismarriedeasily understands the notion of various degrees of independence. Again, theUK's approach, with its two parallel committees within the central bank, onefocusingonmonetarypolicy,theotheronfinancialpolicywithalimitedsetofmacroprudential tools (not including, for example, LTV ratios), seems like a

reasonableapproach.

TheSustainableLevelofDebt

Therateoffiscalconsolidationdependson,amongotherthings,whatwethinkasustainablelevelofdebtis.Manycountrieswillbemanaginglevelsofdebtcloseto 100 percent of GDP for many years to come. There is a standard list oftextbookanswersastowhyhighdebtiscostly,fromlowercapitalaccumulationtotheneedforhigheranddistortionarytaxes.Isuspectthecostsareelsewhere.Iseetwomaincosts.

Thefirst isdebtoverhang.Thehigher thedebt, thehigher theprobabilityofdefault,thehigherthespreadongovernmentbonds,andtheharderitisforthegovernment to achieve debt sustainability.But the adverse effects do not stopthere. Higher sovereign spreads affect private lending spreads, which in turnaffectinvestmentandconsumption.Higheruncertaintyaboutdebtsustainability-and thus about future inflation and future taxation-affects all decisions. I amstruck by how limited our understanding is of these channels. Reduced formregressionsofgrowthondebtcantakeusonlysofar.

Thesecondrelatedcostistheriskofmultipleequilibria.Athighlevelsofdebttheremaywellbe twoequilibria:a"goodequilibrium,"atwhichratesare lowanddebtissustainable,anda"badequilibrium,"atwhichratesarehighand,asaresult,theinterestburdenandtheprobabilityofdefaultarehigher.Whendebtisveryhigh,itmaynottakemuchofachangeofheartbyinvestorstomovefromagoodtoabadequilibrium.

Isuspect that thisphenomenonispartlybehindtheItalianandSpanishbondspreads.Inthiscontext,MartinWolfinchapter20asksaprovocativequestion:whyarethespreadssomuchhigherforSpainthanfortheUK?Debtanddeficitsare actually slightly lower in Spain than in the UK. No doubt, the overalleconomicsituationofSpainisworsethanthatoftheUK,butdoesthisexplainfully the difference in spreads? Could the answer lie in the difference inmonetarypolicy?IntheUK,investorsexpecttheBankofEnglandtointerveneifneededtomaintainthegoodequilibrium,whereastheybelievetheEuropean

Central Bank (ECB) does not have the mandate to do so. These are centralquestionsthatweneedtostudymore.

MultipleEquilibriaandCommunication

In a world of multiple equilibria, announcements can matter a lot. Take, forexample,thecaseoftheOutrightMonetaryTransactionsprogramannouncedbythe ECB. The announcement of the program can be interpreted as havingremovedoneofthesourcesofmultipleequilibriainthesovereignbondmarkets,namely, redenomination risk, or the danger that investors, assuming that acountryontheperipherywouldleavetheeuro,askforalargepremium,therebyforcing exit from the euro in the process. The announcement has succeededwithouttheprogramactuallyhavingtobeused.

From this viewpoint, the recent announcement by theBank of Japan that itintendstodoublethemonetarybaseisevenmoreinteresting.Whateffectitwillhave on inflation depends very much on how Japanese households and firmschangetheir inflationexpectations.If theyrevisethemup, thiswillaffect theirwage and price decisions, and lead to higher inflation-which is the desiredoutcomeintheJapanesedeflationcontext.Butiftheydonotrevisethem,thereisnoreasontothinkthatinflationwillincreasemuch.

Themotivation for this dramaticmonetary expansion is primarily to give apsychologicalshockand toshiftperceptionsandpricedynamics.Will itwork,togetherwith theothermeasures takenby the Japaneseauthorities?Let'shopeso.Butweareveryfarfromthemechanicaleffectsofmonetarypolicydescribedinthetextbooks.

AlthoughIhavehadtosparementionofmanycontributionsandinsightsfromtheconferencecapturedelsewhereinthisvolume,theconferencehasleftuswithaclearresearchagenda.WeattheIMFfullyintendtotakeupthechallenge.

DavidRomer

AsIlistenedtothepresentationsanddiscussions,Ifoundmyselfthinkingaboutthe conference from two perspectives. One is intellectual: Did we askprovocativequestions?Wereinterestingideasproposed?Werewetalkingaboutimportant issues? By that standard, the conference was very successful: thecontributionsanddiscussionswereextremelystimulating,andI learnedagreatdeal.

Thesecondperspective ispractical:Wheredowestand in termsofavertinganother financial andmacroeconomicdisaster?By that standard, I fearwe arenotdoingnearlyaswell.AsIwilldescribe,myreadingoftheevidenceisthatthe events of the past few years are not an aberration, but just an extrememanifestationofabroaderpattern.Andtherelativelymodestchangesofthetypediscussedattheconference-andthatpolicymakersareputtingintoplaceinsomecases-are helpful but unlikely to be enough to prevent future financial shocksfrominflictinglargeeconomicharms.

Thus, I believe we should be asking whether there are deeper reforms thatmighthavealargeeffectonthesizeoftheshocksemanatingfromthefinancialsectororontheabilityoftheeconomytowithstandthoseshocks.Buttherehasbeenrelativelylittleseriousconsiderationof ideasforsuchreforms,not justattheconferencebutinthebroaderacademicandpolicycommunities.

TheFinancialSectorasaContinuedSourceofShocks

Myviewthatweshouldthinkoffinancialshocksasclosertocommonplacethantoexceptionalisbasedonhistory.ConsidertheUnitedStatesoverthepast30orsoyears.Bymycount,therehavebeensixseparatetimesoverthatperiodwhenfinancialdevelopmentsposedimportantmacroeconomicrisks.Inthreeofthem,

the riskswere largelyavertedand thecostsendedupbeingminor. In two, thecostsweremodesttomoderate.Andinone,thedamagewasenormous.

Concretely:

•InthethroesoftheVolckerdisinflationintheearly1980s,thecombinationofthesevere recessionandbanks' exposure toLatinAmericandebtcausedmanymajorbankstobeinserioustrouble.Itwasonlyalast-minuteturninpolicyandthe willingness of regulators to ignore the banks' extremely shaky financialcondition for a fewyears that kept the financial system from falling apart. Sothatwasadangeraverted.

•The1987stockmarketcrashwasasignificant financialshock,but rapidandhighlyvisibleresponsesbytheFederalReservetokeepmarketsfunctioningandreduceinterestratesagainpreventedlargedamagetotheeconomy.

•Thesavingsandloancrisisofthelate1980sandearly1990sdidsomedamagetotheeconomythroughmisallocationof investmentandimpairedlending,andsomewhatmoredamagetothegovernmentbudgetthroughdirectbailoutcosts.

•TheRussiandebt crisis and the collapseofLong-TermCapitalManagement(LTCM) in1998causedcentralbankers somesleeplessnightsas theyworriedaboutthestabilityoftheworldfinancialsystem.StabilitywaspreservedthroughthearrangedrescueofLTCM,lowerinterestrates,andotheractions.Thatisthethirdcaseinwhichthedangerwasaverted.

• The dot-com bubble and bust of the late 1990s and early 2000s caused aconsiderablemisallocationofinvestmentand,moreimportant,arecession.

•And,obviously,wehadthehousing-pricecollapseandfinancialmeltdownofthepastfewyears,whichhavehadcatastrophiceffects.

Inlightofthatrecordforjustonecountryoverathirdofacentury,theideathatlarge financial shocks are rare, and thatwe therefore shouldnotworrygreatlyaboutthem,seemsfundamentallywrong.

WhatIfindstrikingaboutthislistisnotjustitslengthbutitsvariety.AndifyoulookoutsidetheUnitedStates,itiseasytofindexamplesofotherkindsoffinancial shocks.Yousee IcelandandCyprus,where the financial shockcamefrom a vastly expanded banking sector with huge foreign deposits. You seeGreece,wheretheproblemwasdisguisedfiscalprofligacy.Youseetheclassicsuddenstops.And I amsure thatwitha littlemorework,youcouldaddevenmoretypesoffinancialshockstothelist.

Inshort, therangeofpotentialfinancialshocksislongandvaried.Thereareonlyafewonmyillustrative listofdomesticandforeignfinancialshocks thattook the form of big run-ups in asset prices followed by some kind of crash.Indeed, there areonly two, thedot-comepisodeand the recent crisis, thatonecouldreasonablycall"bubbles."SoI think therightconclusion todrawis thatfinancial shocks are likely to be both frequent and hard to predict, not just intheirtimingbutalsointheirform.

Small-ScaleSolutions

Thequestion,then,iswhattodo.Letmestartwithtwosmall-scalepolicies,oneofwhichIthinkislargelyanonstarterandoneofwhichIthinkwillbehelpfulbutveryfarfromacompletesolutiontotherisksoffuturecrises.

The nonstarter is using the short-term policy rate as a tool for dealingwithfinancial imbalances and risks. Even if that were the only objective we wereusing the policy rate for, it is much too crude. Often the concern with thefinancialsysteminvolvesapotentialprobleminonepartoffinancialmarkets,ordifferenttypesofproblemsindifferentmarkets.Insuchsituations,asingletoolthataffectsallmarketsisoflimitedvalue.Indeed,asJanetYellenpointedoutatthe conference, often it is not even clear which direction you would want tomovethepolicyratetoaddressapotentialfinancialrisktotheeconomy.And,ofcourse,wewanttouseitforotherveryimportantpurposesaswell.Sowecandebatewhether there is a little bit of benefit to taking financial developmentsmore into account in the setting of interest rates, but at best it can improveoutcomesonlymarginally.

The type of small-scale policy that I think is more promising is the oneadvanced in the discussions of macroprudential policies and capital accountmanagement. The positive way to put it is that it is the wise central bankermodel; the negativeway to describe it is that it is theWhac-A-Mole strategy.Regardlessofhowonelabelsit,theideaistouseregulationsandinterventionscreatively to address potential problems as they develop. For example, if youthink a bubble is developing in the real estate market in Seoul, you adoptregulationsdirectedspecificallyatmortgagesinSeoul.

Iwas very impressedwith the descriptions of policymakers' actions in suchcountriesas Israel,Korea,andBrazil indealingwithawiderangeof financialdevelopments,andsomethingIlearnedfromtheconferenceisthatsuchtargetedactionsareausefuladdition to thepolicy toolkit.But in lightof theenormousrangeof potential financial shocks, the idea thatwe can stabilize the financialsystembycountingonverysmartpolicymakerstoperceiveeachproblemasitisdeveloping and design a specific intervention to target it quickly is surelywishfulthinking.

What I take from this is that we need to be thinking more broadly andcreatively, looking for more fundamental solutions rather than particularinterventions.Atagenerallevel,thesecantaketwoforms.

DeeperSolutionsontheFinancialSide

Thefirstapproachistoreformthefinancialsystemsothattheshocksitsendstotherealeconomyaremuchsmaller.Thediscussionofmicroregulationshowedthattherearepromisingideasinthatarea.HereIamthinkingofstrongercapitaland liquidity requirements, special rules for institutions that create moresystemic risk, and restrictions on the form or capabilities of what financialinstitutionscando,suchasring-fencingintheUnitedKingdomandtheVolckerrule in the United States. Those approaches are broader than responding toindividualproblemsastheyarise,andtheyappearpromising.

But at the end of the day, it is hard to believe that the relatively modestchangesalongthesedimensionsthatwerediscussedattheconferencearereally

bigenoughtogiveusafinancialsystemthatissorobustthatitisnotgoingtoperiodically cause severe problems. Shadow financial institutions may escapethe rules altogether; rules canbegamed; and shocks canbe so large that theyoverwhelmthemoderatechangesthatwereconsidered.

Thus, Iwas disappointed to see little consideration ofmuch larger financialreforms.Letmegivefourexamplesofpossibletypesoflargerreforms:

• There were occasional mentions of very large capital requirements. Forexample,AllanMeltzernoted thatatone time25percentcapitalwascommonforbanks.Shouldwebemovingtosuchasystem?

•AmirSufiandAdairTurnerdiscussedthefeaturesofdebtcontractsthatmakethem inherently prone to instability. Should we be working aggressively topromotemore indexation of debt contracts,more equity-like contracts, and soon?

•Wecanseethecoststhatthemodernfinancialsystemhasimposedontherealeconomy. It is not immediately clear that the benefits of the financialinnovations of recent decades have been on a scale that warrants those costs.Might amuch simpler, 1960s - or 1970s-style financial system be better thanwhatwehavenow?

•The fact that shocks emanating from the financial system sometimes imposelarge costs on the rest of the economy implies that there are negativeexternalities to some types of financial activities or financial structures. Thissuggests thepossibilityofPigoviantaxes.So,should therebesubstantial taxeson certain aspects of the financial system? If so, what should be taxed-debt,leverage, size, other indicators of systemic risk, a combination, or somethingelsealtogether?

I do not know the answers to these questions, but it seems tome that theydeserveseriousanalysis.Yetradicalredesignofthefinancialsystemwaslargelymissingfromtheconference.

Larger-ScaleSolutionsontheMacroeconomicSide

Theotherwaytomakelargechangesistotrytomakethemacroeconomymoreresilienttofinancialshocks.Ithoughtthelackofdiscussionofpossiblechangesinthisdimensionwasthelargestgapintheconference.Letmediscussthisissuein three areas of macroeconomic policy: measures to deal with shocks to acommoncurrencyarea,monetarypolicy,andfiscalpolicy.

Withregardtoacommoncurrencyarea,imaginethatatsomepointinthenottoodistantfuture,theeuroareaishitwithanotherlargefinancialshockthathasasymmetriceffectsacrossdifferentcountries.Arethingsgoingtoplayoutverydifferentlythantheyhaveoverthepastfewyears?

Therewouldsurelybefewerlate-nightmeetings,becausepolicymakershavelearned more about how to do short-term crisis management. But I see littleprogress towardmeasures thatwouldcause fundamentalchanges in theeffectsthe shock would have. Policymakers have taken, at most, baby steps towardaddressingtheinstabilitiescreatedbythefactthattheresponsibilityforcleaningupinsolventbanksisatthelevelofindividualcountriesratherthanoftheeuroarea as awhole. And even less has been done in terms of a fiscal union andmechanismstodealwithlargedifferencesincompetitiveness.

Concerningmonetarypolicy,inflationtargetingappearedtobeanalmostidealframeworkforitsfirst15or20years.Butwehavenowhadanextendedperiodduringwhichithasshownitselfincapableofprovidingaggregatedemandatthelevel that iswidely recognized to have been needed. So it seems important tothink aboutwhetherwe should have a different approach tomonetary policy.But again,we have not gotten very far.The idea of targeting a nominalGDPpath has been mentioned on and off for a few years, but the debate has notproceeded to serious quantitative analysis of its costs and benefits and ofwhetheritcouldmaketheeconomysubstantiallymoreresilient.Otherideasforsignificantchangesinthemonetarypolicyframeworkhavebeendiscussedevenless.

With regard to fiscal policy, the biggest idea that has achieved substantialsupport is that itwouldbedesirable tohavemorefiscalspace.Buthowtogetfromhere to there, given the challenges of just getting back to the amount of

fiscalspacewehadbeforethecrisis,isahardissue,andoneonwhichprogresshasbeenminimal.Andinlightoftheterribleproblemsthathaveafflictedsomecountriesthatenteredthecrisiswithveryresponsiblefiscalpolicies,fiscalspaceisclearlynotamagicbullet.

Iheardvirtuallynodiscussionoflargerchangestothefiscalframework.Thepossibility of measures to make automatic stabilizers stronger (for example,through macroeconomic triggers for changes in fiscal policy) was notmentioned. And the status of this idea in the broader policy communityresembles the status of targeting a nominal GDP path: the idea is mentionedfromtimetotime,butthediscussionhasnotproceededtothepointofconcreteproposalsandquantitativeevaluation.

Anotherfiscalideathathasreceivedlittleattention,eitherattheconferenceorin the broader policy debate, is the idea of fiscal rules or constraints. Forexample, one can imagine some type of constitutional rule or independentagency(oracombination,withaconstitutionalruleenforcedbyanindependentagency)thatrequireshighlyresponsiblefiscalpolicyingoodtimesandprovidesa mechanism for fiscal stimulus in a downturn that is credibly temporary.RobertoPerottiandAvinashDixitraisedtheideaoffiscalrulesorcouncilsverybriefly,butitgotnofurtherthanthat.

Thefactthatwearemakingsolittleprogressintermsoflargerchangestoourapproaches tomacroeconomicpolicyappears to furtherstrengthen thecase forthinking about deeper financial reforms. But I also think we need broaderthinkingaboutthemacroeconomicside.

Conclusion

After five years of catastrophic macroeconomic performance, "first steps andearly lessons" (to quote the conference's subtitle) is not what we should beaiming for.Rather,we should be looking for solutions to the ongoing currentcrisis and strong measures to minimize the chances of anything similarhappeningagain.Iworrythatthereformswearefocusingonaretoosmalltodothat,andthatwhatisneededisamorefundamentalrethinkingofthedesignof

ourfinancialsystemandofourframeworksformacroeconomicpolicy.

JosephE.Stiglitz

In analyzing the financial crisis that began in 2007 and led to the GreatRecession,weshouldtrytobenefitfromthemisfortuneofrecentdecades:Theapproximately 100 crises that have occurred during the last 30 years-asliberalization policies became dominant-have given us a wealth of experienceand mountains of data. If we look over a 150-year period, we have an evenricherdataset.

Withacenturyandahalfofclear,detailed informationoncrisisaftercrisis,theburningquestion isnotHowdid thishappen?butHowdidwe ignore thatlonghistory,andthinkthatwehadsolvedtheproblemswiththebusinesscycle?Believing thatwehadmadebigeconomicfluctuationsa thingof thepast tookremarkablehubris.

MarketsAreNotStable,Efficient,orSelfCorrecting

Thebig lesson that thiscrisis forciblybroughthome-oneweshouldhave longknown-is that market economies, on their own, are not necessarily efficient,stable, or selfcorrecting. One of the reasons there were not only failures inpreventing and forecasting the downturn but also in responding to it,was thatmany of the predominant models employed special assumptions, leading toviewsthatmarketswereefficient,stable,andselfcorrecting.Becauseourmodelsdidn'tadequatelyanalyzethecausesofthecrisis,wecouldneitherrespondtothecrisisinwaysthatwouldensureaquickandstrongrecoverynortakeappropriateactions tosignificantly reduce the likelihoodofa recurrence.The result is thatwecontinuetofaceasignificantriskofanothercrisisinthefuture.'

That unfettered markets are not, in general, efficient whenever there areinformationimperfectionsandasymmetries,and/orwhenthereisanincomplete

setofriskmarkets,and/orwhencapitalmarketsareimperfectorincompletehasbeenwellestablishedformorethanthreedecades.These"marketimperfections"areimportantineveryeconomy,eventhemostadvanced;andyetmanyofourregulators and the advocates of deregulation ignored not only the lessons ofhistory but also these advances in our understandings of the limitations ofmarkets.

Moreover, predominant macroeconomic models before the crisisunderestimatedmarketinstability.Theyhadfocusedonexogenousshocksasthesourceoftheperturbationsgivingrisetofluctuations,andyetit'sveryclearthataverylargefractionoftheperturbationstooureconomy-includingthosegivingrise to the worse downturns-are endogenous. The housing bubble and itsbursting, like somany bubbles that preceded it, was a creation of themarketitself. The models that focused on exogenous shocks simply misled us-themajority of the really big shocks come from within the economy. Moreover,someof themost important shocksarepersistentandassociatedwith long-runstructural transformations-closely linked with the economy's (endogenous)innovativeactivities.

Finally,economiesarenotselfcorrecting.Whetherornot theyarecouldnot,of course, be addressedwithinmodels that assumed that the economywas inequilibrium.It isnot just that theeconomydoesnot return to fullemploymentquickly after a strong, adverse shock.There are economic forces thatmay, ontheirown,exacerbate thedownturn.Unemployment leads to lower realwages;lowerrealwagesleadtoloweraggregatedemand;andloweraggregatedemandcan lead to still more unemployment. The implication of this is that a quickrecoverymayrequirestronggovernmentintervention.It'sclearthatwehaveyettofullytakeonboardthiscruciallesson:itisobviousthattheattemptsto"fix"theeconomiesoftheUnitedStatesandEuropeintheaftermathofthecrisishavefailedtorestoretheeconomytofullemployment.ThelossinGDPbetweenourpotentialandouractualoutputisinthetrillionsofdollars.Ofcourse,somewillsaythatitcouldhavebeendoneworse,andthat'strue,evenifit'scoldcomfort.

The reason for this failure-in spite of the unprecedented loose monetarypolicy-isthatfiscalpolicywastootepid;theKeynesianpoliciesweretoosmall

and not of long enough duration; and they were followed by contractionarypolicies,withfarmoreausterepoliciesinEuropethanintheUnitedStates,andfarworseeconomicperformanceinEuropethanintheUnitedStates.Ofcourse,thefiscalpoliciescouldhavebeenbetterdesigned.Butpoordesignwasnotthemain sourceof failure.Evenwith the imperfect design, the stimulative effects(themultipliers)werestrong.

More Than Deleveraging, More Than a Balance Sheet Crisis: The Need forStructuralTransformation

Wehaveroughly thesame levels todayofhumanresources,capital stock,andnatural resources as we did before the crisis. But many countries have notregained theirprecrisisGDPlevels, tosaynothingof returning to theprecrisisgrowth paths. It is clear that we are not using our resources well. In a veryfundamental sense, the crisis is still not fully resolved-and there's no goodeconomictheorythatexplainswhythatshouldbethecase.

Real business cycle theory and modern-day descendants of those modelssuggest therewas a negative productivity shock, a collective bout of amnesiathatresultedinareducedcapacitytoproduceoutputsfrominputs.Puttingasidethe absurdity of such a position, the irony is that in this downturn, individualfirms continued to increaseproductivity at a rapid rate.At themicroeconomiclevel,thereisnoevidenceofsuchamnesia.2

Somefocusonthehighlevelofdebt(especiallyatthehouseholdlevel)asanimpediment to recovery.But it isworthnoting that in thestandardmodels (onwhichpolicymakers relied soheavily in theyearsbefore the crisis, andwhichcontinue to be relied on in some circles) debt plays little role: it just changesclaimsonresources,transferringincomefromoneindividualtoanother.Andinthose models, such redistributions have no consequences. But even if suchtransfers lead to a lower level of aggregate demand (because creditors have alowermarginal propensity to consume than debtors), standard theory suggeststhat there is a change in prices that would restore the economy back to fullemployment.(Standardtheorydoesn'thavemuchtosayaboutthedynamicsofthese price adjustments. Indeed, one strand simply assumes that wages and

pricesarerigid,whenofcourseintheGreatDepressiontheywerechangingatavery rapid rate. The problem, as I hinted above, is that the adjustments maythemselveshavebeencounterproductive.)

Much of the popular discussion sees no prospects of wage and priceadjustmentsrestoringtheeconomyquicklytofullemployment.Ithaseffectivelyabandonedthe"standard"model.It isarguedthat ifwecouldonlydeleverage-get ridof the excessivedebt-wecould return to someversionofnormality. Inthisview,theprolongeddownturnisaresultoftheslowpaceofdeleveraging.

Butevenastheeconomydeleverages, thereiseveryreasontobelievethatitwill not return to full employment, even if it does lead to some increase inaggregatedemand.Wearenotlikelytoreturntotheprecrisishouseholdsavingsrate of zero-nor would it be a good thing if we did.' Moreover, even ifmanufacturinghasaslightrecovery,mostofthejobsthathavebeenlostinthatsectorwillnotbe regained.Norwill largenumbersofconstructionworkers intheUnitedStatesthatwereemployedatthepeakofthehousingbubblequicklyregaintheirjobs.

Some,lookingatpastdata,havesuggestedthatweshouldresignourselvestothisunfortunatestateofaffairs.Economiesthathavehadseverefinancialcrises4typically recover slowly.But the fact that thingshaveoftengonebadly in theaftermathofafinancialcrisisdoesn'tmeantheymustgobadly.Tomakemattersgowell,though,onehastounderstandwhyrecoveriesareoftensoslow.

Inearlierwork,GreenwaldandIexplainedwhyrecoveriesfrombalancesheetrecessions(wheretherewereadverseshockstofirms'equity)weresosluggish,andwhythiswasespeciallysowhenbanks'balancesheetswerebadlyhit.'

Butthis ismorethanjustabalancesheetcrisis.Thereisadeepercause: theUnitedStatesandEuropearegoingthroughastructuraltransformation.Thereisa structural transformation associatedwith themove frommanufacturing to aservice sector economy (just as earlier in the twentieth century there was astructural transformation from agriculture to manufacturing). Additionally,changingcomparativeadvantagesrequiresmassiveadjustments in thestructure

oftheNorthAtlanticcountries.Suchtransformationsoccurslowly.Thisispartlybecause the human and physical capital stock has to be restructured (workershavetoberetrained,andoftenrelocated).

Further, markets do not make such adjustments easily on their own, partlybecausethosewhohavetoshiftdonotthemselveshavetheresourcestofinancetherequisiteinvestments,havinglostmuchoftheirhumanandothercapitalasaresult of theunderlying forcesgiving rise to the structural transformation; andthere are natural imperfections in capitalmarkets arising out of imperfect andasymmetricinformation.

Keynesian policies to stimulate the economy are not only able to increaseGDP but can also facilitate restructuring. This is especially so if publicexpenditure is appropriately directed. Conversely, austerity measures, such asthosethatmanycountriesareundertakingtoday,impedetherestructuringthatisnow required.With austerity, someof the sectors thatwouldnaturally expand(as manufacturing contracts, along with its share of employment) are servicesectorsinwhichpublicfinancialsupporthastraditionallyplayedakeyrole,forunderstandablereasons.6

ReformsThatAre,atBest,HalfwayMeasures

As Ihaveobserved,marketsby themselvesdonot ingeneral lead toefficient,stable,andsociallyacceptableoutcomes.Thismeanswehavetothinkalittlebitmoredeeplyaboutwhatkindofeconomicarchitectureswillleadtogrowth,realstability,andagooddistributionofincome.

There is an ongoing debate about whether we simply need to tweak theexistingeconomicarchitectureorwhetherweneed tomakemore fundamentalchanges.Ihavetwoconcerns.OneIhintedatearlier:thereformsundertakensofarhaveonly tinkeredat theedges.Thesecond is that someof thechanges inoureconomicstructure (bothbeforeandafter thecrisis) thatweresupposed tomaketheeconomyperformbettermaynothavedoneso.

Therearesomereforms,for instance, thatmayenable theeconomytobetterwithstand small shocks but at the same time make it less able to absorb big

shocks. This is true of many financial sector developments, which may haveallowedtheeconomytoabsorbsomeofthesmallershocksbutclearlymadetheeconomy less resilient to fatter-tail shocks. Many of the "improvements" inmarketsbeforethecrisisactuallyincreasedcountries'exposuretorisk.Whateverthebenefitsthatmightbederivedfromcapitalandfinancialmarketliberalization(and theyarequestionable), therehavebeenseverecosts in termsof increasedexposuretorisk.Weought toberethinkingattitudes towardthesereforms-andthe IMF should be commended for its rethinking in recent years. One of theobjectivesofcapitalaccountmanagement, inallof its forms,canbe to reducedomesticvolatilityarisingfromacountry'sinternationalengagements.

More generally, the crisis has brought home the importance of financialregulationformacroeconomicstability.ButasIassesswhathashappenedsincethecrisis,Ifeeldisappointed.Withthemergersinthefinancialsectorthathaveoccurredintheaftermathofthecrisis, theproblemoftoo-big-to-failbankshasbecomeevenworse.Buttheproblemisnotjustwithtoo-big-to-failbanks.Therearebanksthataretoointertwinedtofailandbanksthataretoocorrelatedtofail.Wehavedonelittleaboutanyoftheseissues.7Therehas,ofcourse,beenahugeamountofdiscussionabouttoobigtofail.Butbeingtoocorrelatedisadistinctissue. There is a strong need for a more diversified ecology of financialinstitutionsthatwouldreduceincentivestobeexcessivelycorrelatedandleadtogreaterstability.

Also,wehaven'tdoneenoughtoincreasebankcapitalrequirements.Missinginmuchofthediscussionisanassessmentofthecostsversusbenefitsofhighercapitalrequirements.Weknowthebenefits:alowerriskofagovernmentbailoutandarecurrenceofthekindsofeventsthatmarked2007and2008.Butonthecost side, we've paid too little attention to the fundamental insights of theModigliani-Miller theorem, which explains the fallacies of arguments thatincreasingcapitalrequirementswillincreasethecostofcapital.'

DeficienciesinReformsandinModeling

Ifwehadbegunourreformeffortswithafocusonhowtomakeoureconomymore efficient and more stable, there are other questions we would have

naturally asked. Interestingly, there is some correspondence between thesedeficiencies inourreformeffortsandthedeficiencies in themodels thatweaseconomistsoftenuseinmacroeconomics.

TheImportanceofCredit

Wewould,forinstance,haveaskedwhatthefundamentalrolesofthefinancialsectorare,andhowwecanget it toperformthoserolesbetter.Clearly,oneofthekeyrolesistheallocationofcapitalandtheprovisionofcredit,especiallytosmallandmedium-sizedenterprises,afunctionitdidnotperformwellbeforethecrisisandarguablyisstillnotfulfillingwell.

Thismightseemobvious.Butafocusontheprovisionofcredithasnotbeenatthecenterofpolicydiscourseorofthestandardmacromodels.

I believe we have to shift our focus frommoney to credit. In looking at abank'sbalancesheet,thetwosidesareusuallygoingtobehighlycorrelated.Butthat is not always the case, particularly in the context of large economicperturbations. Especially in deep recessions, we ought to be focusing on theimpediments tocreditcreation. I findremarkable theextent towhich therehasbeenan inadequateexamination instandardmacromodelsof thenatureof thecreditmechanism.'

Butfailingtoanalyzecreditmarkets--andtomanagecreditcreation-isnottheonly lacuna in "monetary" theory and policy. There is also a lack ofunderstandingofdifferentkindsoffinance.Amajorareaintheanalysisofriskin financialmarkets is thedifferencebetweendebtandequity.But instandardmacroeconomics,thishasbarelybeengivenanyattention.

Stability

As I have already noted, in the conventionalmodels (and in the conventionalwisdom)market economieswere stable. So itwas perhaps not a surprise thatfundamentalquestionsabouthowtodesignmorestableeconomicsystemswereseldomasked.Wehavealreadytouchedonseveralaspectsofthis:howtodesigneconomicsystemsthatarelessexposedtoriskorthatgeneratelessvolatilityon

theirown.

Oneofthenecessaryreforms,thoughitisnotemphasizedenough,istheneedformoreautomaticstabilizersandfewerautomaticdestabilizers,notonlyinthefinancial sector but also throughout the economy. For instance, themovementfrom defined benefit to defined contribution systems may have led to a lessstableeconomy.

Elsewhere I have explained how risk-sharing arrangements (especially ifpoorly designed) can actually lead to more systemic risk: the precrisisconventional wisdom that diversification essentially eliminates risk is simplywrong.10

Distribution

Distributionmattersaswell-distributionamongindividuals,betweenhouseholdsandfirms,amonghouseholds,andamongfirms.Traditionally,macroeconomicsfocusedoncertainaggregates,suchastheaverageratioofleveragetoGDP.Butthatandotheraveragenumbersoftendon'tgiveapictureofthevulnerabilityoftheeconomy.Itwasthefactthatalargenumberofpeopleatthebottomwereatriskofbeingunabletomaketheirdebtpaymentsthatshouldhavetippedusoffthatsomethingwaswrong.

Across theboard,ourmodelsneed to incorporateagreaterunderstandingofheterogeneityanditsimplicationsforeconomicstability.

PolicyFrameworks

Flawed models lead not only to flawed policies but also to flawed policyframeworks.

ShouldMonetaryPolicyFocusjustonShort-TermInterestRates?

Inmonetarypolicy,thereisatendencytothinkthatthecentralbankshouldonlyintervene in setting the short-term interest rate.Adherents to this viewbelieve"oneintervention"isbetterthanmany.Sinceatleast80yearsagowiththework

ofRamsey,'wehaveknownthatfocusingonasingleinstrumentisnotgenerallythebestapproach.

The advocates of the "single intervention" approach argue that it is bestbecause it leastdistorts theeconomy.Ofcourse, the reasonwehavemonetarypolicy in the first place-the reason why government acts to intervene in theeconomy-is that we don't believe thatmarkets on their ownwill set the rightshort-terminterestrate.Ifwedid,wewouldjustletfreemarketsdeterminethatinterest rate.Theoddthing is thatwhile justabouteverycentralbankerwouldagreeweshouldinterveneinthedeterminationofthatprice,noteveryoneissoconvincedthatweshouldstrategicallyinterveneinothers,eventhoughweknowfromthegeneraltheoryoftaxationandthegeneraltheoryofmarketinterventionthatinterveninginjustonepriceisnotoptimal.

Onceweshiftthefocusofouranalysistocredit,andexplicitlyintroduceriskinto the analysis,we become aware thatwe need to usemultiple instruments.Indeed,ingeneral,wewanttousealltheinstrumentsatourdisposal.Monetaryeconomistsoftendrawadivisionbetweenmacroprudential,microprudential,andconventional monetary policy instruments. In our book, Toward a NewParadigm in Monetary Economics, Bruce Greenwald and I argue that thisdistinction is artificial. The government needs to draw on all of theseinstruments,inacoordinatedway.(I'llreturntothispointshortly.)

Of course, we cannot "correct" every market failure. The very large ones,however-themacroeconomicfailures-willalwaysrequireourintervention.BruceGreenwald and I have pointed out that markets are never Pareto efficient ifinformation is imperfect, if there are asymmetries of information, or if riskmarketsareimperfect.Andsincetheseconditionsarealwayssatisfied,marketsareneverParetoefficient.12Recentresearchhashighlightedtheimportanceofthese and other related constraints for macroeconomics-though again, theinsightsof this importantworkhaveyet tobeadequately integrated intoeithermainstream macroeconomic models or policy discussions. For instance,privately profitable contracts (e.g., credit default swaps) may, as we notedearlier, enhance systemic risk. The reason we have financial and bankingregulationispreciselybecauseprofitmaximizationonthepartofprivateagents

does not, in general, lead to socially optimal outcomes. There are largeexternalities, for instance, associated with the actions taken by certain agentswhich they naturally don't take into account: the bankers did not take intoaccount the costs that their excessive risk takingwould impose on the rest ofsociety.

PriceversusQuantitativeInterventions

Thesetheoreticalinsightsalsohelpustounderstandtheerroneousnessoftheoldpresumption among some economists that price interventions are preferable toquantity interventions. There are many circumstances in which quantityinterventionsleadtobettereconomicperformance.13

RethinkingTinbergen'sAnalysisofTargetsandInstruments

Apolicyframeworkthathasbecomepopularinsomecirclesarguesthatsolongasthereareasmanyinstrumentsasthereareobjectives,theeconomicsystemiscontrollable,andthebestwayofmanagingtheeconomyinsuchcircumstancesis tohavean institution responsible forone target andone instrument. (In thisview, central banks have one instrument, the interest rate, and one objective,inflation.14We have already explained why limiting monetary policy to oneinstrumentiswrong.)

Drawingsuchadivisionmayhaveadvantagesfromanagencyorbureaucraticperspective, but from the point of view of managing macroeconomic policy-focusing on employment, growth, stability and distribution, in a world ofuncertainty-it makes no sense. There has to be coordination among all theinstruments at our disposal, taking into account the impacts on all societalobjectives."sTheequilibriumthatariseswhendifferentpeoplecontroldifferentinstruments and focus on different objectives is, in general, not optimal inachieving overall societal objectives. In particular, there needs to be closecoordination between monetary and fiscal policy. Better coordination-and theuseofmoreinstruments-can,forinstance,enhanceeconomicstability.

TakeThisChancetoRevolutionizeFlawedModels

Itshouldbeclearthatwecouldhavedonemuchmoretopreventthecrisisthatbeganin2007andtomitigate itseffects. Itshouldbeclear toothatwecandomuchmoretopreventthenextone.Still,weareatleastbeginningtoidentifythereally big market failures, the big macroeconomic externalities, and the bestpolicy interventions for achieving high growth, greater stability, and a betterdistributionofincome.

To succeed,wemust constantly remindourselves thatmarketson theirownarenotgoingtosolvetheseproblems,andneitherwillasingleinterventionsuchaschanging short-term interest rates.That this is sohasbeenproved timeandagainoverthelastcenturyandahalf.Weshouldnotletourselvesbedeceivedagainbyoverlysimplisticmodelsthatsuggestotherwise.

Andasdauntingas theeconomicproblemswenowfaceare,acknowledgingthiswillallowustotakeadvantageoftheonebigopportunitythatthisperiodofeconomic traumahasafforded: thechance to revolutionizeour flawedmodels,andperhapsevenexitfromaninterminablecycleofcrises.

Notes

This chapter, an extendedversionof adiscussionpresentedApril 17,2013, atthe IMF Conference "Rethinking Macro Policy II: First Steps and EarlyLessons,"isbasedonjointworkwithanumberofmycolleagues,citedbelow.Iespecially want to thank my long term coauthor Bruce Greenwald. Researchassistance from Laurence Wilse-Samson and Eamon Kircher-Allen is alsogratefullyacknowledged. I alsowish toacknowledge financial assistance fromthe Institute for New Economic Thinking, and to acknowledge its agenda toreexaminethefoundationsofmacroeconomicsinthelightofthecrisis.

1. Ihaveelaboratedatgreater lengthonsomeof the lessons tobedrawnfromthe crisis for macroeconomics in "The Financial Crisis of 2007-2008 and ItsMacroeconomicConsequences," inTimeforaVisibleHand:Lessonsfromthe2008WorldFinancialCrisis,ed.S.Griffith-Jones,J.A.Ocampo,andJ.E.Stiglitz,Initiative for PolicyDialogue Series (Oxford:OxfordUniversity Press, 2010),19-49; "Rethinking Macroeconomics: What Failed and How to Repair It,"Journal of the European Economic Association 9, no. 4 (2011): 591-645;"RethinkingMacroeconomics:WhatWentWrongandHowtoFixIt,"Journalof

Global Policy 2, no. 2 (2011): 165-175; "Stable Growth in an Era of Crises:LearningfromEconomicTheoryandHistory,"Economi-tek2,no.1(2013):1-39;and"Macroeconomics,MonetaryPolicy,andtheCrisis,"inIntheWakeofthe Crisis, ed. O.Blanchard, D. Romer, M.Spence, and J.Stiglitz (Cambridge,MA:MITPress,2012).

2. Economic downturns can lead to the destruction of organization andinformational capital, as Bruce Greenwald and I have emphasized. See ourTowards a New Paradigm in Monetary Economics (Cambridge: CambridgeUniversityPress,2003).

3.Itisimportanttodistinguishbetweenargumentsconcerningtherestorationofgrowthandthosefocusingontherestorationoftheeconomytofullemployment.It isconceivable that theeconomycouldreturn tonormalgrowth-creatingnewjobs in tandem with new entrants to the labor force-but that the level ofunemploymentremainselevated.Hence,itispossiblethatoncetheeconomyhasdeleveraged, growth might be restored. Here we are asking, would aggregatedemandbesufficienttoreturntheeconomytofullemployment?

4. We note too that the fact that downturns that are associated with deepfinancialcrisesarelonglastingtellsusverylittle:iftherearedeeperfundamentalcauses to crises (as suggested in the next paragraph), then these deep andlonglasting crises will result in financial crises; the financial crises are theconsequence,not the (underlying) cause. If that is thecase, then the statementthat"deepfinancialcrisesarelonglasting"saysnothingmorethan"deepcrisesarelonglasting,"astatementthat,whiletrue,isnotveryinformative.

5.B.GreenwaldandJ.E.Stiglitz,"FinancialMarket ImperfectionsandBusinessCycles," Quarterly Journal of Economics 108, no. 1 (1993): 77-114; andGreenwaldandStiglitz,TowardsaNewParadigminMonetaryEconomics.

6. The ideas in this paragraph are elaborated in D.Delli Gatti, M.Gallegati,B.Greenwald, A.Russo, and J.E.Stiglitz, "Mobility Constraints, ProductivityTrends, and ExtendedCrises," Journal of EconomicBehavior&Organization83, no. 3 (2012): 375-393; and idem, "Sectoral Imbalances and Long RunCrises,"inTheGlobalMacroEconomyandFinance,ed.EAllen,M.Aoki,J.-P.Fitoussi,N.Kiyotaki,R.Gordon,andJ.E.Stiglitz,IEAConferenceVolume150-III(Houndmills,UK,andNewYork:Palgrave,2012),61-97.

7.SeeJosephE.Stiglitz,"WitnessTestimonyofJosephE.Stiglitz,CongressionalOversightPanel,"HearingonImpactoftheTARPonFinancialStability,March4, 2011, http://cybercemetery.unt. edu/archive/cop/20110401230935/http://cop.senate .gov/documents/ testimony-030411-stiglitz.pditranscript(accessed September 30, 2013) and "Too Big to Fail or Too Big to Save?ExaminingtheSystemicThreatsofLargeFinancialInstitutions,"testimonyatahearing of theUnited States Congress's Joint EconomicCommittee,April 21,2009, http://www.jec.senate .gov/public/?a=Files.Serve& File_id= 6b50b609-89fa-4ddf-a799-2963b31d6f86(accessedSeptember30,2013).

8.SeeJ.E.Stiglitz,"OntheNeedforIncreasedCapitalRequirementsforBanksandFurtherActionstoImprovetheSafetyandSoundnessofAmerica'sBankingSystem," testimony before the Senate Banking Committee, August 3, 2011,http://www.banking.senate. gov/public/index.cfm? FuseAction=Files.View&FileStore_id= 97cec3el-2d1d-44fa-acd9-a0albc640bc4 (accessedSeptember 30,2013); A.Admati, P.M.De Marzo, M. E Hellwig, and P.Pfleiderer, "DebtOverhang and Capital Regulation," Stanford Working Paper, 2012,http://www.gsb .stanford.edu/news/packages/ PDF/AdmatiDebt032612.pdf(accessed July16, 2013); andAnatAdmati andMartinHellwig,TheBankers'NewClothes:What'sWrongwithBankingandWhattoDoaboutIt(Princeton,NJ:PrincetonUniversityPress,2013),andthereferencescitedthere.Therearenumerous other regulatory issues (e.g., dealing with the transparency ofderivativesandhavingthemtradeoverwell-capitalizedexchanges).Evenabriefexplorationofthesewouldtakemewellbeyondtheconfinesofthisshortnote.Isimplyobservethatinvirtuallyeachofthesearenas,theregulatoryreformshavefallenshortofwhatisdesiredorneeded.

9.Doing sowas the objective ofmy bookwithBruceGreenwald, Towards aNew Paradigm in Monetary Economics. There is, of course, a largemicroeconomic literature on banking and credit, but for the most part, theinsights of this literature have not been taken on board in standard macromodels.

10. See J.E.Stiglitz, "Contagion, Liberalization, and the Optimal Structure ofGlobalization,"JournalofGlobalizationandDevelopment1,no.2(2010),art.2;idem,"RiskandGlobalEconomicArchitecture:WhyFullFinancialIntegrationMaybeUndesirable,"AmericanEconomicReview100,no.2(2010):388-392;and S.Battiston, D.Delli Gatti, M.Gallegati, B.Greenwald, and J.E.Stiglitz,

"Liaisons Dangereuses: Increasing Connectivity, Risk Sharing, and SystemicRisk,"JournalofEconomicDynamicsandControl36(2012):1121-1141.

The intuition behind these results is simple: when we identify a group ofindividualswitha contagiousdisease,wedon't "diversify"-sending them toallquarters of the earth. We quarantine them. Economists have intuitivelyrecognized that economic crises can spread like a contagious disease, but thathappens because of one form of interdependence or another. High levels ofinterdependenceenableadisturbanceinonepartofthesystemtobetransmittedelsewhere. Well-designed "architectures" balance the advantages ofinterdependencewiththedisadvantages,andattempttointroducedesignfeatures("circuitbreakers,"capitalcontrols)thatmitigatetherisksofadversecontagion.Inthestandardmodels,therisksofcontagionwereignored(atleastbeforecrisesoccurred,whichtheydidwithincreasingfrequency),andthusnoattentionwaspaid topolicies thatmight reduce these risks-and indeed,capitalcontrolswereadamantlyopposed.

11.EP.Ramsey,"AContributiontotheTheoryofTaxation,"Economicjournal,1927,47-61.

12. See J.E.Stiglitz and B.Greenwald, "Externalities in Economies withImperfect Information and Incomplete Markets," Quarterly Journal ofEconomics101,no.2(1986):229-264.

13.Theclassicreference isWeitzman(M.L.Weitzman,"Pricesvs.Quantities,"ReviewofEconomicStudies41,no.4 [1974]:477-491).Since then, therehasbeen a wealth of literature in different contexts showing that quantityinterventionsmaybepreferable;forexample,quotasmaybepreferabletotariffs(P.Dasgupta and J.E.Stiglitz, "Tariffs Versus Quotas As Revenue RaisingDevicesUnderUncertainty,"AmericanEconomicReview67,no.5[1977]:975-981),orquantityinterventionsincapitalaccountmanagementmaybepreferabletopriceinterventions(J.E.Stiglitz,JoseAntonioOcampo,ShariSpiegel,RicardoFrenchDavis, and Deepak Nayyar, Stability with Growth: Macroeconomics,Liberalization, and Development, The Initiative for Policy Dialogue Series[Oxford:OxfordUniversityPress,2006]).

14.Afortiori,simplerules,likeinflationtargeting,thatcallforanincreaseintheinterest rate when the inflation rate exceeds the targeted level are even more

misguided.Suchrulesdonottakeintoaccountthesourceofthedisturbancetotheeconomyor themostefficientwayofrestoringtheeconomyto thedesired"equilibrium" after a perturbation. The obvious illustration was provided byinflationindevelopingcountriesarisingoutoftheglobalfoodandoilpricerisesof 2007. The increase in the interest rate in a small country obviouslywouldhave a negligible effect on these global prices; if the average rate of inflationwere to be brought down to the desired level, it would require such largeconstrictions in the nontraded sectors that the cure would be worse than thedisease. Fortunately, most governments have recognized this, and even thosethat have retained an inflation-targeting framework have adopted "flexible"frameworks.

15.JanTinbergen'sanalysiswas,ofcourse,basedonaverysimplemodel,withvery stringent assumptions. Evidently, those relying on his insights did notappreciatejusthowcriticaltheseassumptionswere,andthattheresultswerenotevenapproximatelycorrectinmoregeneralcontexts.

GeorgeA.Akerlof,a2001NobelLaureate,isGuestScholarattheInternationalMonetaryFundandDanielE.KoshlandSr.DistinguishedProfessorEmeritusatBerkeley.

Sheila Bair served as Chairman of the FederalDeposit InsuranceCorporationfromJune2006throughJune2011.

LorenzoBiniSmaghi iscurrentlyaVisitingScholaratHarvard'sWeatherheadCenter for InternationalAffairs.Hewas formerly aMember of the ExecutiveBoardoftheEuropeanCentralBank.

Olivier Blanchard is the Economic Counselor and Director of the ResearchDepartment at the IMF.He is also theRobertM.SolowProfessorEmeritus atMIT.

AndersBorgisaSwedisheconomistwhoservesasMinisterforFinanceintheSwedishgovernment.

ClaudioBoric,isHeadoftheMonetaryandEconomicDepartmentattheBankforInternationalSettlements(BIS).

AgustinCarstensisGovernoroftheBancodeMexico.HeisalsoamemberoftheSteeringCommitteeoftheG-20FinancialStabilityBoard,andtheChairmanoftheBISEconomicConsultativeCouncilandtheGlobalEconomyMeeting.

JoseDeGregorioisProfessorattheUniversityofChile.HewastheGovernoroftheCentralBankofChilefrom2007to2011.

GiovanniDell'Ariccia isAssistantDirector in theResearchDepartment of theIMF.

JaniceEberly is the JamesR. andHelenD.RussellDistinguishedProfessor ofFinanceattheKelloggSchoolofManagementatNorthwesternUniversity.Shewas Assistant Secretary for Economic Policy and Chief Economist at the USTreasuryfrom2011to2013.

Stanley Fischer is currently at the Council on Foreign Relations. He wasGovernorof theBankofIsraelfrom2005to2013andFirstDeputyManagingDirectoroftheIMFfrom1994to2001.

Andrew Haldane is Executive Director for Financial Stability at the Bank ofEngland.

MarcioHolland is Secretary ofEconomicPolicy for theBrazilianMinistry ofFinance since 2011 and also Professor at the Sao Paulo School ofEconomics/FGV.

ChoongsooKimistheGovernoroftheBankofKoreaandtheChairmanofitsMonetaryPolicyCommittee.

MervynA.KingwasGovernoroftheBankofEnglandfrom2003to2013.

PaoloMauroisanAssistantDirectorintheAfricanDepartmentoftheIMF.

RobertoPerottiisProfessorofEconomicsatUniversityBocconiinMilan.

HeleneReyisProfessorofEconomicsatLondonBusinessSchool.

David Romer is the Herman Royer Professor of Political Economy at theUniversityofCalifornia,Berkeley.

Nouriel Roubini is Professor of Economics at New York University's SternSchoolofBusinessandChairmanandCofounder,RoubiniGlobalEconomics.

Jay C.Shambaugh is Professor of Economics and International Affairs at theGeorgeWashingtonUniversity.

Jeremy C.Stein is a member of the Board of Governors of the US FederalReserve System and has served as Professor of Economics at HarvardUniversity.

Joseph E.Stiglitz is University Professor at Columbia University. He wasawardedtheNobelPrizeinEconomicSciencesin2001.

Duvvuri Subbarao was Governor of the Reserve Bank of India from 2008 to

2013.

Jean Tirole is Chairman of the Foundation J.J.Laffont-Toulouse School ofEconomics (TSE), Chairman of the Executive Committee of the Institute forAdvancedStudyinToulouse(IAST),andScientificDirectoroftheInstituteforIndustrialEconomics(IDEI).

AdairTurnerisSeniorFellowattheInstituteforNewEconomicThinking.From2008 to 2013 he was the Chairman of the UK Financial Services Authority(FSA).

JohnVickersisWardenofAllSoulsCollege,Oxford,andwasChairmanoftheUKIndependentCommissiononBanking2010-2011.

Martin Wolf is Associate Editor and Chief Economics Commentator at theFinancialTimes,London.

MichaelWoodford is the JohnBates Clark Professor of Political Economy atColumbiaUniversity.

Janet L.Yellen is the Vice Chair of the Board of Governors of the FederalReserveSystem.

Gang Yi is Deputy Governor of the People's Bank of China and theadministratoroftheStateAdministrationofForeignExchange(SAFE).

Date post:	11-Sep-2021
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times