II [~IF~MAY/JUNE 1995

R. Glenn Hubbard is Russell 1. Carson professor of economics and finance at Columbia University. The author is grateful to Allen Berger, Phillip

Cogan, Richard Cantor, Mark Gertler, Simon Gilchrist, Anil Kashyop, Don Morgan, Glenn Rudebusch, Bruce Smith and participants at the confer-ence for helpful comments and suggestions. The authar also acknowledges financial support from the Center for the Study of the Economyand the State of the University of Chicaga, and the Federal Reserve Bank of New York.

Is There a“Credit Channel”for MonetaryPolicy?

P. Glenn Hubbard

nderstanding the channels through which/ monetary pohcy affects economic van

ahies has long been a key research topicin macroeconomics and a central element ofeconomic policy analysis. At an operationallevel, a “tightening” of monetary policy by theFederal Reserve implies a sale of bonds by theFed and an accompanying reduction of bankreserves. One question for debate in academicand public policy circles in recent years iswhether this exchange between the central

bank and thebanking system has consequencesin addition to those for open market interestrates. Ac the risk of oversimplifying the debate,

the question is often asked as whether thetraditional interest rate or “money view”channel presented in most textbooks is aug-mented by a “credit view” channel.’

There has been a great deal of interest inthis question in the past several years, moti-vated both by developments in economic

models (in the marriage of models of infor-mational imperfections in corporate finance

with traditional macroeconomic models) andrecent events (for example, the so-called creditcrunch during the 1990-91 recession),’ As I

elaborate below, however, it is not alwaysstraightforward to define a meaningful creditview alternative to the conventional interestrate transmission mechanism. Similar diffi-culties arise in structuring empirical tests of

credit view models,This paper describes and analyzes a broad,

though still well-specified. version of a creditview alternative to the conventional monetarytransmission mechanism. in so doing, I

sidestep the credit view language per se, andinstead focus on isolating particular frictionsin financial arrangements and on developingtestable implications of those frictions, Toanticipate that analysis a bit, I argue thatrealistic models of “financial constraints” onfirms’ decisions imply potentially significanteffects of monetary policy beyond thoseworking through conventional interest ratechannels. Pinpointing the effects of a narrow“bank lending” channel of monetary policyis more difficult, though some recent modelsand empirical work are potentially promisingin that regard.

I begin by reviewing the assumptionsand implications of the money view of themonetary transmission mechanism and bydescribing the assumptions and implicationsof models of financial constraints on borrow-ers and models of bank-dependent borrow-ers- The balance of the article discusses thetransition from alternative theoretical modelsof the transmission mechanism to empiricalresearch, and examines implications for

monetary policy

/7/

— 7

Before discussing predictions for theeffects of alternative approaches on monetarypolicy, it is useful to review assumptionsabout intermediaries and borrowers in thetraditional interest rate view of the monetarytransmission mechanism. In this view, finan-cial intermediaries (banks) offer no specialservices on the asset side of their balance sheet -

On the liability side of their balance sheet,banks perform a special role: The bankingsystem creates money by issuing demanddeposits. Underlying assumptions aboutborrowers is the idea that capital structuresdo not influence real decisions of borrowersand lenders, ahe result of Modigliani and

Miller (1958). Applying the intuition of theModigliani and Miller theorem to banks, Fama(1980) reasoned that shifts in the public’sportfolio preferences among bank deposits,

‘For descriptons of the debate, see

Bernanke and Blinder (19f8) andBernanke (1993).

‘for on analysis of the ‘creditcrnnch’ episode, see Kliesen andTotam (1992) and the studies inthe Federal feserne Bank of NewYork (1994). The paper by Cantarand Rndrigues in the New York Fedstudies considers the possibility of a

credit cmoch far nonbonk interme-diaries.

FEDEEAL RESERVE SANK OF ST. LOUIS

63

DII~I1t~MAY/JUNE 199$

Fama’s insight amplifies the endiercontribution of Brainard and Tobin(1963) that monetary policy con

he analyzed through its effects oeinvestor portfalias.

Mare fenerolly, in a model withmany assets, this description wouldassign to the money view of thetransmission mechanism effects onspending arising fmm any changesin the relative prices of assets,

While this simple tmoasset-mndeldescription of the money view ishighly stylized, it is consistent witha number of alternative modelsbeyond the texihank IS-IM model(see, for example, Hubbard, 1994),including dynamic-equilibrium cash-in-advance madels (far example,Rotemberg, 1984; and Christianoand fichenbaum, 1992).

‘Far an empidral descripton of thistransmission mechanism in the con-text of the Federal Reserve’s forecast-ing model, see Mauskapf (1990).

6 See, for example, “limited partici

ponien’ models as in Lucas (1990)and Chrisfana and lichenbaum(1992).

See, for example, analyses of

innentory irvestnnent ir Kashyap,

Stein and Wilcox (1993) andGerfer and Gilchist (1993). Seealso the review of empirical studiesof business fined investment inChirinko (1993) and Cammins,Hassett and Hubbard (1994).

a This current fashion actually has

long pedigree in macroecenemics,with iniporlont cantrihutions by Fisher(1933), Gudey and Show (1955,

960), Minsky (1964, 19/5) andWalnilawer 11980), Some econa-metric ferecosfing medels have alsofocosed on financial factors in prop-agafon mechaaisms (see, forexample, the description for the DRImodel in Eckstein and Sinai, 19861.

Cogan (19/2) provides an ernpiicolanalysis of money and bank lending

views. Mr early contributor tu thecontemporary credit view literatureis Bemanke (1983).

bonds or stocks should have no effect on realoutcomes; that is, the financial system ismerely a veil.’

To keep the story simple, suppose thatthere are two assets—money and bonds,4

In a monetary contraction, the central bankreduces reserves, hmiting the banking system’sability to sell deposits. Depositors (house-holds) must then hold more bonds and lessmoney in their portfolios. If prices do notinstantaneously adjust to changes in the moneysupply the fall in household money holdingsrepresents adechne in real money balances, Torestore equilibrium, the real interest rate onbonds increases, raising the user cost of capitalfor a range of planned investment activities,and interest-sensitive spending falls.’

While the money view is widely acceptedas the benchmark or “textbook” model foranalyzing effects of monetary policy on eco-nomic activity, it relies on four key assump-tions: (1) The central bank must control thesupply of “outside money,” for which there areimperfect substitutes; (2) the central bank canaffect real as well as nominal short-tenn interestrates (that is, prices do not adjust instanta-neously); (3) policy-induced changes in realshort-term interest rates affect longer-terminterest rates that influence household andbusiness spending decisions; and (4) plausiblechanges in interest-sensitive spending inresponse to a monetary policy innovationmatch reasonably well with observed outputresponses to such innovations.

In this stylized view, monetary policy isrepresented by a change in the nominal supplyof outside money Of course, the quantityof much of the monetary base is likely to beendogenous.° Nonetheless, legal restrictions(for example, reserve requirements) maycompel agents to use the outside asset forsome transactions, In practice, the centralbank’s influence over nominal short-ternninterest rates (for example, the federal fundsrate in the United States) is uncontroversial,There is also evidence that the real federalfunds rate responds to a shif’t in policy (see,for example, Bernanke and Blinder, 1992).

Turning to the other assumptions,that long-term rates used in many savingand investment decisions should increase ordecrease predictably in response to a change

in short-term rates is not obvious a prioribased on conventional models of the termstructure, Empirical studies, however, havedocumented a significant, positive relationshipbetween changes in the (nominal) federalfunds rate and the 10-year Treasury bond rate

(see, for example, Cohen andWenninger, 1993;and Escrella and Hardouvelis, 1990). Finallyalthough many components of aggregate

demand are arguably interest-sensitive (suchas consumer durables, housing, business fixedinvestment, and inventory investment), output

responses to monetary innovations are largerelative to the generally small estimated effectsof user costs of capital on investment.’

I shall characterize the money view asfocusing on aggregate, as opposed to distribu-tional, consequences of policy actions. In

this view, higher default-risk-free rates ofinterest following a monetary contractiondepress desired investment by firms and

households. While desired investment falls,the reduction in business and householdcapital falls on the least productive projects.

Such a view offers no analysis of distribu-tional, or cross-sectional, responses to policyactions, nor of aggregate implications of this

heterogeneity I review these points not tosuggest that standard interest rate approachesto the monetary transmission mechanism areincorrect, but to suggest strongly that one

ought to expect that they are incomplete.

HOW REASONABLE IS THECREDIT VIEW?

The search for a transmission mechanismbroader than that just described reflects twoconcerns, one “macro” and one “micro,” Themacro concern, mentioned earlier, is thatcyclical movements in aggregate demand—

particularly business fixed investment andinventory investment—appear too large tobe explained by monetary policy actions that

have not generally led to large changes inreal interest rates, This has pushed somemacroeconomists to identify financial factors

in propagating relatively small shocks, fac-tors that correspond to accelerator modelsthat explain investment data relatively well,0

Indeed, I use the term “financial accelerator”(put forth by Bernanke, Gender and Gilchrist,

FEDERAL RESERVE SANK OF ST. LOUIS

64

MAY/JUNE 1995

forthcoming) to refer to the magnification ofinitial shocks by financial market conditions,

The micro concern relates to the emer-

gence of a growing literature studying infor-mational imperfections in insurance and credit

markers. In this line of inquiry problems ofasynimetric information between borrowersand lenders lead to agap between the costof external finance and internal finance.The notion of costly external finance standsin contrast to the more complete-marketsapproach underlying the conventional interest

rate channels, which does not consider hnksbetween real and financial decisions,r

Although a review of this literature isbeyond the scope of this article, I want tomention three common empirical implica-tions that have emerged from models of thefinancial accelerator’0 The first, which Ijustnoted, is that uncollaceralized external financeis more expensive than internal finance,Second, thespread between the cost of external

and internal finance varies inversely with theborrower’s net worth—internal funds and

collateralizable resources—relative to theamount of funds required. Third, an adverseshock to a borrower’s net worth increases the

cost of external finance and decreases theability of the borrower to implement invest-ment, employment and production plans.

This channel provides the financial accelerator,magnifying an initial shock to net worth,(See, for example: Fazzari, Hubbard andPetersen, 1988; Gertler and Hubbard, 1988;

Cantor, 1990; Hoshi, Kashyap and Scharfstein,1991; Calomiris and Hubbard, forthcoming;Hubbard and Kashyap, 1992; Oliner and

Rudebusch, 1992; Fazzari and Petersen, 1993;Hubbard, Kashyap and Whited, forthcoming;Bond and Meghir, 1994; Cummins, Hassettand Hubbard, 1994; Carpenter, Fazzari andPetersen, 1994; and Sharpe, 1994.)” Linksbetween internal net worth and broadly definedinvestment (holding investment opportunitiesconstant) have been corroborated in a numberof empirical studies.rr

Let me now extend this argument toinclude a channel for monetary policy” In

the money vie~policy actions affect theoverall level of real interest rates and inter-

est-sensitive spending. The crux of modelsof information-related financial frictions is

a gapbetween the cost of external and internal

finance for many borrowers, In this context,the credit view offers channels through whichmonetary policy (open market operations orregulatory actions) can affect this gap. Thatis, the credit view encompasses distributionalconsequences of policy actions, because the

costs of finance respond differently for differenttypes of borrowers. Two such channels havebeen discussed in earlier work: (1) financial

constraints on borrowers and (2) the exis-tence of bank-dependent borrowers,

Any story describing a credit channel formonetary policy must have as its foundation

the idea that some borrowers face high costsof external finance, In addition, models of

a financial accelerator argue that the spreadbetween the cost of external and internal fundsvaries inversely with the borrowers’ net worth,

It is this role of net worth which offers achannel through which policy-inducedchanges in interest rates affect borrowers’net worth (see, for example, Gertler and

Hubbard, 1988). Intuitively increases in thereal interest rate in response to a monetary

contraction increase borrowers’ debt-serviceburdens and reduce the present value ofcollateralizable net worth, thereby increasing

the marginal cost of external finance andreducing firms’ ability to carry out desiredinvestment and employment programs, This

approach offers a credit channel even if openmarket operations have mo direct quantityeffect on banks’ ability to lend, Moreover,

this approach implies that spending by low-net-worth firms is likely to fall significantlyfollowing a monetary contraction (to theextent that the contraction reduces borrowers’net worth).

t~5~.flceOt H~cyncc4f’pndànrt

The second channel stresses that someborrowers depend upon banks for externalfunds, and that policy actions can have a directimpact on the supply of loans. When banksare subject to reserve requirements on liabihties,a monetary contraction drains reserves,

Potential effects of advene selec-

tion problems on market allocationhave been addressed in importantpapers by Akedof (1970) andRothschild and Stglitz (1976), andhave been applied to moo marketsby laffee and Russell (1976) andStiglitz and Weiss (1981), and toequity markets by Myers andWailaf (1984). Research an piacipal-agent problems in finance hasfollowed the contribotion of larsonand Meckling (1976). Gertler11988), Bernanke (1993) andKing and Lenine (1993) pmvidereviews of related models of infor-mational imperfectiens in capitalmarkets.

“See also the review in Bernanke,Gender and Gilchrist (forthcoming).These implications ore consistentwith a wide class of models, in rInd-ing those of Townsend (1979),Blinder and Stiglitz (1983),Farmer (1985), Williamsen(1987), Bereunke and Gender(1989, 1990), Calominis andHubbard (1990), Sharpe (19901,Hart and Meow (19911, Kiyotakiand Moore (1993), Gender(1992), Greerwald and Stiglitz(1988, 19931 and Iomoat(1993).

‘‘Far households, Mishkin 119/7,1978) and Leldes (19891 pmnideevidence of effects of hauseholdbalance sheet conditions on con-sumer enpenditures.

The appendin presents a simple

model that illustrates these predic-tions.

“For broader desciptiens of creditview arguments, see Bemanke(1993), friedman and Kottner(1993), Gender (1993), Gerferand Gilchrist (1993) and Kashyapand Stein (1994). An early eupo-sition of a role for cmdit availabilityappears in fooso (1951).

FEDERAL RESERVE BANK OF ST. LOUIS

65

o FYI F~MAY/JUNE 1995

“Models of equlibrium creditrationing onder adverse selection(for enample, Stiglitz and Weiss,

1981) offer another mechanismthrough which on increase in thelevel of defanlt-risk-free real interestmtes reduces loon sapply. Creditrationing is rot reqaired for thebonk-dependent-barrower channelno be operative. Insteod, what isrequired is that loans tu these bor-rawers are an imperfect substitute

for other assets and that tine bar-roman lack alternative sources offinance.

° Colomiris and Kahn (1991) offermodel of demandable debt tuhaooce bank lending.

‘‘A substantial body of empitical evidence sopports the idea that banksoffer special services in the lendingprocess. For example, James(19871 and lummer andMcCaneell (1989) find that theannouncement of a bank loon, oilelse equal, mises the share ptice ofthe barmwing firm, likely reflectingthe information content ol thebank’s assessment, In a similarspirit, loma (1985) and James(1987) find Chat honks’ borramers,

rather than banks’ depositors, bearthe incidence ol reserve reqaire-ments (indicating that borrowersmast not hove easy access tu othersources of funds). Petersen andRalan (1994) show that smallbusinesses tend to rely on lacalhanks for enternal funds.

‘‘See, far euomple, the discussion inPetersen and Roman mi 994).

re Omens and Schreft (1992) discuss

the ident’dication of ‘credit crunch-es.’ See also the description inHubbard 11994),

possibly decreasing banks’ ability to lend. As

a result, credit allocated to bank-dependentborrowers may fall, causing these borrowersto curtail their spending. In the IS-LMframework of Bernanke and Blinder (1988),both the IS and LM curves shift to the left inresponse to a monetary contraction, Alterna-tively an adverse shock to banks’ capital coulddecrease both banks’ lending and thespendingby bank-dependent borrowers. Such banklending channels magnify the decline in out-put as a result of the monetary contraction,and the effect of the contraction on the realinterest rate is muted. This basic story raisesthree questions, relating to: (1) why certainborrowers may be bank-dependent (that is,unable to accessopen market credit or borrowfrom nonbank financial intermediaries orother sources), (2) whether exogenous changesin banks’ ability to lend can be identified, and(3) (for the analysis of open market operations)whether banks have access to sources of fundsnot subject to reserve requirements.

The first question is addressed, thoughnot necessarily resolved, by the theoreticalliterature on the development of financialintermediaries”, In much of this research(see especially Diamond, 1984; and Boydand Prescott, 1986), intermediaries offerlow-cost means of monitoring some classesof borrowers. Because of informational fric-

tions, non-monitored finance entails dead-weight spending resources on monitoring. Afree-rider problem emerges, however, in pub-lic markets with a large number of creditors.The problem is mitigated by having a finan-cial intermediary hold the loans and act as adelegated monitor, Potential agency problemsat the intermediary level are reduced by havingthe intermediary hold a diversified loan port-folio financed principally by publicly issueddebt.” This line of research argues rigorouslythat borrowers for whom monitoring costsare significant will be dependent upon inter-mediaries for external finance,” and that costsof switching lenders will be high.” It does not,however, necessarily argue for bank dependence(for example, finance companies are interme-diaries financed by non-deposit debt).

Second, even if one accepts the premisethat some borrowers are bank-dependent inthe sense described earlier, one must identify

exogenous changes in banks’ ability to lend.Four such changes have been examined inprevious research. The first focuses on therole played by banking panics, in which

depositors’ flight to quality—converting hankdeposits to currency or government debt—reduces banks’ ability to lend (for empiricalevidence, see Bernanke, 1983, and BernankeandJames, 1991, for the 1930s and Calomiris

and Hubbard, 1989, for the National Bankingperiod).

A second argument emphasizes regulatory

actions, such as that under binding RegulationQ ceilings in the United States (see, forexample, Schreft, 1990; Kashyap and Stein,1994; and Romer and Romer, 1993) and reg-ulation of capital adequacy (see, for example,Bernanke and Lown, 1992; and Peek andRosengren, l992).,~Empirical evidence for

this channel is quite strong. Third, Bizer(1993) suggests that increased regulatoryscrutiny decreased banks’ willingness tolend in the early 1990s, all else equal.

The fourth argument stresses exogenous

changes in bank reserves as a result of shiftsin monetary policy In principle, such a shiftin monetary policy could be identified with adiscrete change in the federal funds rate in theaftermath of a dynamic open market operationor with achange in reserve requirements.Because the effects on reserves of changes in

reserve requirements are generally offset byopen market operations, bank-lending-chan-nel stories are generally cast in terms of open

market operations.An illustration of the gap between models

and practice surfaces in addressing the thirdquestion of the ease with which banks can raisefunds from non-deposit sources (for example,CDs), when the Fed decreases reserves. Romerand Romer (1990) have pointed out, forexample, that if banks see deposits and CDsas perfect substitutes, the link between openmarket operations and the supply of credit tobank-dependent borrowers is broken. Banks

are unlikely however, to face a perfectly elasticsupply schedule for CDs at the prevailing CDinterest rate, Since large-denomination CDsare not insured at the margin by federal depositinsurance, prospective lenders must ascertainthe quality of the issuing bank’s portfolio.Given banks’ private information about at

FEDERAL RESERVE BANK OF St. LOUIS

66

DFYIE~MAY/JUNE 3005

least a portion of their loan portfolio, adverseselection problems will increase the marginalcost of external finance as more funds areraised (see, for example, Myers and Majluf,1984; and Lucas and McDonald, 1991). Inaddition, as long as some banks face constraintson issuing CDs and those banks lead tobank-dependent borrowers, a bank lendingchannel will be operative.

While the foregoing discussion centerson open market operations, regulatory actionsby the central bank—credit controls, forexample—represent another way in whichmonetary policy can have real effects throughinfluencing the spending decisions of bank-dependent borrowers. Here the effects arelikely to be more pronounced than for thecase of open market operations, since thequestion of the cost of non-deposit sourcesof funds is no longer central, and the effec-tiveness of such regulatory actions dependsonly on the existence of bank-dependentborrowers.

flQffsfl~ly~tsnCrr?it°.MOLWLC TQ‘2 °2~2er‘2’2inø°i’24srts’2i,.fd’2c2,22

Both the financial-constraints-on-bor-rowers and bank-lending-channel mechanismsimply significant cross-sectional differencesin firms’ shadow cost of finance and in theresponse of that cost to policy-induced changesin interest rates. Accordingly empiricalresearchers have attempted to test thesecross-sectional implications. As I examinethis literature, I explore how Modigliani-Millerviolations for nonfinancial borrowers, financialintermediaries or both offer channels formonetary policy beyond effects on interestrates. The appendix frames this discussionusing a simple model; an intuitive presenta-tion follows.

EMPIIU.CAL !flSEARt*I, oN

THE CREDIT VIEW

Studies Using Aqqregote Da.tu

The microeconomic underpinnings ofboth financial accelerator models and thecredit view of monetary policy hinge oncertain groups of borrowers (perhaps including

banks or other financial intermediaries) facingincomplete financial markets, Examininglinks between the volume of credit and eco-nomic activity in aggregate data (with an eyetoward studying the role played by bank-dependent borrowers) requires great care.Simply finding that credit measures lead outputin aggregate time-series data is also consistentwith a class of models in which credit is passive,responding to finance expected future output(as in King and Plosser, 1984), Consider thecase of a monetary contraction, for example.The effect of the contraction on interest ratescould depress desired consutnption andinvestment spending, reducing the demandfor loans.

In a clever paper that has stimulated anumber of empirical studies, Kashyap, Steinand Wilcox (1993)—henceforth, KSW—examine relative fluctuations in the volumeof bank loans and a close open market sub-stitute, issuance of commercial paper. In theKSW experiment, upward or downwardshifts in both bank lending and commercialpaper issuance likely reflect changes in thedemand for credit, However, a fall in banklending while commercial paper issuance is

rising might suggest that bank loan supply iscontracting. To consider this potential co-movement, KSW focus on changes over timein the mix between bank loans and commer-cial paper (defined as bank loans divided bythe sum of bank loans and commercial paper).They find that, in response to increases inthe federal funds rate (or, less continuouslyat the times of the contractionary policyshifts identified by Romer and Romer, 1989).the volume of commercial paper issues rises,while bank loans gradually decline, Theyalso find that policy-induced changes in themix have independent predictive power forinventory and fixed investment, holding con-

stant other determinants.”The aggregate story told by KSW masks

significant firm-level heterogeneity however.The burden of a decline in bank loans fol-lowing a monetary contraction is borne bysmaller firms (see Gertler and Gilchrist, 1994).”Moreover, the evidence in Oliner andRudebusch (1993) indicates that once tradecredit is incorporated in the definition ofsmall firms’ debt and once firm size is held

Oliner and fadebasch (1993) and

Friedman and Kattaner (1993) honedisputed the KSW interpretation ofthe mia as measadag a substitutionbetween bank loans and commer-cial paper. They argue than, during

recession, shifts in the miu areerplained by an increase in conmerciol paper issuance rather thanby a decrease in bank loans,

“Morgan (1993) finds a similar

result in en 000lysis of loan corn-mitments, After an episode ofmonetary contraction, firms withontloan commitments receive a small-er share of bank loans.

FEDERAL RESERVE BANK OF ST. LOUIS

67

OFYI[~MAY/JUNE 1995

1’ Toward this end, more direct com-

parisons of borrowing by bark-dependent and nonboak.dependentbarrowers hove been offered.Using firm-lenel data, Kashyop,tamont end Steir (1994)—henceforth KLS—follow theFozzari, Hubbarri and Petersen11988) approach of classifyinggroups of firms as a priori finance-constrained hr this case, bank-dependent) or not. Ir particular,they stody inventory investment bypeblicly traded firms mitlr and with-out bond rotirgs, as a prony forbank dependence. focusing on the1982 recession (as an indirectwears of identifying a period fol-lowing a tight money episode),they find that inventury innestnentby nan-rated firms was influenced,all else equal, by the firms’ owncash holdings, on effect not presentfor the innertory investment byrated firms, tn subsequent boomyears (which KLS identify with oneasy money episode), they lied lit-in effect of cash holdings on inven-tory investment for either non-rotedor rated companies. These patternslead KtS to conclude that a banklending channel was operative inresponse to the monetary contrac-tion. Hawener, the KLS resolts areconsistent with a more generalmodel in which low-nat-worth firmsfoce more costly erternalfinance indowntnrns.

constant, monetary policy changes do notalter the mix.

It also does not appear that hank-depen-dent borrowers switch to the commercialpaper market following a monetary contrac-tion. Instead, the increase in commercialpaper issuance reflects borrowing by largefirms with easy access to the commercialpaper market, possibly to smooth fluctua-tions in their flow of funds when earningsdecline (Friedman and Kuctner, 1993) or tofinance loans to smaller firms (Calomiris,Himmelherg and Wachtel, forthcoming).

More convincing empirical tests focus

on the cross-sectional implication of theunderlying theories—namely that credit-market imperfections affect investment,employment or production decisions of someborrowers more than others. At one level,existing cross-sectional empirical studies havebeen successful: There is asubstantial bodyof empirical evidencedocumenting that proxiesfor borrowers’ net worth affect investmentmore for low-nec-worth borrowers than forhigh-net-worth borrowers (holding constantinvesnnent opportunities). This suggests

that, to the extent that monetary policy canaffect borrowers’ net worth, pure interestrate effects of open market operations willbe magnified.

The second body of empirical analysis ofinformation-related imperfections focuses onthe effects of monetary policy on borrowers’balance sheets. Gercler and Hubbard (8988)conclude that, all else equal. internal fundshave a greater effect on investment by non-dividend-paying firms during recessions.The evidence of Gertler and Gilchrist (1994)

is particularly compelling here. Analyzingthe behavior of manufacturing firms summa-rized in the Quarterly Financial Reports data,

Gertler and Gilchrist consider differences insmall and large firms’ responses to tightmoney (as measured by federal funds rateinnovations or the dates identified by Romerand Romer, 1989). In particular, small firms’

sales, inventories and short-term debtdecline relative to those for large firms over a

two-year period following a monetary tight-ening, results consistent with the financialaccelerator approach. They also demonstratethat the effects of shifts in monetary policyon the small-firm variables are sharper inperiods when the small-firm sector as a wholeis growing more slowly also consistent withthe financial accelerator approach. Finallythey show that the ratio of cash flow to interestexpense (a measure of debt-service capacity)is associated positively with inventory accu-mulation for small, but not for large, manu-facturing firms.

The Gertler and Gilchrist results, whichare very much in the spirit of the earliercross—sectional tests of financial acceleratormodels, have been borne out for studies offixed investment by Oliner and Rudebusch

(1994) and for inventory investment byKashyap, Lamont and Stein (1994),” In addi-tion, Ramey (1993) shows that, for forecast-ing purposes, the ratio of the sales growth ofsmall firms to that for large firms offers sig-nificant information about future GDP.

Finally using the firm-level dataunderlying the aggregates summarized inthe Quarterly Financial Reports, Bernanke,Gertler and Gilchrisc (forthcoming) analyzethe differences in sales and inventories betweenlarge and small manufacturing firms by

two-digit industry They find that fluctua-tions in the large firm-small firm differencesareroughly the same size as fluctuations in

the corresponding aggregate fluctuations forthe manufacturing sector. Because small firms’sales (as they define small firms) compriseabout one-third of the sales of the manufac-turing sector, roughly one-third of cyclicalfluctuations in manufacturing sales can be

explained by large firm-small firm differences,

M.S’MSSSIflST ‘2220. Durtk t.uncnnq

While the principal empirical predictions

of the financial accelerator approach have beencorroborated in micro-data studies and low-nec-worth firms appear to respond differen-

tially to monetary contractions, the questionof the role of banks remains. I consider this

question below in three steps.First, is there evidence of significant

FEDERAL RESERVE BANK OF ST. LOUIS

68

II FYI F~MAY/JUJIS n99S

departures from Modigliani and Miller’sresults for certain groups of banks in thesense that have been identified for firms?Second, is there evidence that small- or low-net-worth firms are more likely to he the

loan customers of such banks? Finally dolow-net-worth firms have limited opportuni-ties to substitute credit from unconstrained

financial institutions when cut off by con-strained financial institutions?

0

Kashyap and Stein (1994) apply theintuition of the models of effects of internalnet worth on investment decisions by nonfi-nancial firms to study financing and lendingdecisions by banks. This is an importantline of inquiry in the hank lending channel

research agenda, because ft addresses the easewith which banks can alter their financingmix in response to a change in bank reservesand the effect of changes in the financingmix on the volume of bank lending. Just asearlier studies focused on cross-sectional dif-ferences in financing and real decisions ofnonfinancial firms of different size, Kashyapand Stein analyze cross-sectional differencesin financing and lending decisions of banksof different size. To do this, they use data

drawn from the quarterly “Call Reports”collected by the Federal Reserve,

Kashyap and Stein construct asset sizegroupings for large banks (those in the 99thpercentile) and small banks (defined as thoseat or below the 75th, 90th, 95th or 98thpercentiles). They first show that contrac-tionary monetary policy (nneasured by anincrease in the federal funds rate) leads toa similar reduction in the growth rate of nom-inal core deposits for all hank size classes.They find significant heterogeneity acrosshank size classes, however, in the responseof the voluene of lending to a change in mon-etary policy In particular, a monetary corn-traction leads to an increase in lending int.he short run by very large banks. This is incontrast to a decline in lending in the shortrun by smaller banks. These do not simplyreflect differences in the type of loans madeby large and small banks. A sinnilar pattern

emerges when loans are disaggregated toinclude just commercial and industrial loans.

One possible explanation for the Kashyapand Stein pattern is that a monetary contrac-tion weakens the balance sheet positions of

small firms relative to large firms. If smallfirms tend to be the customers of small banksand large firms tend to be the custoaners oflarge banks, a fall in loan demand (by smallborrowers) for small banks could be consis-tent with the differential lending responsesnoted by Kashyap and Stein, To examinethis possibility Kashyap and Stein analyzewhether small banks increase their holdings

of securities relative to large banks during amonetary contraction, They actually findchat small banks’ securities holdings are less

sensitive to monetary policy than large banks’securities holdings, though the difference inthe responses is not statistically significant.

The use of hank size as a measure to gen-erate cross-sectional differences does not corre-spond precisely to the underlying theoreticalmodels, which stress the importance of networth. In this context, bank capital may hea better proxy Peek and Rosengren (forth-

coming) analyze the lending behavior of NewEngland banks over the 1990-9 1 recession.Their results indicate that the loans of wellcapitalized banks fell by less than the loansof poorly capitalized banks.” Hence, as withthe Kashyap and Stein findings, their evidencesuggests there are effects of infonnationalimperfections in financial markets on thebalance sheets of intermediaries as wellas horrowers.

tanrnvnu ti~rrcJvePt-cnr,d ~

The last Iwo questions relate to thematching of borrowers and lenders. Theformer asks whether the firms identified byempirical researchers as finance-constrainedare the loan customers of the constrained(small) banks such as those identified byKashyap and Stein. This line of inquiry

requires an examination of data on individ-ual loan transactions, with information oncharacteristics of the borrower, lender andlending terms.” One could establish whetherconstrained firms are the customers of con-strained banks and whether such firms

“Using dotu on commercial banksnatonwrle onerthe 1979-92 period,Berger and Udell 119941 bond lit-tle enidence that the introduction oftisk-irased capitul requirements per Se

affected credit ollocotior. Hancock,Laing and Wilcan 119941 also rseqvortedy data on iudividuul honk’sportfolios to estimate the respan-sireress of portfolio composition to

changes in capitol requirements.They find that ‘copitul shortfall’institutions reduced their Cal loorsresponse by larger total amounts,all else equal, titer ‘copitol surplus’institutions.

‘2 Aril Koshyop, Darius Polio and lore

cunertfy ergoged ir such an anolyis.

FEDERAL RESERVE BANK OF ST. LOUIS

69

DFYIF~MAY/JUNE 1995

‘2 Another possibil’ty is that the

weakened balance sheet positionsof many borrowers precipituteso ‘flight to quality’ by lendersgenerally, increasing the demandfor commercial paper issues oflarge firms,

‘5The dates of monetary policycontractions saggested by Rowerand Rawer (19891 have generatedsignificant controversy. Shapiro(1994) argues, for erawple,that empirical evidencefarors thehypothesis thnt several Rower datesare predictable asing weaseresof unemployment and infanionas deterninants of actions by theFederal Open Market Committee;see also the discussion in Cecchett

(1995). Hoover and Perez (1994)offer a number of criticisms of thelowers’ appraach.

‘°Such relationships are lypically

estimated as:

Yht)”a+ bY(.t— il

— cH(t— il—dEbt—i),where Y is the percentage changein real GOP miatine to potentialGOP, His the percentuge chatge inthe high-employmentfederal bad-get surplus, F is the change in thefederol funds rate, tis the currerottime period, and idenotus lags.See, for enample, Hirtle and Kelleher(1990), Pen-y and Schultze (1992)and Cohen and Wenninger (19931.

“Cover (1992) finds still strongerevidence of asymmetric effectswhen monetary aggregates areused as the policy indicator insteadof the federal funds rnte,

switch from constrained banks to uncon-strained ones during episodes of monetarycontractions. Theories emphasizing the impor-tance of ongoing borrower-lender relationshipsimply that such switches are costly and unhkely

If true, part of themonetary transmissionmechanism takes place through reductionsin loan supply by constrained banks.

The latter of the two questions suggests

the need to study a broader class of lendersthan banks, Ifborrowers from constrained

banks can switch at low cost to nonbanklenders following a monetary contraction,the narrow bank credit channel of monetary

policy is frustrated. In this vein, Calomiris,Flimmelberg and Wachtel (forthcoming) ana-lyze finn-level data on commercial paper

issuance and argue that large, high-qualitycommercial paper-issuing flrnms increasepaper borrowings during downturns to

finance loans to smaller firms.’2 They notethat accounts receivable rise for paper-issu-ing firms, supporting the notion chat these

firms may serve as trade credit intermedi-aries for smaller firms in some periods.From the standpoint of the bank lendingchannel, it is important to establish whathappens to the costs and terms imposed bythese intermediaries. If, on the one hand,such terms are no more costly than bankintermediary finance, then the switch of bor-rowers from being bank customers to beingtrade credit customers entails very limitedmacroeconomic effects. On the other hand,if large, paper-issuing firms accept theirintermediary role reluctantly very costlytrade credit may exacerbate a downturnby raising the cost of funds for constrainedfirms. More empirical investigation oftrade credit terms is needed to resolvethis question.

More empirical research is also neededto assess the validity of the basic moneyview A central problem is that, while mostempirical studies focus on monetary aggre-gates such as M2, the theoretical descriptionoffered in the first section suggests an emphasison outside money and, importantly on com-

ponents of outside money over which thecentral bank can exercise exogenous control.First identifying exogenous changes in mone-

tamy policy is difficult.’5 Recent research byBernanke and Blinder (1992) and Christiano,Eichenbaum and Evans (forthcoming) offerspromising strategies for studying the effects

of monetary policy shocks.In addition, recent analyses of policy-

reduced-form models document a significant,negative relationship in quarterly data betweenthe percentage change in real GDP relative to

potential GDP and the change in the federalfunds rate.mu Such studies must first confrontthe possibility that the measured interest

sensitivity of output reflects links betweeninterest race and net worth changes for certaingroups of borrowers/spenders. A secondissue, noted by Morgan (1993) and Cohen

and Wenninger (1993), is that quarterlyresiduals from estimated policy-reduced-form

equations display large negative errors duringrecessions, suggesting the possibility of anasymmetric response of economic activity toincreases or decreases in the federal funds~ Finally more theoretical and empiricalresearch is needed to examine links between

changes in short-term real interest rates (whichare significantly influenced by pohcy actions)and changes in long-term real interest races(which affect firms’ cost of capital).

fl~-r~~ryq Intlsi:fl ii1yiyiifliui,,,t’aW’tt0ns”*~nC’ i.anns 222222

~ .~22o -

RgMJrtYThis survey argues that the terms money

view and credit view are not always well-defined in theoretical and empirical debatesover the transmission mechanism of mone-

tary policy Recent models of informationand incentive problems in financial marketssuggest the usefulness of decomposing thetransmission mechanism into two parts: onerelated to effects of policy-induced changeson the overall level of real costs of funds; andone related to magnification (or financialaccelerator effects) stemming from impactsof policy actions on the financial positions ofborrowers and/or intermediaries.

Two observations emerge clearly fromthe literature, First, the spending decisions

FEDERAL RESEEVE SANK OF ST. LOUIS

70

llF~IFWMAY/JUNE 1995

of a significant group of borrowers are influ-enced by their balance sheet condition in the

ways described by financial accelerator models.Second, even in the presence of more sophis-

ticated financial arrangements, there are stillinformation costs of screening, evaluationand monitoring in the credit process, impart-ing a special role for intermediaries (be theybanks or other lenders) with cost advantagesin performing these tasks.’°

The first observation suggests thatfinancial factors are likely to continue to playa role in business fluctuations. The second

suggests that regulatory policies affectinginformation-specializing intermediaries arelikely to affect the cost of credit for at least

some borrowers, In part because of interestin alternative views of the monetary trans-mission mechanism and in part because ofconcern over the effects of institutional changein the financial system, academics and poli-cymakers are analyzing whether the scopefor monetary policy to affect real outcomesis becoming narrower. Both observationsnoted above are consistent with a heightenedrole for monetary policy in affecting realdecisions of firms with weak balance sheetpositions. Developing ways to incorporateborrower heterogeneity in both economicmodels of money and credit and in forecastingis an important, practical task for economicmodelers and policymakers.

Whether the simplest bank lendingchannel—that a fall in banks’ reserves fol-lowing contractionary open market operationsdecreases both banks’ ability to lend and bor-rowers’ ability to spend—is operative is notclear, however, More micro-evidence at thelevel of individual borrower-lender transac-tions is needed to resolve this question. Atthe same time, proponents of the simplestcharacterization of an interest rate channelmust address both the cross-sectional hetero-geneity in firms’ response to monetary policyand the extent to which observed interest rateeffects on output reflect differentially largeeffects of policy on certain classes of borrowers,

Akerlob, George. ‘The Market for lemons: Quality Uncertuinty ond theMarket Mechanism,” Quarterly Journal of Economics (August 1970),pp. 488-500.

Berger, Allen N., and Gregory F. Udell. ‘Did Risk-Based Capital Allocate

Bank Credit and Cause a ‘Credit Crunch’ in the U.S.?” (port two)Journal atMoney, Credit and Banking (August 1994), pp. 585-628.

Bernanke, Ben S. ‘Credit in the Macroecanatny,” Federal Reserve Bankof New York Quarterly Review (spring 1993), pp. 50-70,

__________ ‘Nonmonetary Effects of the Financial Crisis in thePropagation of the Great Depression,” The American Economic Review(June 1983), pp. 257-76.

_________ and Alan S. Blinder. ‘The Federal Funds Rate and theChannels of Monetary Transmission,’ The American Economic Review(September 1992), pp. 901-21.

__________ and __________- ‘Crecit, Money, and AggregateDemand,” The American Economic Review (May 1988), pp. 435-39.

___________ and Mark Gerfier. “Agency Cost, Net Worth, and BusinessFluctuations,” The American Economk Review (March 1989), pp. 14-31.

______ and ______. “Financial Fmgility and EconomicPerformance,” Quarterly Journal atEconomics (February 1990),pp.87-il4.

_____________________ and Simon Gilchrist. “The FinancialAccelerator and the Flight to Quality,” Review of Economics andStatistics (forthcoming).

__________ and Harold James. ‘The Gold Standard, Deflation, andFinancial Crisis in the Great Depression: An International Comparison,”in R. Glenn Hubbard, ed., Financial Markets and Financial Crises.University of Chicago Press, 1991.

_________ and Cara S. lawn, ‘The Credit Crunch.” Braakings Papersan EcanamicActivity(1992:2), pp. 205-39.

Bizer, Dovid S. ‘Regulatory Discretion and the Credit Cwnch,” workingpaper (April 1993), U.S. Securities and Exchange Commission.

Blinder, Alan S., and Joseph F. Stiglitz. ‘Money, Credit Constraints, andEconomic Activity,” TheAmerican Economic Review (May 1983), pp.

297-302.

Bond, Stephen, and Castos Meghir. ‘Dynamic Investment Models andthe Firm’s Financial Policy,” Review of Econamk Studies (April 1994),pp. 197-222.

Boyd, John H., and Mark Gertler, “Are Banks Dead? Or, Are the ReportsGreatly Exaggerated?’ working paper (Moy 1994), Fedeml ReserveBank of Minneapolis.

___________ and Edward Prescott. ‘Financial lntermediory-Coalitions,’Journal of Economic Theory (April 1986), pp. 211-32.

_________ and Bruce D. Smith. “Capitul Market Imperfections in aMonetury Gmwth Model,” Working Paper Na. 533 (August 1994),Federal Reserve Bank of Minneapolis.

Brainord, William C., and James Tabin. ‘Financial Intermediaries andthe Effectiveness of Monetary Controls,’ The American EconomicReview (May 1963), pp. 383-400.

Cagan, Phillip. The Channels of Monetary Effects an Interest Rates.Columbia University Press, 1972.

on For recent analyses of the future of

banking, see Gortan and Pennacchi(1993), Edwards (1993) andBoyd and Gertler (1994).Thornton (1994) discusses likelyeffects of recent financial innova-tions on the bank lending chunnel,

FRDERAL RESERVE SANK OF ST. LOUIS

?1

DF\’IF~MAY/JUNE 1995

Calomiris, Charles W., Charles P Himmelberg and Paul Wachtul. Eckstein, Otto, and Allen Sinai, “The Mechanisms of the Business Cycle‘Commercial Paper and Corparote Finance: A Microeconamic in the Postwar Era,” in Robert J. Gordon, ed., The American BusinessPerspective,” Carnegie-Rochester Conference Series on Public Policy Cyde: Continuity and Change. University of Chicago Pmss, 1986.

(farthcomtng) - Edwards, Franklin R. ‘Financial Markets in Transitiorn—Or the Decline

_________ and R. Glenn Hubbard. ‘Firm Heterogeneity, Inturnol of Commercial Banking,” in Changing Capital Markets: Implications forFinance, and ‘Credit Ratianing,’” Economic Journal (March 1990), Monetary Policy. Federal Reserve Bank of Kansas City, 1993.

~ 90-1 04. Estrella, Arturn, and Gikas Hordouvelis. “Possible Roles of the Yield

_________ and _________. ‘Price Flexibility, Credit Availability, and Curve in Monetory Policy,” in Intermediate Targets and Indicators ofEconomic Fluctuations: Evidence from the United States, 1894-1909,’ Monetary Policy: A Critical Survey, 1990, Federal Reserve Bank ofQuarterlyJournal of Economics (Angnst 1989), pp. 429-52. New York, 1990, pp. 339-62.

___________ and ___________. ‘Tax Policy, Internal Finance, and Fama, Eugene F. “Banking in the Theory of Finance,” Journal ofInvestment: Enidence from the Undistributed Profits Tax of 1936-37,” Monetary Economics (January 1980), pp. 39-57.Journal of Business (forthcoming). Fama, Eugene F. “What’s Different About Banks?’ Journal of Monetary

_________ and Charles Kahn. ‘The Role af Demandable Debt in Economics (June 1985), pp. 29-39.Structuring Optimal Banking Arrangements,” TheAmerican Economic Farmer Roger E.A. “Implicit Contracts with Asymmetric Information andReview (June1991), pp. 497-513. Bankruptcy: The Effect of Interest Rates an Layoffs,” Review of

Cantor, Richard. “Effects of levemge on Corporate Intestment and Economic Studies (July 1985), pp. 427-42,Hiring Decisions.” Federal Reserve Bank of New Ynrk Quortedy Fozzoti, Steven M., and Bruce C. Petersen. ‘Working Capital and FixedReview (summer 1990), pp. 3141. Investment: New Evidence on Financing Constraints,’ RAND Journal

_________ and Anthony P. Rodrigues. ‘Nonbank lenders and of Economics (ontnmn 1993), pp. 328-42.the Credit Slowdown,’ in Studies on Causes and Consequences Fazzari Steven M., R. Glenn Hubbard, and Bruce C. Petersen.of the 1984-1992 Credit Slowdown. Fedeml Reserve Bank of “Financing Constroints and Corpomte Investment.” Brookings PapersNew York, 1994, on Economic Activity (1988:1), pp. 141-95,

Carpenter, Robert, Steven M. Fazzori and Bruce C. Petersen. ‘Inventory Federol Reserve Bank of New York. Studies on Causes and(Dis)investment, Internal Finance Fluctuations, and the Business Consequences ofthe 1989-1992 Credit Slowdown, Febmary 1994.Cycle,” Bmokings Papers on Economic Activity (1994:2).

Fisher, Irving. ‘The Debt-Deflation Theory of Great Depressrons.”Cecchetti, Stephen G. ‘Distinguishing Theories of the Monetory Econometrica / (Octuber 1933), pp. 337-57.

Transmission Mechanism,” this Review (May/June 1995), .

nn 83-100 Frtedman, Benlamtn M., and Kenneth N. Kuttner, Economtc Actnvntyand the ShortTerm Credit Markets: An Analysis of Prices and Quantities,”

Chirinko, Robert S. ‘Business Fixed Investment Spending: A Critical Survey,” Brookings Papers an Economic Activity (1993:2), pp. 193-266.Journal of Economic Literature (December 1993), pp. 1875-911.

Gertler, Mark. Comment on Chrrstina 0. Romer and Dannd H. Romer,Christiono, lawrence, and Martin S. Eichenboam. ‘liquidity Effects and ‘Credit Channel or Credit Actions?: An Interpretation of the Postwar

the Monetary Transmission Mechanism,’ The American Economic Transmission Mechanism,” in Chonging Capital Markets: ImplicationsReview (May 1992), pp. 34653. for Monetary Policy. Federal Reserve Bank of Kansas City, 1993.

__________ __________ and Charles Evans. ‘The Effects of Monetary __________. ‘Financial Capacity and Output Fluctuations in anPolicy Shocks: Evidence from the Flow of Funds,” Review of Economy with Multi-period Financial Relationships,” Review ofEconomics and Statistics (forthcoming). Economic Studies (1992), pp. 455-72,

Cohen, Gerald, and John Wenninger. ‘The Relationship Between the __________, ‘Financial Structure and Aggregate Economic Activity: AnFederal Funds Rate and Economic Activity,” working paper (December Overview,’ Journal of Money, Credit and Banking (August 1988, part 2),

1993), Federal Reserve Bank of New York. pp. 559-88.

Cover, John Pery. “Asymmetric Effects of Positive and Negative Money __________ and Simon Gilchrist, “Monetary Policy, Business Cycles, andStpply Shocks,” QuarterlyJournal of Economics (November 1992), the Behavior of Small Manufacturing Firms,” Quarterly Journal ofpp. 1261-82. Economics (May 1994), pp. 309-40,

Cummins, Jason G., Kevin A, Hasseti and R. Glenn Hubbard. ‘A __________ and __________. ‘The Role of Credit Market ImperfectionsReconsideration of Investment Behavior Using Tax Reforms as Natural in the Monetary Transmission Mechanism: Arguments and Evidence,”Experiments,” Srookings Papers on EconomicActivity (1994:2), Scondinavian Journal of Economics (No. 1, 1993), pp. 43-64.

1~174, Gertler, Mark, and R. Glenn Hubbard. “Corporate Financial Policy,Diamond, Douglas W. “Financial Intermediation and Delegated Taxation, and Macroeconomic Risk.” RAND Journal of Economics

Monituring,” Review of Economic Studies (July 1984), pp.393-414. (summer 1993), pp. 286-303.

FE~SR*L RESERVE SANK OF ST. LOUIS

72

II E~IU~MAV/JUHE 1995

___________ and ___________- “Financial Factors in BusinessFluctuations,” FinancialMarket Volatility. Fedeml Reserve Bank ofKansas City, 1988.

Gorton, Gory, and George Pennacchi. ‘Money Market Funds and FinanceCompanies: Are They Banks of the Future?” in Michael Klousner andlawrence J. White, eds., Structural Change in Banking. Irwin Publishing,1993, pp. 173-214.

Greenwald, Bruce C., and Joseph E. Stiglitz. “Financial MarketImperfections and Business Cycles,” Quarterly Journal ofEconomics(February 1993), pp. 77-114.

___________ and ___________. “Information, Finance Constsaints, and

Economic Activity,” in Meir Kohn and S.C. Tsiang, eds., FinanceConstraints, Expectations, and Economic Activity. Oxford UniversityPress, 1988.

Gurley, John, and Edward Show. Money in a Theory of Finance.Brookings Institution, 1960.

___________ and ___________- “Financial Aspects of EconomicDevelopment,” The American Economic Review (September 1955),pp. 515-38.

Hancock, Diana, Andrew J. loing and James A. Wilcox. ‘Bank CapitalShocks: Dynamic Effects on Secutities, loans, and Capital,” workingpaper (August 1994), Board of Governors of the Federal ReserveSystem.

Hardauvelis, Gikos, and Thierry Wizman. ‘The Relative Cost of Capitalfor Marginal Firms Over the Business Cycle,” Federal Reserve Bank ofNew York Quarterly Review (autumn 1992), pp. 59-68.

Hart, Oliver, and John Moore. “A Theory of Debt Based on theInalienability of Human Capitul,” Discussion Paper No. /29(1991),london School of Economics, Financial Markets Group.

Hirtle, Beverly, and Jeanette Kelleher. ‘Financial Market Evolution andthe Interest Sensitivity of Output.’ Federal Reserve Bank of New YorkQuarterly Review (summer 1990), pp. 56-70.

Hoover, Kevin 0., and Stephen J. Perez. “Post Hoc Ergo Propter Once More:An Evaluation of ‘Does Monetary Policy Matter?’ in the Spirit of JamesTabin,’ Journal of Manetury Economics (August 1994), pp. 47J3.

Hoshi, Tokeo, Mdl K. Kashyop and David Schanistein. ‘Corporate Structure,liquidity and Investment: Evidencefrom Japanese Panel Data,”Quarterly Journal of Economics (February 1991), pp. 33-6D.

Hubbard, R. Glenn. Money, the Financiol System and the Economy.AddisomWesley, 1994.

__________ and Anil K. Kashyap. “Internal Net Worth and theInvestment Process: An Application to U.S. Agriculture,’ Journal ofPolitical Economy (June1992), pp. 5D6-34.

_____________________ and Toni Whited. ‘Internal Finance and FirmInvestment,’ Journal of Money, Credit andBanking (forthcoming).

Jaffee, Dwight, and Thomas Russell. “Imperfect Information,Uncertainty and Credit Rationing,” QuarterlyJournal of Economics(November 1976), pp. 651-66.

James, Christopher. ‘Some Evidence an the Uniqueness of Bank loans,”JournolafFinoncialEconamics(1987), pp. 217-36.

Jensen, Michael, and William Meckling. “Theory of the Firm:Management Behavior, Agency Costs, and Ownership Stuuctum,’Journal of Financial Economics (October 1976), pp. 305-60.

Koshyop, Anil K., and Jeremy C. Stein. ‘Monetary Policy and Banklending,” in N. Gregory Mankiw, ed,, Monetary Policy. University ofChicago Press for the National Bureau of Economic Research, 1 994a,pp. 221-62.

_______ and _______. “The Impact of Monetary Policy on BankBalance Sheets,” Camegie-Rochester Conference Series on PublicPolicy (forthcoming).

_____________________and David Wilcox. ‘Monetary Policy andCredit Conditions: Evidence from the Composition of ExternalFinance,” The American Economic Review (March 1993), pp. 78-98,

_________ Owen A. lamont and Jeremy Stein. “Credit Conditions andthe Cyclical Behavior of Inventuries: A Case Study of the 1981-82Recession,” QuarterlyJournal of Economics (August1994), pp. 565-92,

King, Robert G., and Charles Plosser, “Money, Credit and Ptices in a

Real Business Cycle,” The American Economic Review (June 1984),pp. 363-80.

___________ and Ross levine. “Finance and Growth: Schumpeter MightBe Right,” Quarterly Journal of Economics (August 1993), pp. 717-38.

Kiyotaki, Nobuhiro, and John Moore. “Cmdit Cycles.” mimeo (March1993), london School of Economics.

Kliesen, Kevin L, and John A. Totam. ‘The Recent Credit Crunch: TheNeglected Dimensions,” this Review (September/Dctuber 1992),pp. 18-36.

lamont, Owen S. “Corporate Debt Overhang and MacroeconomicVulnerablity,” working paper (1993), Massachusetts Institute of

Technology.

lucas, Robert E., Jr ‘liquidity and Interest Rates,’ Journal of EconomicTheory (1990), pp. 237-64.

lucas, Deborah J., and Robert L McDonald. ‘Bank Financing andInvestment Decisions with Asymmettic Information About loanQuality,’ RAND Journal of Economics (winter 1991), pp. 86-1 05.

lummer, Scott, and John McConnell. “Further Fvidence on the BankLending Process and Capital Market Response to Bank loanAgreements,” Journal ofFinancial Economics (1989), pp. 99-1 22.

Mauskopf, Eileen. “The Transmission Channels of Monetary Policy:How They Hove Changed,’ Federal Reserve Bulletin (December1990), pp. 985-1008.

Minsky, Hyman P. John Maynard Keynes. Columbia University Press, 1975.

___________ “Financial Crisis, Financial Systems, and the Performanceof the Economy,” in Commission an Moneyond Credit Private CapitulMarkets. Prentice-Hall, 1964,

FEDERAL RESERVE RANK OF ST. LOUIS

73

MAV/JUNE 7995

Mishkin, Frederic S. ‘The Household Balance Sheet and the Great _________ and _________, ‘Does Monetary Policy Matter?: A NewDepression,” Journal of Economic History (1978), pp. 918-37. Test in the Spirit of Friedman and Schwartz,” in Olivier J, Blanchard

and Stanley Fischer, eds., National Bureau of Economic ResearchMishkin, Frederic S. “Whatoepressed the Consumer?: The Household Macroeconomics Annual. MIT Press, 1989, pp. 121-70.

Balance Sheet and the 1973-75 Recessian,” Brookings Popecs onEconomic Activity (1977:1), pp. 123-64. Roosa, Robert V. “Interest Rates and the Central Bank,” in Money,

Trade, and Economic Growth: Essays in Honor of John H. Williams.Modigliani, Franca, and Merton Miller, “The Cost of Capital, Corporation Macmillan, 1951.

Finance and the Theory of Investment,” The American EconomicReview (June 1958), pp. 261-97. Ratemberg, Julio J. “AManetary Equilibrium Model with Transactions

Costs,” Journal of Political Economy (February 1984), pp. 40-58.Morgan, Donald P. “The lending View of Monetary Policy and Bank

loan Commitments,” working paper (July 1993), Federal Reserve Rothschild, Michael, and Joseph E. Stiglitz. “Equilibrium in CompetitiveBank of Kansas City. Insurance Markets: An Essay an the Economics of Imperfect

Information,’ QuarterlyJournal of Economics (November 1976),_________ “Asymmetric Effects of Monetary Policy.” Federal Reserve j. 63~49.

Bank of Kansas City Economic Review (second quarter 1993), pp. 21-33.Schreft, Stacey. “Credit Controls: 1980,” Federal Reserve Bank of

Myers, Stewart, and Nicholas Majiuf. “Corporate Financing and Richmond Economic Review (November-December 1990), pp. 25-55.Investment Decisions When Firms Have Information that Investors DoNot Have,” Journal of Financial Economics (1984), pp. 1 87~221. Shapiro, Matthew 0. “Federal Reserve Palicy: Cause and Effect,” in N.

Gregory Mankiw, ed., Monetary Policy. University of Chicago Press,Oliner, Stephen 0., and Glenn 0. Rudebusch. “Is There a Bank Credit 1994, pp. 207-34.

Channel for Monetary Policy?” Board of Governors af the FederalReserve System, Finance and Economics Division Discussion Paper No. Sharpe, Steven. ‘Asymmetric Information, Bank lending, and Implicit93-B (February 1993). Contracts: A Stylized Model of Customer Relationships,” Journal of

Finance (September 1990), pp. 1069-87.___________ and ___________. “Sources af the Financing Hierarchy for

Business Investment,” Review of Economics and Statistics (November Stiglitz, Joseph E., and Andrew Weiss. “Credit Rationing in Markets with1992), pp. 643-54. Imperfect Information,” The Americon Economic Review (June 1981),

pp. 393-410.Owens, Raymond, and Stocey Schreft. “Identifying Credit Crunches,”

Federal Reserve Bank of Richmond Working Paper No, 92-I Thornton, Daniel L ‘Financial Innovation, Deregulation and the “Credit

(March 1992). View” of Monetary Policy,” this Review (January/Febmary 1994),pp. 31-49.

Peek, Joe, and Eric Rosengren. ‘The Capital Crunch: Neither a BorrowerNor a lender Be,” Journal of Money, Credit and Banking (forthcoming). Townsend, Robert. “Optimal Contracts and Competitive Markets with

Cosify State Verification,” Journal of Economic Theory (Octaber 1979),_________ and _________, “The Capital Crunch in New England.” pp. 265-93.

Federal Reserve Bank of Boston New England Economic Review(May-June 1992), pp.2l-3l. Williamsan, Stephen. “Cosdy Monitoring, Optimal Contmcts, and

Equilibtium Credit Rationing,” Quarterly Journal of EconomicsPerry, George L, and Chades L Schultze. ‘Was This Recession Different? (February 1987), pp. 135-45.

Are They All Different?” Brookings Papers on Economic Activity(1993:1), pp. 145-211, Wojnilower, Albert, ‘The Central Role of Credit Crunches in Recent

Financial History,” Broakings Papecs on Economic Activity (1980:2),Petersen, Mitchell A., and Raghuram G. Ralan. ‘The Benefits of Firm- pp. 277-326,

Creditor Relationships: Evidence from Small Business Data,” Journal ofFinance (March 1994), pp. 3-38. Zeldes, Stephen P “Consumption and liquidity Constraints: An

Empitical Investigation,” Journal ofPolitical Economy (April 1989),Ramey, Valerie. “How Important is the Credit Channel in the Transmission pp. 30546

of Monetary Policy?” Carnegie-Rochester Conference Series on PublicPolicy (December 1993), pp. 1-45.

Romer, Chtistino 0., and David H. Romer. ‘Credit Channel or CreditActions?: An Interpretation of the Postwar Transmission Mechanism,”in Changing Capital Markets: Implications for Monetary Policy.Federal Reserve Bank of Kansas City, 1993.

___________ and ___________. ‘New Evidence on the MonetaryTransmission Mechanism,” Brookings Papecs on Economic Activity(1990:1), pp. 149213.

FEDERAL RESERVE SANK OF ST. LOUIS

74

II II ~I II ~MAY/JUNE 1005

THE FINANCIAL ACCELERATOR AND THE CREDIT VIEW

There are three basic conclusions ofmodels of financial frictions relating to asym-metric information between borrowers and)enders: (1) Uncollateralized external financeis more costly than interna) finance; (2) the

spread between the cost of external andinternal funds varies negatively with the levelof the borrower’s internal funds; and (3) areduction in internal funds reduces the bor-rower’s spending, holding constant underlyinginvestment opportunities. I illustrate cheseconclusions (and hnk them to empirical testsof credit view models) below in a simplemodel of firm investment decisions adaptedfrom Gertler and Hubbard (T988).

Consider two periods—zero and one.In thefirst, arisk-neutral borrower uses inputsto produce output Y to sell in the secondperiod. These inputs are hard capital, K—say, machinery—and soft capital, C—inputswhich improve the productivity of hard capital(such as organizational or maintenanceexpenditures). The production technologyis risky, with two possible productivity states,“good” and “bad”; uncertainty is realizedafter the investment decision is made.

To make the example as simple as possi-ble, suppose the firm can increase the chanceof a good output realization if it uses a stnffi-cient quantity of soft capital, where sufficientis defined by a level proportional to thequantity of hard capital used, In particular,let output Y satisfy:

(1A) Y =f(K), with probability tar,

and

if

and

(2A)

if

Y = cvf(K), ~th probability lrr~,

C vK,

Y= crf(K),

CCvK,

wherejKK) is twice continuously differen-tiable, strictly increasing, and strictly con-

cave (wheref(O) = O,f’(O) = oo, andf’(z) — 0

as z —~ 00); ira + gb = 1; 0 <cc <1; v> 0; andthe random productivity realization is idio-syncratic.

The structure of the problem guarantees

that the firm will either use vK units of softcapital or none, For simplicity, assume thatit is always efficient to employ soft capital.(Formally, this requires one to assume that(irs + zrbc4I(1+v) >ct).

If there are no informational imperfec-

tions, the firm’s investment decision is intu-itive, It chooses K to satisfy

(3A) (ir5

+ ,rta)f~(K)—(1 + v)r = 0,

where a- is the gross interest rate faced by thefirm. Equation 2A simply states that, at the

optimum, the expected marginal benefitfrom an additional unit of hard capital (givena complementary addition of v units of softcapital) equals the marginal cost of investing.The value of K that satisfies equation A2 —

call it K*_ does not depend on any financialvariables; that is, the Modigliani and Millertheorem applies.

The traditional interest rate channeloften identified with the money view mecha-nism is easy to illustrate in this example.Suppose for simplicity that the interest ratepaid on deposits is zero, so that a- representsthe gross required rate of return on lending.To the extent that an open market sale raisesa-, investment demand falls, This is the usualtextbook interest rate channel for monetarypolicy

Under asymmetric information, thestory is more complicated. Consider, forexample, a simple agency problem:Expenditures on hard capital are observableby outside lenders, while expenditures onsoft capital are not. In this case, the managermay be tempted to divert soft capital fundsto personal gain. Such perquisite consump-tion can take a number of forms. For sim-plicity assume that the manager can investthe funds (say in a Swiss bank account) toyield a gross interest rate, a-.

FEDERAl RESERVE BANK OF ST. LOUIS

75

llF~IF~MAV/JUNE 1995

Lenders understand this temptation,

and modify the financial contract to mitigateincentives to cheat. As shown below, oneconsequence of this modification is that desired

capital, K*, may exceed actual capital, K, andthis gap will depend inversely on the bor-rower’s net worth. Suppose the firm signs aloan contract with a competitive financial

intermediary The firm has some initial liq-uid asset position, W, and collaceralizablefuture profits, V, in period one, worth a pre-sent value of Via-. Hence, the firm’s initial networth is (W + Via-). To make the story inter-esting, assume that W <K~that is, the firmwould like to borrow, (For a richer descrip-tion of the role of internal net worth in thecontracting problem, see Gertler, 1992.)

The firm-intermediary loan contractspecifies the amount borrowed (in this case,(1 + v)K — W)), a payment P5 to the inter-mediary in the event that the project yieldsthe “good” output level, and a payment Pb inthe event of the “bad” output level. Thesecontractual features are chosen to maximizethe firm’s expected profits:

(4A) (irr+zrba)f(K)_ZVSPS_ irbpb,

From the intermediary’s perspective, theloan contract must offer an expected returnequal to its opportunity cost of funds, which

equals the gross interest rate a- times thequantity borrowed:

(5A) ir5P5+ ir~p~= r[(i + v)K — w].

That is, for simplicity assume that the inter-mediary simply channels funds from saversto borrowers, and uses no resources.

Given the underlying incentive problem,the contract must give the firm the incentiveto invest in soft capital as a complementaryinput to hard capital. That is, the contractmust satisfy the “incentive constraint:”

(6A) (ir5+ ir5a)f(K)- (zr5P5+2~

hPr~)

Equation 6A just states that themanager’s

expected gain from honest action exceedsthe gain from diverting the soft capital fundsto personal use.

One way in which the intermediary

could reduce the entrepreneur’s temptation

to cheat is to increase the amount of ph that

the firm must pay the intermediary in theevent of a bad outcome, The firm, however,can only credibly promise to pay availableassets in the bad state, That is, a limited lia-

bility constraint influences the contract:

(7A) p5

~ af(K)+V

To summarize, the contracting probleminvolves the selection of K, P5 and P5 to max-imize equation 4A subject to equations 5A,flA and 7A. One case is easy: As long as theincentive constraint does not bind, actualinvestment, K, simply adjusts to desiredinvestment K*, In addition, the pattern ofcontract payments is indeterminate. (Forsimplicity I am abstracting from a richerstructure that would lead to both debt andequity contracts and tax considerations; see,for example, Gertler and Hubbard, 1993, forsuch a treatment.)

When the incentive constraint in equa-

tion 6Abinds, financing and investmentdecisions are no longer independent. First,note that when the incentive constraint

binds, it is desirable to raise P5 to the maxi-mum extent possible; therefore, the limitedliability constraint in equation 7A also binds.Using 5A and 7A, one can eliminate ~5 andp5 from equation 6A, and thereby obtain arelation among K, the interest rate and inter-

nal net worth:

(8A) (~a+ ir~ct)f(K)

+ 2v)]K + a-( w + via-) = 0.

As long as equation SA holds, investmentK is an increasing function of the borrower’s

net worth (W+ Via-), holding investmentopportunities constant:

(9A) d(W±V/r)

=1(1+2v)_(inr~)1’(K)un1’>0,

The explanation for this effect is that,when the incentive constraint binds, an

increase in internal net worth increases theamount of feasible investment.

The existence of the net worth channel

precludes neither the traditional interest ratechannel nor the bank lending channel, To

FEDERAL RESERVE BANK OF ST. LOUIS

76

llFfl[~MAY/JUNE 1095

see the former, note an increase in lenders’opportunity cost of funds on account of amonetary contraction reduces desired invest-ment K* (since K* is determined by (ia

5+

zr5cc)f’(K) = (1 + v)a-). To see the latter, notethat, to the extent that banks face ahighermarginal opportunity cost of funds because ofa less than perfectly elastic supply schedulefor managed liabilities (and borrowers lackaccess to nonbank finance), the increase in a-lowers both desired and actual investment.

This simple framework is consistentwith the description of the flitancial acceleratormechanism: The cost of uncollateralizedexternal finance exceeds that for internalfinance. This gap varies inversely with theinternal net worth of the borrower and adecline in net worth reduces the borrowers’spending, all else equal. The framework alsoyields simple testable predictions related tothese money view and credit view arguments:(1) When informational imperfections are

ignored, an increase in real interest ratesfollowing a monetary contraction should

affect investment (broadly defined) simi-larly for borrowers of a given type (forexample, with similar technology and

risk characteristics).(2) If informational imperfections are signif-

icant only on the borrower side, all else

equal, spending by borrowers with lowerlevels of internal net worth should fallrelative to spending by borrowers withhigher levels of internal net worth.

(3) For bank-dependent borrowers, the avail-

ability of monitored bank credit can bethought of as a substitute for internal networth. Changes in the availability of bank

credit can influence the ability of bank-dependent borrowers to finance spending.

(4) The model’s intuition can apply to banksas well as nonfinancial borrowers. Adecline in banks’ net worth raises banks’opportunity cost of external funds (say inthe CD market). Asa result, the cost offunds to bank-dependent borrowers rises.

(5) If relationships between borrowers andspecific banks are important, shocks tothe balance sheet positions of individuallenders affect credit availability (at anygiven open market interest rate) to their

borrowers,

FEDERAL RESEEVE BANK OF ST. LOUIS

77