Rethinking Central Banking Web

8/4/2019 Rethinking Central Banking Web

1/42

Think Tank 20:Macroeconomic Policy Interdependence

and the G-20

Rethinking CentralBankingCommittee on

International EconomicPolicy and Reform

Barry Eichengreen

Mohamed El-Erian

Arminio Fraga

Takatoshi Ito

Jean Pisani-Ferry

Eswar Prasad

Raghuram Rajan

Maria Ramos

Carmen Reinhart

Hlne Rey

Dani Rodrik

Kenneth Rogoff

Hyun Song Shin

Andrs Velasco

Beatrice Weder di Mauro

Yongding Yu

SEPTEMBER 2011


2/42

Rethinking CentralBankingCommittee on

International EconomicPolicy and Reform

Barry Eichengreen

Mohamed El-Erian

Arminio Fraga

Takatoshi Ito

Jean Pisani-Ferry

Eswar Prasad

Raghuram Rajan

Maria Ramos

Carmen Reinhart

Hlne Rey

Dani Rodrik

Kenneth Rogoff

Hyun Song Shin

Andrs Velasco

Beatrice Weder di Mauro

Yongding Yu

SEPTEMBER 2011


3/42

Rethinking Central Banking

ii

Preace

The Committee on International Economic and

Policy Reorm is a non-partisan, independent

group o experts, comprised o academics and

ormer government and central bank ocials. Its

objective is to analyze global monetary and nan-

cial problems, oer systematic analysis, and ad-

vance reorm ideas. Te Committee attempts to

identiy areas in which the global economic archi-

tecture should be strengthened and recommend

solutions intended to reconcile national interests

with broader global interests. Trough its reports,

it seeks to oster public understanding o key issues

in global economic management and economic

governance. Each Committee report will ocus on

a specic topic which will emphasize longer-term

rather than conjunctural policy issues.

Te Committee is grateul to the Alred P. Sloan

Foundation* or providing nancial support and

to the Brookings Institution or hosting the com-

mittee and acilitating its work. Quynh onnu pro-

vided excellent administrative and logistical sup-

port to the Committee.

Committee Members

Barry Eichengreen, University o Caliornia, Berkeley

Mohamed El-Erian, PIMCO

Arminio Fraga, Gavea Investimentos

akatoshi Ito, University o okyo

Jean Pisani-Ferry, Bruegel

Eswar Prasad, Cornell University and Brookings Institution

Raghuram Rajan, University o Chicago

Maria Ramos,Absa Group Ltd.

Carmen Reinhart, Peterson Institute or International Economics

Hlne Rey, London Business School

Dani Rodrik, Harvard University

Kenneth Rogo, Harvard University

Hyun Song Shin, Princeton University

Andrs Velasco,Columbia UniversityBeatrice Weder di Mauro, University o Mainz

Yongding Yu, Chinese Academy o Social Sciences

*Brookings recognizes that the value it provides to any donor is in its absolute commitment to quality, independence and impact. Activitiessponsored by its donors reect this commitment and neither the research agenda, content, nor outcomes are inuenced by any donation.


4/42


iii

Executive Summary

This report lays out a ramework or rethink-

ing central banking in light o lessons learned

in the lead-up to and afermath o the global

nancial crisis.

By the early 2000s, a growing number o central

banks, in advanced countries and emerging mar-

kets alike, had converged on a policy ramework,

exible ination targeting, which seemed capable

o achieving price stability and delivering mac-

roeconomic stability at the national and interna-

tional levels. Tis ramework had many practical

achievements, including bringing price stability

to many emerging markets. Now, however, there

is growing recognition that the conventional ap-

proach to central banking needs to be rethought.

Te relationship between price stability and the

broader goals o macroeconomic and nancial sta-

bility clearly needs to be redened. Moreover, the

evolution o monetary and exchange rate regimes

has resulted in incompatibilities among the poli-

cies o some key countries. Central banks are also

being pulled into new roles by the post-crisis envi-

ronment, which eatures high levels o public and

private debt in advanced economies and concerns

about capital inows and currency appreciation

in emerging markets. While some aspects o these

roles are not new, they are risky, as central bank

actions can inict collateral damage on domestic

nancial systems and have the potential o raising

new domestic and international tensions.

Te report analyzes these issues rom academ-

ic and practical policy-oriented perspectives.

Drawing on this analysis, it recommends changes

to the dominant ramework guiding central bank-

ing practice.

Te rst recommendation is that central banks

should go beyond their traditional emphasis on

low ination to adopt an explicit goal o nancial

stability. Macroprudential tools should be used

alongside monetary policy in pursuit o that ob-

jective. Mechanisms should also be developed to

encourage large-country central banks to inter-

nalize the spillover eects o their policies. Spe-

cically, we call or the creation o an International

Monetary Policy Committee composed o repre-

sentatives o major central banks that will report

regularly to world leaders on the aggregate conse-

quences o individual central bank policies.

Tere is substantial pressure on central banks to

acknowledge the importance o still other issues,

such as the high costs o public debt management

and the level o the exchange rate. Central banks

are more likely to saeguard their independence

and credibility by acknowledging and explicitly

addressing the tensions between ination target-

ing and competing objectives than by denying

such linkages and proceeding with business as

usual. Central banks should make clear that mon-

etary policy is only one part o the policy response

and cannot be eective unless other policiess-

cal and structural policies, nancial sector regula-

tionwork in tandem.


5/42


1

The Golden Age o Infation Targeting

High ination in the advanced economies in the

1970s and in emerging economies in the 1980s and

1990s was instrumental in shaping modern think-

ing about the practice o central banking. Te tenets

o the resulting ramework are amiliar and, to a

large extent, uncontroversial. First, there is no per-

manent tradeo between ination and unemploy-

menta sustained higher level o ination does not

lead to higher growth and a sustained lower level o

unemployment. Second, high and volatile ination

depresses growth and distorts the allocation o re-

sources. Tird, ination disproportionately harms

the poorest segments o society, which lack instru-

ments or protecting themselves rom its disruptive

eects. For all o these reasons, price stability is the

cornerstone o monetary policy.

Te actions needed to achieve price stability, such

as the maintenance o high interest rates, can be

politically unpopular, among other reasons be-

cause they slow growth. It ollows that the pur-

suit o price stability can be made more credible

and thus more eective by granting independence

or at least operational autonomy to the central

bank. Otherwise, central banks may be subject to

political pressure to attach greater weight to other

objectives, making it harder or them to containinationary expectations and deliver desirable

outcomes.

By the early 2000s, a growing number o central

banks, in advanced countries and emerging mar-

kets alike, had converged on a policy ramework,

exible ination targeting, that seemed capable o

achieving these desiderata and delivering macro-

economic stability at the national and internation-

al levels. In the conventional view, there are our

explanations or this happy outcome:

Flexible ination targeting, under which

the central bank aims to stabilize ination

around its target but also minimize the

output gap, delivers low ination at the

national level, thereby avoiding the need

or large nominal exchange-rate adjust-

ments and the kind o overshooting that

characterized the 1980s.1

Flexible ination targeting, by allowing

or exchange rate variability, acilitates in-

ternational adjustment. Countries experi-

encing demand shocks can cushion them

1 Although neither the Fed nor the ECB had ormally endorsed ination targeting (I), both were aiming at price stability, which made theirpolicies similar to those o the central banks on a strict I regime.

CHAPTER 1

Introduction


6/42


2

through interest-rate changes and associ-

ated movements in exchange rates.

Flexible ination targeting makes re-

serve accumulation unnecessary, since

exchange-rate intervention is rare andlimited to short-term responses to market

disruptions and to a signaling role in cases

o serious misalignments.

Te combined policy stance o the coun-

tries ollowing this strategy is supposed to

ensure an appropriate level o aggregate

demand at the global level.

Te generalization o ination targeting cum oat-

ing exchange rates could thus be regarded as thetriumph o the own house in order doctrine in

the international monetary eld. National macro-

economic stability was seen as sucient or inter-

national macroeconomic stability. Te domestic

and international aspects were essentially regarded

as two sides o the same coin.2

An added benet o exible ination targeting,

according to the emerging orthodoxy, was that it

allowed the objectives o price stability and nan-

cial stability to be pursued through separate toolsmonetary policy or the ormer and micro-pru-

dential regulatory and supervisory measures or

the latter. inbergens separation principle, i.e. the

idea that each goal should be pursued with a sepa-

rate and dedicated instrument, was widely invoked

in this context.

In this orthodox view, monetary policy ocuses

on controlling ination and works by managing

expectations o uture policy rates, which by the

expectations theory o the yield curve determinethe long-term interest rates that inuence ag-

gregate demand. Financial stability is attained by

microprudential regulation o bank capital that

counteracts the moral hazard generated by deposit

insurance, together with periodic supervisory as-

sessments and the necessary strictures meant to

prevent excessive risk taking and maleasance. Re-

gardless o whether the microprudential regulator

is situated in the central bank or a separate special-

ized regulatory agency, nancial regulation is seen

as a separate activity.

Central bankers nowadays ofen observe that ex-

ible ination targeting was never as straightor-

ward as this ramework suggests and that issues

o nancial stability and spillovers were always on

their minds. Still, it remains accurate to say that

the basic theoretical ramework sketched above

did much to shape their thinking. Its clarity and

simplicity enabled it to gain adherents in academia

and nancial markets as well as in central banks.

Rethinking the Framework

Some o the practical achievements o the exible

ination targeting ramework are indisputable.

Te adoption o price stability objectives by coun-

tries at dierent levels o economic development

was a major step orward afer decades o domes-

tically-generated instability. Tis ramework can

be credited, at least in part, or the drop in globalination and the abatement o exchange-rate con-

troversies among the advanced economies.3

Now, however, there is growing recognition that

the conventional approach to central banking

needs to be rethought. Critics reach this conclu-

sion or several related reasons:

Te conventional approach ails to ac-

count adequately or nancial-sector risk

and is thereore too narrowly ocused.

Te conventional ramework assumes lim-

ited or nonexistent cross-border spillovers

2 Looking ahead, some even regarded this regime as the solution to perennial international monetary controversies (Rose, 2007).3 o what extent I can be credited or the disination o the 1990s and the early 2000s is a matter or discussion. Another important actor was

the disinationary pressure coming rom the emerging countries exports. We return to the issue below.


7/42


3

o monetary policies, while in act spill-

overs are requently o rst-order impor-

tance. Tey can complicate monetary pol-

icy management, accentuate the volatility

o real activity and increase nancial-sec-

tor risk.

Te incompatibility o national monetary

policies in the ace o spillovers is height-

ened when countries ollow dierent de

acto monetary policy regimes (e.g., ina-

tion targeting and exchange rate target-

ing).4

Spillovers may be urther accentuated

when central banks pursue unconven-

tional monetary interventions (e.g., wheninterest rates are at their oor and con-

strained by the zero bound). Because o

weak domestic demand, as well as dis-

tressed banks that are unwilling to lend,

the portolio adjustments prompted by

unconventional policies may largely serve

to increase capital ows to countries with

stronger growth prospects rather than

boosting domestic credit as intended.

High levels o government debt in ad-vanced countries and the slowing growth

o traditional export markets or develop-

ing countries create new sources o po-

litical pressure that central banks will nd

dicult to ignore.

In this report, we start by considering the valid-

ity o these criticisms. We then go on to ask how

central banking theory and practice need to be up-

dated in light o this shif in thinking. Te report

consists o our chapters (afer this one) ollowedby our recommendations.

In Chapter 2, we describe how the global nancial

crisis has recast the debate over central banking.

We ocus on the relationship between the tradi-

tional emphasis on price stability and the broader

goals o macroeconomic and nancial stability. We

discuss why the traditional separation, in which

monetary policy targets price stability and regula-

tory policies target nancial stability, and the two

sets o policies operate largely independently o

each other, is no longer tenable.

I central banks do in act embrace the goal o

nancial stability in addition to price stability,

monetary policy-making and policy communica-

tion will become more challenging. We thereore

consider the practical issues that arise when the

central bank is orced to juggle multiple mandates.

We then turn in Chapter 3 to a criticism o theconventional policy ramework: it assumes not

just that central banks practice exible ination

targeting but also that they allow the exchange rate

to oat reely. Under these assumptions, each cen-

tral bank has the independence necessary to target

price stability and ull employment.

Te problem is that policy independence in theory

may exceed policy independence in practice. In

other words, the conventional ramework ails to

take into account that national policies can havepowerul cross-border repercussions that the a-

ected partner may not be able to adequately o-

set with exchange rate movements. In part this

is because the existing system is not, in act, one

o ully exible exchange rates. In practice, some

countries eectively target exchange rates (Chinas

tight management o its currencys value relative

to the US dollar being a prominent case in point).

In part it is because international transmission oc-

curs even under exible exchange rates, through

both trade channels and capital ows. Te conse-quences include the prospective re-emergence o

global imbalances as well as the prolieration o

trade and capital controls when countries seek ur-

ther insulation rom cross-border spillovers.

4 Tough the choice o regime itsel may partly be a reaction to spillovers.


8/42


4

o analyze these issues, in this chapter we provide

a global perspective on the evolution o monetary

policy and exchange rate regimes. We examine the

problems that arise out o the incompatibility o

national regimes with similar domestic objectives.

We then discuss the challenges that arise in rec-

onciling domestic monetary policies with global

macroeconomic stability.

In Chapter 4, we describe how central banks are

being pulled into new roles by the post-crisis en-

vironment and by the unavailability o alternative,

potentially more suitable instruments.While some

aspects o these roles are not new, they nonetheless

move central banks into risky territory insoar as

central bank actions can inict collateral damage

on domestic nancial systems and have the po-tential o raising new domestic and international

tensions. We highlight two sets o issues: (a) the

consequences o high levels o public and private

debt in the advanced economies and the attendant

pressures towards nancial repression; and (b)

the perceived dangers o currency misalignments

and overvaluation, and the attendant pressures to-

wards currency intervention and capital controls.

In Chapter 5, we draw on the analysis in previous

chapters to recommend changes in the dominantramework guiding central banking practice. In

the ramework we propose, central banks should

go beyond their traditional emphasis on low ina-

tion to adopt an explicit goal o nancial stabil-

ity. Macroprudential tools should be used along-

side monetary policy in pursuit o that objective.

Mechanisms should also be developed to encour-

age large-country central banks to internalize the

spillover eects o their policies. Specically, we

call or the creation o an International Monetary

Policy Committee composed o representatives o

major central banks that will report regularly to

world leaders on the aggregate consequences o

individual central bank policies.

While this report suggests more responsibilities or

central banks, we also recognize the environment is

one where there is substantial pressure on centralbanks to acknowledge the importance o still other

issues, such as the high costs o public debt manage-

ment and the level o the exchange rate. While these

pressures, i internalized, can make central bank ob-

jectives hopelessly diuse, they are not reasons to

postpone rethinking the overall policy ramework.

o the contrary, a ramework that is seen as de-

cient will become an easier political target.

For all these reasons, we believe it is time to re-

think the existing paradigm. Te rest o the reportlays out what this rethinking should entail.
http://instruments.while/http://instruments.while/


9/42


5

CHAPTER 2

The Scope o Monetary Policy

I

n this chapter we describe how the global nan-

cial crisis has recast the debate over the scope ocentral banking unctions. We ocus on the re-

lationship between the traditional narrow goal o

monetary policyprice stabilityand the broader

goals o macroeconomic and nancial stability. We

explain why the traditional separation, in which

monetary policy targets price stability and regula-

tory policies target nancial stability and the two

sets o policies operate independently o each oth-

er, is no longer tenable. We then review some prac-

tical issues that arise in connection with attempts

to coordinate the two sets o policies.

Central Banks and Financial Stability

Te global nancial crisis shook condence in mi-

croprudential tools o regulation as the primary in-

strument or ensuring nancial stability. Yet many

central bankers still subscribe to the traditional di-

chotomy between monetary policy and nancial

stability, except that microprudentialtools have given

way to an embrace omacroprudential tools o -

nancial regulation (countercyclical capital adequacyrequirements, or example). Tese tools or policies,

which mitigate risks to the nancial system as a

whole rather than solely at the level o the individual

institution, are to be developed and implemented by

specialists in nancial stability, not by central bank-

ers responsible or the conduct o monetary policy.

Te case or this separation rests on the belie that

interest rates are too blunt an instrument or the

eective pursuit o nancial stability. Te question

is commonly ramed as whether the central bankshould raise interest rates in response to asset bub-

bles. In the 1990s and early 2000s, central bankers

discussed at length whether and how to respond

to asset market developments.5 Te conclusion o

that debate was that central banks had a mandate

to react to bursting bubbles but not to target asset

prices. Not everyone, however, shared this conclu-

sion. Te lean vs. clean debate remained active in

the run-up to the crisis.6

5 Te early debate was ramed by the stock market boom o the late 1990s. Arguments in avor o leaning against the wind when it comesto nancial developments have been given by Blanchard (2000), Bordo and Jeanne (2002), Borio and Lowe (2002), Borio and White (2003),Cecchetti, Genberg, Lipsky and Wadhwani (2000), Crockett (2003), Dudley (2006) and Goodhart (2000) among others. Te argument against isgiven in Bean (2003), Bernanke and Gertler (1999, 2001), Bernanke (2002), Greenspan (2002), Kohn (2005), Mishkin (2008) and Stark (2008).

6 A policy school, primarily associated with economists rom the Bank or International Settlements and the Bank o Japan, was critical o narrowination targeting and maintained that central banks could not orgo their responsibility or nancial stability. Bank o Japan economistsregretted having allowed the bubble to become too large in the second hal o the 1980s. Te European Central Bank never ully endorsed thestandard ormulation o ination targeting and argued that the growth o monetary aggregates and credit developments were also importantindicators o potential risks to price stability over a longer-term horizon.


10/42


6

Te case against attempting to prick bubbles rests

on the ollowing arguments.

Identiying bubbles is hard.

Even i there is a bubble, monetary policy

is not the best tool with which to address

it. An asset price bubble will not respond

to small changes in interest rates; only a

sharp increase will suce to prick a bub-

ble. However, a drastic increase in interest

rates can cause more harm than good by

depressing output growth and increasing

output volatility.

Te claim that an asset price bubble will not re-

spond to a small change in interest rates has beenmade in the context o stock market bubbles,

where the proposition is most plausible. When the

stock market is rising by 20 percent a year, a small

increase in interest rates will not outweigh the e-

ects o rapid asset price increases.

However, the stock market may not be the best

context in which to discuss the nancial stability

role o monetary policy. Te housing market, with

its more prominent role or leverage and credit,

and markets in the derivative securities associatedwith housing investment may be more pertinent.

Monetary policy stands at the heart o the lever-

age decisions o banks and other nancial inter-

mediaries involved in lending or housing-related

investments. In this setting, even small changes in

unding costs may have an impact on risk-taking

and unding conditions. Financial intermediaries,

afer all, borrow in order to lend. Te spread be-

tween borrowing and lending rates is thereore a

key determinant o the use o leverage and has im-

portant implications or the interaction betweenbanking sector loan growth, risk premia, and any

ongoing housing boom.7

Focusing on risk taking by banks and other nan-

cial intermediaries will lead the policy maker to ask

additional questions about risks to the stability o

economic activity. Rather than waiting or incontro-

vertible proo o a bubble in housing markets, or

example, a policy maker could instead ask whether

benign unding conditions could reverse abruptly

with adverse consequences or the economy. Even

i policy makers are convinced that higher housing

prices are broadly justied by secular trends in pop-

ulation, household size, and living standards, policy

intervention would still be justied i the policy

maker also believed that, i lef unchecked, current

loose monetary conditions signicantly raise the

risk o an abrupt reversal in housing prices and o

nancing conditions, with adverse consequences

or the nancial system and the economy.

Not responding in this way has led to a dangerous-ly asymmetric response to credit market develop-

ments. Central banks have allowed credit growth

to run ree, ueling booms, and then ooded mar-

kets with liquidity afer the crash, bailing out -

nancial institutions and bondholders. Tis asym-

metry has contributed to stretched balance sheets,

with aster lending growth and leverage in times o

low risk premia, more violent deleveraging when

risk premia rise, and requent booms and busts.

For all these reasons, there is a case or centralbanks to guard against credit market excesses. An

ination-targeting central bank may argue that it

does so automatically insoar as higher asset prices

boost aggregate demand through wealth eects

and create inationary pressures. However, some

additional leaning against credit market develop-

ments would be advisable even in the absence o

aggregate demand eects once it is determined

that unding conditions and reduced risk premia

indicate a nascent credit boom. Put dierently,

ination-targeting central banks may want to straybelow target when conditions are boom-like

when rapid asset price growth is accompanied by

substantial credit expansionsince policy would

otherwise become asymmetric and execerbate

macroeconomic volatility.

7 See Adrian and Shin (2011) or a discussion o these linkages.


11/42


7

Retiring the Separation Principle

A consequence o this doctrine o leaning against

the wind is that the neat inbergen assignment o

dierent tools to dierent objectives becomes more

dicult to implement in practice. Interest rates aectnancial stability and, hence, real activity. Equally,

macroprudential tools impact credit growth and ex-

ternal imbalances with consequences or macroeco-

nomic and price stability. When consumer credit is

growing rapidly and the household debt ratio is high,

or example, restraining credit growth by changing

guidance on loan-to-value (LV) or debt service-to-

income (DI) ratios over the business cycle will have

important macro-stabilization eects.

Rather than viewing the allocation problem as hav-ing a corner solution where one instrument is de-

voted entirely to one objective, the macro-stabiliza-

tion exercise must be viewed as a joint optimization

problem where monetary and regulatory policies

are used in concert in pursuit o both objectives.

Believers in a strict interpretation o inbergen

separation will ret that blurring the assignment o

instruments to targets will jeopardize the central

banks operational autonomy, the central banks

mandate will become uzzier, and its actions willbecome more dicult to justiy.

Tese are valid concerns. Central bankers will ex-

perience more political pressure than i monetary

policy were primarily targeted at price stability.

Here, however, it is important to remember that

central bank independence is a means to an end

rather than an end in itsel. Limiting the scope o

monetary policy purely or the sake o deending

central bank independence risks undermining the

institutions legitimacy by giving the impressionthat the central bank is out o touch and that it is

pursuing a narrow and esoteric activity that does

not square with its democratic responsibilities.

Ultimately, political reality will thrust responsibility

or nancial stability on the central bank. As hap-

pened in the UK ollowing the ailure o Northern

Rock, the central bank will be blamed or nancial

problems whether or not it was ormally responsi-

ble or supervision and regulation. As lender o last

resort, it will be charged with cleaning up the mess.

It ollows that it would be better o devoting more

o its resources and attention to attempting to pre-

vent the crisis, the elegance and analytical appeal o

the inbergen principle notwithstanding.

Macroprudential Policy Tools

Macroprudential tools are designed to buttress the

stability o the nancial system as a whole, which

is distinct rom ensuring the stability o individ-

ual institutions. Tese tools are intended to help

mitigate externalities and spillovers at the level o

the system as a whole. For example, interlockingclaims and obligations create externalities i the

ailure o one higly leveraged institution threatens

the solvency o other institutions and the stability

o the entire nancial system. Fire sales o assets

may magniy an initial shock and lead to vicious

circles o alling assets prices and the need to de-

leverage and sell o assets. Externalities also arise

over the course o the cycle i the structure o capi-

tal regulation allows an increase in leverage in -

nancial booms while dampening it in busts.

It is useul to distinguish between dierent mac-

roprudential tools that address these dierent as-

pects o nancial risk. In particular, dierent tools

should be used address the time- and cross-sec-

tional dimensions o risk.

Te ime Dimension in Macroprudential

Supervision

In terms o the time dimension, the macropruden-

tial supervisor should develop a range o tools ca-pable o tempering nancial procyclicality. Coun-

tercyclical capital buers, as recommended by the

Basel Committee, are a case in point, although they

are conned to the banking system. A supplement

would be to impose a systemic levy or all levered

nancial institutionsthat is, an additional charge

levied on the unstable (non-core) portion o a


12/42


8

nancial institutions unding, as suggested by the

IMF (2010). Tis levy could be varied over the the

lie o the cycle.

Restraints on bank lending such as loan-to-value

(LV) or debt service-to-income (DI) guide-

lines could useully complement traditional tools

o bank regulation, such as capital requirements.

Capital requirements can themselves consist o

a core o long-dated equity or equity-like instru-

ments supplemented with an additional buer o

contingent capital instruments.

Te interaction between these prudential mea-

sures, as well as their cumulative costs, need to be

careully considered while rolling them out, with

a view to adjusting measures based on experience.And governments should guard against the temp-

tation to use such levies as just a revenue-gener-

ating mechanism rather than a tool to promote

nancial stability.

Some measures (e.g., capital requirements) are

likely to have implications or cross-border compe-

tition between nancial institutions and thereore

may need to be harmonized across countries. Tis

will make it harder to tie them to local economic

conditions, or such harmonization will have to bedone in an objective and mutually agreeable way

across countries. Others like LV or DI guid-

ance need not be harmonized across countries and

could vary substantially with the domestic cycle.

Te systemic levy is a orm o capital charge, mak-

ing harmonization important or countries with

many cross-border banks, something that will ad-

mittedly make it more dicult to tie it to the cycle.

Te Cross-sectional Dimension in

Macroprudential Supervision

In terms o the cross-sectional dimension, policy

should ocus on systemically important nancial

institutions (SIFIs). Better resolution regimes to

deal with ailing nancial institutions could re-

duce the need or reliance on ex ante buers such

as capital. Following the near collapse o Northern

Rock, the United Kingdom was among the rst

to enact a resolution regime that provides super-

visors extensive authority to stabilize a ailing in-

stitution.8 Germany enacted a similar law in Janu-

ary 2011 and the United States is in the process

o empowering regulatory agencies to deal with

uture insolvencies o systemically relevant insti-

tutions. An important complication is that many

systemically relevant institutions are active across

geographical and product borders. Tese new laws

have not been coordinated, and they are unlikely

to be adequate or dealing with a large cross-bor-

der or cross-market ailure. Te new resolution

regimes consequently do not solve the moral haz-ard problem implicit in too big to ail (BF). It

ollows that the implicit public subsidy or BF

institutions remains intact; hence the need or ex

ante measures.

Macroprudential tools could be used to reduce

this incentive to become too big to ail. Tey could

include a systemic risk tax as suggested by the IMF

(2010). Eorts to quantiy systemic risk exposure

or the purposes o regulation are now underway,

but much else remains to be worked out, includingwho would impose this tax, on whom, and under

what circumstances.

Alternatively, surcharges on capital requirements

that vary with the systemic risk they create could

be applied to SIFIs. Te Swiss government com-

mission on BF institutions has shown how this

could be done. In addition to increasing capital bu-

ers to nearly double the level o Basel III, the Swiss

proposal makes the surcharge sensitive to systemic

risk, calculated as a unction o the balance sheetsize and the market share o the institution.

Proposals have also been mooted to eliminate cer-

tain activities o SIFIs (e.g., proprietary trading),

8 Japan enacted an emergency resolution mechanism in 1998, ollowing the banking crisis o 1997. When the emergency term ended, thegovernment set up a permanent resolution mechanism.


13/42


9

ringence certain activities (such as retail banking,

as discussed in the context o the Vickers Commis-

sion in the UK), or even break up SIFIs. Tere is

no consensus among the authors o this report on

what approach is most appropriate. But in devel-

oping all these proposals, care should be taken that

they in act reduce lower systemic risk and do not

just shif risk to entities that are less visible to the

regulatory authorities (including to entities less ca-

pable o managing that risk). Risk that is shunted

out o sight in good times comes back to haunt the

system in bad times.

Finally, supervisors need to identiy direct and in-

direct exposures and linkages, cross border as well

as national, in order to make supervision more

eective. Tey need to identiy institutions andtrades where activity is disproportionately con-

centrated. While collecting the relevant data (on,

or example, inter-bank derivative exposures) or

their own supervisory needs, they should also dis-

seminate more aggregated inormation to market

participants and the general public. Such dissemi-

nation will allow market participants to manage

risks better and allow the public in turn to better

monitor supervisory behavior. While individual

countries now have eorts underway to collect and

disseminate data (or example, the Oce o Finan-cial Supervision in the United States), we are still

some distance rom eective cross-border data

collection and sharing.

Institutional Responsibility

Who should be responsible or nancial stability

at the national level?9 Tere are two answers to this

question. Te coordinated approach gives multiple

institutions (central bank, systemic risk boards,

micro- and macroprudential supervisors) inter-locking mandates, their own instruments, and a

directive to cooperate. In contrast, the unied ap-

proach vests one institution, possibly the central

bank, with multiple mandates and instruments.

Te coordinated approach dominated prior to the

nancial crisis and, despite its ailures, has largely

survived the reorm process. In countries like In-

dia and the United States, administrative bodies

have been set up to coordinate the eorts o mul-

tiple supervisory and regulatory bodies, although

these bodies tend to lack enorcement power. In

Europe, the push or greater regional coordination

has been urther complicated by the superimposi-

tion o an additional layer o supervisory institu-

tions with ew powers o their own. Supervisory

colleges, which collect relevant home- and host-

country supervisors o a large cross-border insti-

tution, are one o the tools or coordination among

countries. But overall, the problem o incomplete

coordination remains.

In particular, the problem that EU-wide banks are

still largely supervised by national regulators is yet

to be ully solved. A new body, the European Sys-

temic Risk Board (ESRB), has been charged with

macroprudential supervision but is endowed with

only weak powers and ew eective instruments.

Te ESRB is large and unwieldy, comprising the

central bank governors and nancial supervisors

o every EU country, plus a number o other unc-

tionaries. Moreover, the ESRB can only issue rec-

ommendations and has no enorcement powers.

While there is little consensus as to the best mod-

el, our contention that nancial stability should

be a core objective o the central bank increases

the weight o arguments or giving central banks

primary responsibility or regulatory matters. I

central banks have a mandate to ensure nan-

cial stability and also the powers needed to wield

macroprudential corrective instruments, they can

optimally choose trade-os between the use o

the interest rate instrument and macroprudentialmeasures. Moreover, the central bank will have,

or should have, its nger on the pulse o nancial

markets through its monetary policy operations. It

possesses a sta with macroeconomic expertise. It

9 Alternatively, at the regional level in places where multiple national economies share a single central bank (e.g., Euroland).


14/42


10

is the one institution with the balance sheet capac-

ity to act as lender o last resort.

Tere are also compelling arguments against a uni-

ed model. One disadvantage is that it makes the

central bank more susceptible to political interer-ence. Te central bank will have to work hard to

establish the legitimacy o its actions in circum-

stances where the nature o threats to nancial sta-

bility may be poorly understood and its actions are

unpopular. Te public and its elected representa-

tives may not be happy, or example, i the central

bank curbs credit growth and causes asset prices

to all, and they will pressure the authorities to re-

verse course.

Te unied model may also pose a conict o in-terest or the central bank, which may, or example,

be tempted to keep interest rates ariticially low in

an eort to aid distressed nancial institutions, or

to treat a bank acing a solvency problem (a mat-

ter properly addressed by the scal authority or its

agents) as i it were acing a liquidity problem.

I, on balance, the decision is to make the central

bank the macroprudential supervisor, this ap-

proach should go hand in hand with measures to

strengthen its independence rom political pres-sure. o this end, it is important or the central

bank to participate in the public discussion o how

its perormance will be evaluated. More regular

communication o the rationale or its policies will

also become increasingly important.

In sum, there are advantages to both models, and

individual countries institutional characteristics

and political settings will determine what works

best. Whatever the mechanism, it is clear that e-

ective coordination between monetary and nan-cial regulatory policies will be the lynchpin o -

nancial stability.

Exchange Rates and Monetary

Policy

Te external dimension o monetary policy is criti-

cally important or small open economies with open

capital accounts. Capital ows and exchange ratemovements are important or price-level develop-

ments. Tey are important or nancial stability as

well: in open economies, monetary policy may have

limited eectiveness in inuencing credit develop-

ments because, inter alia, nancial intermediaries

can substitute external unding or domestic unding.

Macroprudential tools that lean against credit de-

velopments can give the central bank some mea-

sure o monetary policy autonomy, weakening the

link between domestic monetary policy and capi-tal inows. For instance, by leaning against credit

expansion, the central bank may be able to reduce

the incentive or banks to borrow externally when

domestic interest rates are increased.

Te tensions between these dierent acets o eco-

nomic stabilization become more acute when the

currency is strong relative to undamentals and the

government wants to prevent excessive apprecia-

tion. Tis puts the central bank in a corner when

domestic demand is also too strong. Tere is thenthe need to cool an overheating economy by allow-

ing the appreciation o the currency, on the one

hand, but pressure to guard against the erosion o

competitiveness rom what might prove to be only

a temporary appreciation, on the other. Capital

controls that moderate nancial inows, especially

short-term inows that are channeled through the

domestic banking sector, may alleviate the policy

dilemma but their role as a legitimate part o the

policy makers toolbox remains controversial.

Much commentary takes or granted that capi-

tal controls dont work.10 Commentators making

10 See, or instance, the ollowing editorial in the Wall Street Journal: Capital-Control Comeback: As Money Flows to Asia, Politicians Play KingCanute, 2010, June 17. http://online.wsj.com/article/SB10001424052748704289504575312080651478488.html
http://online.wsj.com/article/SB10001424052748704289504575312080651478488.htmlhttp://online.wsj.com/article/SB10001424052748704289504575312080651478488.html


15/42


11

such claims typically assume that the objective is

either to hold down the exchange rate or to sup-

press the total volume o inows. In this approach

the emphasis is on the exchange rates inuence on

the trade balance and thus also the attempt to hold

back currency appreciation by limiting nancial

inows, whatever their precise orm.

But i capital controls and related macropruden-

tial measures are seen not as instruments o ex-

change rate management but as part o a package

o policies targeted at nancial stability, then it is

the composition o capital ows that takes center

stage rather than their volume.11 Foreign direct

investment (FDI) and portolio equity ows are

less likely to reverse direction abruptly. And even

when portolio ows do reverse, the impact onunding may be less damaging than any sudden

loss o access by the banking sector. Foreign sell-

ers o stocks in a crisis ace the double penalty o

lower local currency prices when they sell and a

sharply depreciating exchange rate, the implica-

tion being that the dollar-equivalent outow as-

sociated with repatriation o portolio equity sales

proceeds tends to be small compared to the pre-

crisis marked-to-market value o oreign holdings

o equity. And the typical equity investor (such as

a pension und or mutual und) is not leveraged.

In contrast, when oreign unding o the banking

sector evaporates abruptly, the consequences are

more damaging. I the local bank is leveraged and

debt is denominated in dollars, then outows can

set o the well-known cycle o distress in which

belated attempts by banks to hedge their dollar ex-

posure drives down the value o the local currency,

making the dollar-denominated debt even larger.12

I the crisis erupts afer a long build-up o such

mismatches, the coincidence o the banking crisiswith the currency crisis (the twin crisis) can un-

dermine banking sector solvency, with signicant

economic costs.

Capital controls are not, o course, the only tool or

dealing with inows. Microprudential tools such

as minimum capital ratios should be part o the

policy response. Even these tools, however, may

not be enough to dampen the upswing o the cycle.

Bank capital ratios ofen look strong during booms

when banks are protable and the measured qual-

ity o loans is high. In addition, the application

o discretionary measures, such as higher capital

requirements, must surmount concerted lobbying

by vested interests that benet rom the boom.

Currency appreciation may also help to moder-

ate the size o capital inows, as oreign investors

perceive less o a one way bet. However, when

banking sector ows orm the bulk o the inows,

merely allowing the currency to appreciate maynot suce. Te behavior o banks and other lever-

aged institutions is additionally inuenced by their

capital position and their perception o risks. Cur-

rency appreciation and strong protability coupled

with tranquil economic conditions can be seen by

banks as a cue to expand lending rather than to

curtail their activity.

In sum, capital controls can, under some circum-

stances, be useul or managing maturity and

currency mismatches and, in particular, or ore-stalling dollar shortages in the banking system.

Judiciously employed along with other macropru-

dential policies, they can reduce nancial instabil-

ity as well as boom-bust cycles, thereby serving as

a useul complement to conventional monetary

policy instruments. As with other instruments,

care should be taken that they are used to reduce

macro-economic volatility rather than merely

to suppress it, only to see it emerge in other, po-

tentially more destructive ways. Moreover, with

capital accounts becoming more open and giventhe increasing ungibility o unds across dierent

orms o capital, even controls limited to specic

types o capital ows are becoming an increasingly

11 For an extensive discussion, see Ostry, Ghosh, Habermeier, Chamon, Qureshi, and Reinhardt (2010).12 Figuratively, the attempt to clamber out o the ditch by buying dollars merely drags others into the ditch.


16/42


12

weak substitute or good macroeconomic and pru-

dential policies.

Conclusion

Tis chapter has made the case or augmenting thetraditional narrow price stability ocus o mon-

etary policy with the additional goal o nancial

stability. Te conventional separation in which

monetary policy targets price stability and micro-

prudential policies target nancial stability, and

the two sets o policies operate independently o

each other, is no longer tenable.

Tis has a number o implications.

Policy makers need a new set o policiesthat are macroprudential in nature, target-

ing the build up o risks to nancial stabil-

ity. Tese policies range rom countercy-

clical capital ratios to capital controls.

Te neat inbergen separation o two

tools or two objectives is no longer ea-

sible. Interest rates aect nancial stability

and, hence, real activity. Equally, macro-

prudential tools impact credit growth and

external imbalances, which have conse-

quences or macroeconomic and price

stability. Central bankers thereore will

have to consider tradeos as they optimize

among their policy tools to achieve their

multiple objectives.

We believe that explicit recognition o such trade-

os will, in some cases, move theory closer to prac-

tice. In other cases it will make adopting ination

targeting more attractive insoar as the ramework

now recognizes issues that some policy makershitherto thought were missing. And in the case o

the ew who still adhere to narrow ination target-

ing, it might prompt a welcome reconsideration.


17/42


13

CHAPTER 3

Cross-Border Spillovers

I

n the last chapter we discussed how national

monetary policy rameworks should be re-thought to better incorporate nancial-stability

considerations. But there is another equally im-

portant reason or rethinking the ramework: in-

ternational spillovers.

I national policies have important cross-border

eects, then there is a prima acie case or coordi-

nating them internationally. Tis observation was

o course the main point o the voluminous 1980s

literature on spillovers and policy coordination.

But it has since been rendered more compelling bychanges in the world economy in the last quarter

century. Te world today is more connected than

ever by cross-border nancial ows. Te policy

choices o individual countries, especially those o

large, systemically signicant countries, can have a

substantial impact on their neighbors. When gov-

ernments and central banks change their macro-

economic policy stance dramaticallyas they did

in the recent world nancial crisisthe spillovers

on other nations can be sizeable.

Cross-border spillovers may also have increased as a

result o the nature o policy responses to economic

shocks and business cycle conditions. A commonly

voiced concern is that unconventional monetary

policies may have especially large and complex

cross-border spillovers. For instance, monetary in-

jections when the nominal interest rate is at its zerobound might result in capital outows rather than

in more domestic activity, i domestic demand is

weak and banks are reluctant to lend.13

And while concern in the 1980s centered on the

interaction o the United States and Europe, two

economic blocs with oating exchange rates, spill-

overs today involve one bloc that oatsthe major

advanced countriesand one, led by China, with

xed or semi-xed exchange rates. Tis asymme-

try gives rise to important new issues.

In this chapter we review various channels or in-

ternational transmission o domestic policies and

discuss their implications. We then discuss the

tensions that arise in reconciling domestic mon-

etary policies with the larger objective o global

macroeconomic and nancial stability.

Cracks in the Framework of

(Mostly) Flexible Exchange RatesTe international properties o the de acto regime

o exible exchange rates were never as desirable

as asserted by its champions. o start with, the new

regime was not, in act, universally adopted. It was

13 Tis combination o circumstances is not unusualwitness what happened during the recent nancial crisis.


18/42


14

not widely adopted in Asia, or example, where de

jure or de acto pegging remained the reality and a

large volume o oreign exchange reserves was ac-

cumulated in the 2000s, contrary to the presump-

tion that reserves would become superuous with

the breakdown o the Bretton Woods system o

xed exchange rates.

Moreover, large current-account surpluses and

decits (imbalances) persisted over much o the

last decade without prompting macroeconomic

and exchange-rate responses. Imbalances persist-

ed in countries with very dierent exchange rate

arrangements, including countries that did not

maintain dollar pegs, such as Japan and Germany.

Questions also remained about the ability o ina-tion targeting cum oating exchange rates to cope

with the volatility o international capital ows.

While stability-oriented monetary policies at the

national level could help to limit the magnitude

o sudden inows and reversals, and while strong

regulatory and supervisory rameworks could help

limit their consequences, it was unclear whether

such measures would be sucient to protect

emerging economies rom macroeconomic and

nancial instability.

Nor did the I-oating ramework eliminate the

special role o the dollar as the key international

currency. Te dollar remains the worlds most im-

portant reserve currency and a leading invoicing

currency or international trade. It is also the cur-

rency that underpins the global banking system as

the unding currency or global banks. Tis raises

important questions about access to dollar liquid-

ity by non-US banking systems in times o stress.14

Reconsidering the Conventional Wisdom

In light o the nancial crisis and subsequent de-

velopments, several reasons have emerged or re-

visiting the conventional wisdom:

Convergence towards the ination target-

ing cum exible exchange-rate ramework

remains incomplete. While a large part o

the world economy has adopted this mod-

el, some ast-growing emerging markets

have not. Te coexistence o oaters and

xers thereore remains a characteristic

o the world economy. It can even be said

that the incidence o pegging has risen

over time with the export drive o East

Asia and, toward the end o the most re-

cent decade, the rise o the relative price

o oil.

Te period in which the I regime was test-

ed was exceptionally benign. Chinas en-

try into global trade and other emergingmarkets acted as a strong disinationary

orce, making or price stability globally.

Commodity prices remained subdued un-

til the late 2000s, and there were ew in-

ation spillovers. Since then the situation

has changed. In a new context where com-

modity prices respond strongly to aggre-

gate demand, a major question is whether

central banks take into account spillovers

through global commodity prices when

making monetary policy decisions.

Capital market spillovers between ad-

vanced and emerging economies have

grown. While Obstelds (2009) character-

ization o the world economy as compris-

ing a single nancial system may not apply

to all countries, it is certainly correct or

North America, Europe, East Asia, and

a number o emerging market countries.

Private gross capital ows to and rom

both the US and Europe grew massively inthe course o recent decades. o be sure,

this was in large part or reasons indepen-

dent o monetary policy, including nan-

cial liberalization, the unique role o the

14 For an extensive discussion o these issues see Farhi, Gourinchas, and Rey (2011).


19/42


15

US as supplier o sae nancial assets, and

the attractiveness o emerging markets

as destinations or investment. Still, the

resulting nancial interpenetration im-

plies that the stock o diversiable assets

and cross-border holdings that respond

to changes in monetary conditions have

grown enormously.15 Tis creates chal-

lenges or countries on the receiving end

o capital ows. In practice, many o those

recipients are emerging market economies

that are struggling to prevent the surges in

capital inows rom leading to exchange

rate misalignment and unsustainable

lending booms.

Unconventional monetary policies are likelyto accentuate international spillovers. Such

policies are typically undertaken when

traditional instruments are exhausted and

traditional channels have ceased working.

In such situations, unconventional poli-

cies could result in less domestic demand

creation and more demand shifs between

countries. Critics argue that purchases by

central banks o long-dated bonds and pri-

vate-sector-issued securities create liquid-

ity that can spill abroad (because domesticchannels or credit creation are blocked),

causing capital ows to and undesirable

relative price changes in other countries.16

Central banks in countries conducting

quantitative easingthe US Federal Re-

serve and the Bank o Englandargue

that Quantitative Easing (QE) is no dier-

ent conceptually rom conventional mon-

etary policy but merely its continuation

through other means in a situation where

interest rates approach the zero bound.Central banks in several emerging market

countries, in contrast, claim that QE is a

beggar-thy-neighbor strategy.

Tese observations suggest that convergence to-

wards a common policy template in the 2000s was

not general. Moreover, where convergence has

take place, it may not last long in view o the chal-

lenges currently conronting monetary policy. It is

thereore important to assess whether a reormed

consensus can and will be ormed and to contem-

plate its implications or the conduct o monetary

policy and or the own house in order doctrine in

particular.

Challenges to the IT-plus-foating Regime

1. Uneasy coexistence: oaters and fxers

Te idealized I-plus-oating ramework has not

worked out as anticipated, because countries havenot converged to similar monetary and exchange

rate arrangements.

In Latin America, a substantial number o coun-

tries, some o them large and economically im-

portant, resist moving in this direction. While the

two largest countries Brazil and Mexicoand an

important set o middle-sized and small nations

Colombia, Peru, Chile, Uruguayhave adopted

it, another sizeable group including Venezuela, Ar-

gentina, Bolivia, and Ecuador continues to pursuexed or semi-xed exchange regimes, sometimes

with multiple exchange rates or dierent current

and capital account transactions. Few countries

in the Middle East and Arica have converted to

I plus oating, though economically important

South Arica has adopted it.

In Asia, several countries have adopted the rame-

work, albeit with dierent degrees o commit-

ment. Ination targets are explicit in Tailand,

Korea, Indonesia, and the Philippines. In Tailandand Korea, low and stable ination was achieved

in the 2000s. Singapore has achieved low and sta-

ble ination using a basket-based exchange rate

15 Lane and Milesi-Ferretti (2001) and Kubelec and S (2010) provide a quantitative account o nancial integration and the participation in it omajor emerging economies.

16 See Portes (2010) or a discussion.


20/42


16

regime, since the economy is small and highly open

to nancial ows. Usually, however, Asian central

banks have multiple objectives: growth, price sta-

bility, and exchange rate stability, some o which

temper the conventional ramework. It is air to

say that many East Asian countries deal with in-

ation more on the basis o discretion than pre-

set rules. In Cambodia and Vietnam, dollarization

and the lack o independence o the central bank

is a serious problem in stabilizing ination. India

has a hybrid regime without an explicit ination

objective and with exchange rate management in

principle limited to moderating sharp movements

in the currencys value.

China is the largest nation with a managed ex-

change rate. Te renminbi was delinked rom itsUS dollar peg in 2005 but remains tightly managed

against the dollar. Among the explanations or this

choice o exchange rate regime are the governments

objective o promoting export-led growth. Another

is the desire to sel-insure against external shocks

by accumulating a large stock o reserves. Chinas

oreign exchange reserves now exceed $3 trillion,

dwarng by a wide margin all evaluations o the

reserve buer necessary to insure against sudden

stops o inows or a surge o capital outows.

National and regional dierences aside, a common

eature o policies in these countries is a reluctance

to allow exchange rates to move as much as needed

to accommodate external disturbances, especially

those originating in the capital account. Non-oat-

ers monitor nominal and sometimes also real ex-

change rates and use not just oreign exchange mar-

ket intervention but a whole array o instruments to

prevent unwanted exchange rate movements.

In sum, notwithstanding the perceived success oination targeting with exible exchange rates,

countries operating a reely oating exchange-

rate regime, whether measured in terms o global

GDP or global exports, have not increased over

the last two decades. o the contrary, the share

o such countries, so measured, has actually de-

clined (Figure 1).

Te main consequence is that the adjustment

mechanism implied by the standard I-plus-oat-

ing arrangement has not been allowed to operate.

Tis is one explanation or the size and persistence

o global imbalances. According to the IMFsWorld

Economic Outlook, these imbalances reached 3%

o world GDP in 2007, beore the advent o the

crisis.17 Te subsequent crash then reduced cur-

rent account decits in countries such as the US

and the UK as their demand or imports dropped

sharply. But according to the April 2011 WEO, im-

balances once again began to grow starting in 2010

and will hover around 2% o world GDP betweennow and 2016.

A prominent instance o the uneasy coexistence

o oaters and xers is the tug o war between

US monetary policy and exchange rate policy in

emerging market xers such as China. A highly

stimulative US monetary policy is potentially u-

eling ination elsewhere, including in emerging

markets that have closed their output gaps and are

acing inationary pressures. O course, emerg-

ing market central banks could raise interest ratesmore rapidly, but they would then attract capital

inows and experience aster exchange rate appre-

ciation. Meanwhile, emerging market resistance

to exchange rate appreciation is limiting export

and employment growth in industrial countries

already experiencing high and persistent unem-

ployment. In normal circumstances, the United

States and other advanced economies would ad-

just by cutting interest rates. But these countries

are already at the zero bound. In this context, the

exchange rate policy o emerging market xers isimposing a negative demand externality on the ad-

vanced economies. In tandem with the ination-

ary externality imposed by US monetary policy,

this has created severe policy complications or

17 Tat is the size o the current account surpluses in countries like China, Japan, Germany, Switzerland, and the oil producers, matched (up toerrors and omissions) by the corresponding decits in the US, the UK, Spain and elsewhere.


21/42


17

Figure 1:Shares o countries under alternative exchange-rate regimes in world GDP and

world exports, 1980-2007

World GDP

World Exports

Source: Angeloni et al. (2011). Calculations are based on the Ilzeztki-Reinhart-Rogo classication. Euro area countries aretreated separately throughout in order not to introduce a break in the series.


22/42


18

other countries, especially emerging markets that

are oaters.

Collective action problems arise rom these asym-

metric exchange rate arrangements. Many emerg-

ing market countries in East Asia, even those that

ostensibly oat, explicitly or implicitly monitor

their real exchange rates. Tey are reluctant to see

their currencies appreciate excessively, especially

relative to other countries in the region. Tis reluc-

tance hinders nominal exchange-rate adjustment

between East Asia and the advanced economies at

a time when asymmetries between the two groups

urgently call or real exchange-rate adjustment.

Concerns about exchange rate appreciation and

overshooting are not limited to the emerging mar-kets, o course. Te recent intervention in oreign

exchange markets by committed oaters such

as Japan and Switzerland highlights the tensions

building up in the global economy as public debt

levels in the major reserve currency areasthe

US and Europeimpose more o a burden on the

Federal Reserve and the European Central Bank to

maintain lax monetary policy with attendant spill-

overs to the rest o the world (as discussed in more

detail in the next chapter).

Fixing also creates policy dilemmas or countries

seeking to x. Tese countries are by choice de-

pendent on their partners monetary policy de-

cisions, especially but not only when they have

opened the nancial account. Attempting not to

import oreign monetary conditions while xing

has required extraordinary measures.

ake China, whose capital account is only partial-

ly open. Experiencing large balance o payments

surpluses, the Peoples Bank o China (PBOC)has regularly intervened in the oreign exchange

market to limit the appreciation o the renminbi.

Te resulting increase in Chinas oreign exchange

reserves accounts or almost all the increase in

Chinas monetary base. o sterilize the increase

in the money supply created by its intervention in

the oreign exchange market, the PBOC has been

orced to sell all o its holdings o government se-

curities and to sell central bank bills to state-owned

commercial banks. Tis strategy has been abetted

by repressed interest rates, creating distortions innancial markets and in eect taxing households

who receive negative real returns on their massive

stock o bank deposits.

Te nancial crisis heightened these tensions. Its

size and depth increased the incentive or emerg-

ing markets experiencing sharp capital ow rever-

sals to sel-insure by accumulating even larger re-

serves.18 Moreover, the instability o world demand

has caused a number o countries, not all o them

in Asia, to place an even greater premium on man-aging the level o the real exchange rate. Tis has

led them to deploy a broad array o tools, includ-

ing capital controls, to prevent unwanted apprecia-

tion (or a more detailed discussion o this issue,

see Chapter 4 below).

Tere are two possible assessments o these trends.

One minimizes the importance o the asymme-

try o exchange rate policies on the grounds that

what matters or international adjustment is real

exchange rates, which governments cannot controlin the long run. Tus, recent price and wage ina-

tion in China is causing non-trivial appreciation o

the renminbi in real terms vis--vis the dollar even

while the nominal bilateral exchange rate remains

relatively stable.

Te alternative view, which we share, is that inter-

national adjustment via wage and price ination

is slow and inecient. Te world economy would

be better served by a speedier mechanism involv-

ing greater exchange rate exibility. I exibilityis not easible or domestic political reasons, then

incentives need to be put in place to make sure

large nations among both groupsxers but also

oatersinternalize the international eects o

their actions.

18 Tat actor alone suggests that xed or semi-xed exchange rate arrangements will be around or some time in emerging markets.


23/42


19

2. Controlling ination in a less benign

environment

For the second time in three years, rising commodi-

ty prices are uelling global ination. Tis ination-

ary pressure is superimposed on the background o

still-large output gaps and high unemployment in

virtually all advanced countries. Tis combination

is problematic or an ination-targeting strategy in

which central banks ocus on the components o

ination that are under their direct control. Indeed,

or central banks in commodity-importing coun-

tries, a rise in oil or commodity prices is an exog-

enous supply shock, and the standard model says

that the central bank should only respond to the

extent that the shock has second-round eects and

increases expected uture ination.

argeting domestically-generated ination was an

appropriate strategy and did not raise concerns

about collective action problems in the 1990s and

the early 2000s, when an ample supply o com-

modities and the entry o China and other devel-

oping countries into the global labor orce helped

subdue global ination. Against the background o

a steep global commodity supply curve, however,

expansionary monetary policies by major econo-

miesadvanced and emerging alikemay createnegative externalities that are not adequately inter-

nalized in the standard ramework.

Tis shortcoming is especially evident in the strict

ination-targeting ramework in which the central

bank commits to keeping the orecast rate o ina-

tion (conditional on market expectations or the

policy rate) on target. In this setting, the global en-

vironment is taken as given and is not aected by

domestic monetary policy responses. As a conse-

quence, the global monetary policy stance is likelyto be suboptimal.

In small open economies, monetary policy is rea-

sonably geared to domestic objectives. Te same,

however, does not apply to the large-economy

central banks, such as the Fed, the ECB, and the

PBOC. Tese economies are large enough or their

policy choices to involve signicant externalities.

It would thereore be desirable that these central

banks, and perhaps a handul o others, include

in their policy objective a measure o these e-

ects. Clearly, however, such a move would involve

a collective-action dimension, which calls or anexplicit dialogue among these central banks about

the amendment o their policy rameworks. We re-

turn to this later.

3. Financial channels o transmission

In the idealized world in which all central banks

pursue I and allow their exchange rates to oat,

an individual central banks monetary policy ac-

tionssay, a cut in the interest rateare transmit-

ted to the rest o the world mainly through twochannels:

Te cut in local interest rates stimulates

domestic demand, some o which spills

over to additional imports. Te magni-

tude o this eect on the rest o the world

depends on the countrys share o world

GDP.

Te countrys nominal and real exchange

rates depreciate, shifing demand awayrom the rest o the world. Again, the size

o this cross-border eect depends on the

size o the country in question.

In this stylized model, capital ows only have an

indirect role, with the potential or outows rom

the country undertaking an expansionary mon-

etary policy causing movements in the value o its

currency. Prices bear the burden o adjustment.

In contrast, recent experience points to the exis-tence o additional channels whose role and im-

pact may well be large and potentially destabiliz-

ing. While the act that the impact o capital move-

ments can dwar that o the more traditional trade

eects has long been understood, the new and

novel observation concerns the size o the cross-

border capital movements triggered by the supply


24/42


20

o liquidity or small changes in interest rates in ad-

vanced countries. Tis reects the accumulation o

a huge pool o ootloose assets responsive to small

changes in expected returns.

Te composition o these investment portolios is

interest-rate sensitive and likely to respond sharp-

ly to dierences in expected rates o economic

growth in recipient countries. An example is the

massive capital ows to emerging markets in 2010

in response to the growth slowdown and record-

low interest rates in major advanced countries.

Policy spillovers to the rest o the world can be

sizeable in the case o the United States, which

hosts branches o some 160 oreign banks whose

main unction is to raise wholesale dollar undingin capital markets. Foreign bank branches collec-

tively raise over one trillion dollars o unding, o

which over 600 billion dollars is channeled to their

headquarters outside the United States.19

Although the United States is the single largest net

debtor, it is a substantial net creditor in the glob-

al banking system. In eect, the US borrows long

through the issue o treasury and other securities

while lending short through the banking sector.

Tis is in contrast to countries like Ireland and Spainthat nanced their current account decits through

their respective banking sectors, which subsequent-

ly aced runs by their wholesale creditors.

Some borrowed dollars will nd their way back to

the United States. But many will ow to Europe,

Asia, and Latin America, where global banks are

active local lenders. At the margin, the shadow val-

ue o bank unding will be equalized across regions

through the portolio decisions o global banks,

making global banks the carriers o dollar liquidityacross borders. In this way, permissive US liquidity

conditions are transmitted globally, and US mon-

etary policy becomesglobalmonetary policy.20

An additional channel o transmission is through

commodity prices. Low interest rates in the G-3

countries have a tendency to push up primary-

commodity prices, both because the associated

low borrowing costs mean high consumption and

investment demand or these products, including

rom emerging markets, and because a low inter-

est rate reduces the nancial cost o holding stocks

o storable commodities, thus making them more

attractive as investment vehicles.

From the point o view o a commodity-producing

country, lower world interest rates thus improve

the terms o trade and increase local wealth and

creditworthiness. A rating upgrade may ollow. All

this makes the country even more attractive or

ootloose international capital, creating pressuresor currency appreciation.

Tese cross-border eects can be magnied by

dierences in exchange rate regimes. In recipi-

ent countries with reely oating exchange rates,

standard theory suggests that the local currency

should appreciate in response to a cut in oreign

interest rates. It could even appreciate beyond its

new steady-state level on impact, beore then de-

preciating until reaching its new equilibrium level.

But i the country in question has a managed oat

or semi-xed exchange rate, the required apprecia-

tion will not occur on impact. Even so, expecta-

tions o appreciation will eventually set in, making

it more attractive to shif capital toward the coun-

try. Tis may bring orth additional inows, in

turn creating additional pressure or the exchange

rate to strengthen.

Te situation is even more complicated i interven-

tion in the oreign exchange market is sterilized. Teneed to issue local bonds to mop up the liquidity re-

sulting rom the purchase o oreign exchange may

cause local interest rates to rise, attracting even more

19 Bank or International Settlements (2010).20 See also Cetorelli and Goldberg (Forthcoming).


25/42


21

inward capital ows. And since local interest rates

are likely to be higher to begin with (i the recipient

country is an emerging market), this sterilization

will be expensive. I sustained over a suciently

long period, sterilized intervention can weaken s-

cal accounts, causing expectations o monetization

and higher ination, which in turn will cause local

nominal rates to go up. Tis, in turn, can call orth

yet another round o destabilizing capital inows.

Te conventional view o international spillovers

has also relied on the assumption o smoothly-

adjusting international capital markets, something

that seems less than tenable today. Te 2007-09

nancial crisis serves as a reminder that nancial

ows can reverse abruptly, placing intense pressure

on the unctioning and integrity o markets andmarket participants. Tis has been pointed out re-

peatedly afer recent capital-account currency cri-

sesMexico, Asia, Russia, Brazil, and Argentina.

What is new in the 2007-09 crisis was that it hap-

pened even in some advanced countriesor ex-

ample, some European economies, such as Ireland.

A nation previously ooded with capital can thus

become the subject o a sharp reversal in ows.

Margin and borrowing constraints can suddenly

become binding, leading to a painul process odeleveraging. I the need to raise cash causes one

round o asset sales, the prices o those assets will

all, reducing the value o collateral and calling orth

urther asset sales and additional price drops. Tis

can cause massive destruction o value, as rms nd

themselves liquidity-constrained and abandon un-

nished potentially protable investment projects.

Policy makers in countries on the receiving end o

these ows ace an unappetizing choice. I they al-

low the currency to appreciate, they expose them-selves to accusations o overvaluation, loss o com-

petitiveness, and de-industrialization. But i they

ght the appreciation via intervention, they may

nd themselves on the receiving end o ever-larger

inows. Te central bank may end up allowing

some appreciation anyway, but not beore accu-

mulating a large stock o expensive domestic li-

abilities and a large stock o international reserves

on which it will take a capital loss (in domesticcurrency terms) i and when the exchange rate ad-

justment eventually happens.

While the conventional model o I-plus-oating

acknowledged these complications, it did not place

them at the center o the analysis. o the extent

countries targeted core ination, spillovers through

global commodity prices were lef unattended. Tis

was not a serious concern in the1980s and 1990s,

the period o the Great Moderation, but is a more

serious one in the presence o large global imbal-ances and the need to accommodate large stocks

o internationally mobile capital looking or yield.

4. Normal versus crisis times

Te conventional wisdom was developed in tran-

quil times. In crises, in contrast, central banks have

resorted to an array o non-conventional mon-

etary policies such as quantitative easing (QE)

the printing o money to buy bonds. What do such

policies imply or the question o internationalspillovers o monetary policy?

One view is that unconventional policies are no

dierent rom conventional policies in their cross-

border implications. I oating exchange rates can

adjust to make international coordination o con-

ventional policies unnecessary, then the same must

be true o unconventional policies. Tis was the

view o the United States ollowing the adoption

o QE2. In response to complaints rom emerging

market policy makers who eared the wave o li-quidity coming their way, Fed ocials essentially

argued that, everything will be okay i you just let

your currencies appreciate.21

21 As indicated, or example, by the ollowing excerpt rom the speech by Fed chairman Ben Bernanke on November 19, 2010 at the ECB CentralBanking Conerence: An important driver o the rapid capital inows to some emerging markets is incomplete adjustment o exchange rates inthose economies, which leads investors to anticipate additional returns arising rom expected exchange rate appreciation.


26/42


22

Te alternative view is that beggar-thy-neighbor

impacts are greater when using unconventional

instruments. Te diculty arises in evaluating

whether the use o such instruments is consistent

with the normal policy ramework or represents

an attempt mainly to weaken the currency and

boost exports in the absence o a positive domes-

tic demand response. Te same causes that jus-

tiy recourse to unconventional policies make the

ination-targeting compass lose precision. When

ination signicantly undershoots its target and

central banks resort to instruments with which

they have little experience, it is much harder to

say whether a policy stance is in line with the I

ramework or whether it represents an attempt at

competitive devaluation.

In addition, spillovers may work dierently in times

o crisis. During a crisis, local credit demand is likely

to be weak and banks willingness to lend domesti-

cally will be especially limited. For every additional

dollar o liquidity that is created by monetary policy,

a larger share will end up abroad in crisis times than

in normal times, thereby depreciating the exchange

rate at the expense o trade partners. It ollows that

spillovers are potentially larger during episodes o

local nancial distress.

Te presence o international spillovers suggests

that coordination can lead to better global out-

comes. In addition, the current situation high-

lights the need or principles and procedures or

deciding when an unconventional monetary pol-

icy is beggar-thy-neighbor in its eect. In turn,

these principles should orm the basis or correc-

tive action.

Conclusion

Te cross-border spillovers rom monetary policy

provide yet another reason or rethinking not justthe domestic monetary policy ramework but also

mechanisms or ensuring compatibility between

large-country policies. We will turn to recommen-

dations that ollow rom this analysis in Chapter

Five. But beore oering recommendations, we

turn to a discussion o some additional policy bur-

dens on central banks in the afermath o the crisis.


27/42


23

CHAPTER 4

Additional Pressures on Central Banks

I

n this chapter we describe how central banks are

coming under additional pressures in the post--nancial-crisis environment. While some o these

additional pressures are not entirely new, they

threaten to orce central banks onto risky terrain.

We highlight two sets o pressures: (a) the conse-

quences o high public and private debts; and (b)

the perceived dangers o currency appreciation

and overvaluation.

While maniestations o these pressures are already

evident in individual countries, it is important tounderstand them as part o a broader global picture.

We do so in the next two sections, which look at the

consequences o high public and private debts in the

advanced economies and at worries abo

Date post:	07-Apr-2018
Category:	Documents
Upload:	bruegel
View:	220 times
Download:	0 times

Rethinking Central Banking Web

Documents