Charles W. Calomiris

Charles W Calomiris is associate professor of finance,University of Illinois at Urbana-Champaign; faculty researchfellow, National Bureau of Economic Research; and visitingscholar at the Federal Reserve Bank of St. Lou/s. Greg Chaudoin,Thekla Halouva and Christopher A. Williams provided researchassistance. The data analysis for this article was conducted inpart at the University of Illinois and the Federal Reserve Bankof Chica got

11~Jsthe Discount Window

Necessary? A Penn CentralPerspective

N RECENT YEARS, ECONOMISTS have cometo question the desirability of granting banksthe privilege of borrowing from the FederalReserve’s discount window. The discount win-dow’s detractors cite several disadvantages.First, the Fed’s control over high-powet-edmoney can be hampered. tf bank borrowing be-havior is hard to predict, open market opera-tions cannot pem’fectly peg high-powered money,which somne economists believe the Fed shoulddo. Second, there are microeconomic concernsabout potential abuse of the discount window(Schwartz1 1992). Critics argue that the discounttvindov has been misused as a transfer schemeto hail out (or postpone the failure of) troubledor insolvent financial institutions that should beclosed quickly to prevent desperate acts of fraudom’ excess risk taking by bank managemnent. Inresponse to growing criticism of Fed lending toprop up failing banks, Congress mandated limitson discount lending to distressed banks, whichwent into effect on December 19, 1993.

Some economists (Goodfmiend and King, 1988;Bordo, 1990; Kaufman, 1991, 1992; and Schwartz,1992) have argued that there is no gain from al-lowing the Fed to lend through the discountwindow. These cmitics argue that open marketoperations can accomplish all legitimate policygoals without resort to Federal Reserve lendingto banks. Clearly, if the only policy goal is topeg the supply of high-powered money, openmarket operations are a sufficient tool. Similar-ly, the Fed could peg interest rates on tradedsecurities by purchasing or selling them. Anyargument for a possible role for the discountwindow must demonstrate that pegging the ag-gregate level of reserves in the economy, Orcontrolling the riskless interest rate on tradedsecurities, is insufficient to accomplish a legiti-mate policy objective that can be accomplishedthrough Fed discounting.

In this article, I examine theoretical assumnp-

tions that may justify the existence of the dis-count window. I argue that there is little cum’rent

role for the discount window to protect againstbank panics. ‘I’he main role of the discount win-dow is in defusing disruptive liquidity crisesthat occur in particular normbarmk financial mar-kets. I discuss evidence from the Penn Centralcrisis of 1970, which seems consistent with thatview, and conclude by considering whether thisevidence is relevant for today’s relatively sophisti-cated financial environment.

Backup protection for financial marketsthrough the discount window could he achievedat little cost if access to the discount windowwere confined to periods of financial disruption.During normal times, open market operationsand interhank lending would be sufficient fordetermining the aggregate amount of reservesin the banking system and theim allocationamong banks.

A first step toward envisioning a role for anyfinancial institution or policy instrument, includ-ing the discount window, must be the relaxationof the assumptions of zero physical costs oftransacting and/or symmetric information. Thediscount window’s benefit, if any, must be relat-ed to its role in helping to economize on costsin capital markets, which themselves are a func-tion of physical or informational “imperfections.”I divide the discussion of potential justificationsfor the discount window into two parts—assistance to financial intem’mecliaries andassistance to particular financial markets.

ririu:f;,T,1~ DISc.i1L...r N’I’ %j/lNIIfO’i4 /%~~‘Li)

B.A.i.V.K1.Nfi PANICS

The Federal Reserve System was created in

1913 with three primam’y objectives: to eliminate

the ‘‘pyramiding’’ of reserves in New York City

and replace it with a polvcentric system of 12reserve banks; to create a more seasonally elas-

tic supply of hank credit; and to reduce thepropensity for banking panics. The discount

window was the primary mechanism for achiev-ing these goals. The 12 regional Federal ReserveBanks offered an alternative to the private inter-bank deposit market as depositories of bankreserves. The architects of the Fed expected toeliminate reserve ‘pyramiding,” which chan-neled interhank deposits to New York, wherethey often were used to finance securities mar-ket transactions (White, 1983). lnterbank lend-ing was viewed by some as a problem becauseit encouraged dependency of the nation’s bankson New York hankers and placed funds into thehands of securities market speculators.

The discount window also pm’omised to reducethe seasonal volatility of interest rates and in-crease the seasonal elasticity of bank lending byproviding an elastic supply of reserves, allowingbank balance sheets to expand seasonally withoutincm’easing the loan-to-asset ratio. Prior to thecreation of the Fed, hank expansion of loans inpeak seasons led to costly increases in portfoliorisk (a higher loan-to-asset ratio), or costlyseasonal importation of specie. This implied anupward sloping loan supply function with lam’geinterest rate variation over the seasonal cycle(Miron, 1986; Calomiris and Hubbard, 1989; andCalomiris and Gom’ton, 1991).

Finally, the availability of the discount windowwas also expected to reduce the risk of bank

panics in two ways. First, by increasing theavailability of reserves, the discount windowlimited seasonal increases in portfolio risk andreductions in hank liquidity (luring high-lendingmonths, thus reducing the risk of pam~ics.Sec-ond, the discount window would provide asource of liquidity to banks if an unpredictablewithdm’awal of deposits in the form of currencycreated a shortage of reserves that threatenedthe liquidity of the banking system (as in Dia-mond and Dybvig, 1983).’ But the discount win-dow offered limited protection to banks froma panic induced by adverse economic news.Because access to the discount windo~vwas

1lf money-demand disturbances were the cause of bankingpanics, as envisioned in Diamond and Dybvig (1983), thenopen market operations, as normally defined, would be asufficient tool for policy if the central bank were permittedto purchase bank loans. Since bank loans are not “spe-cial” in that framework (that is, there is no delegated con-trol and monitoring function performed by the banker and,hence, no potential for adverse selection or moral hazard),it is natural to think of standard open market operations asincluding purchases of bank debt in the context of thatmodel. If, however, banking panics are produced by contu-sion over the incidence of shocks to the value of bank as-sets, as argued in Calomiris and Gorton (1991). and if

banks have special information about their portfolios, thena government policy of purchasing bank loans during a cri-sis at pre-panic prices would have the same costs andbenefits as allowing banks access to the discount window.

limited by strict collateral requirements, bankhortowing was limited to the amount of eligiblecollateral the bank possessed.2 Thus, FederalReserve Banks could not use the discount win-dow to shore up banks if their depositors lostconfidence in the quality of the hank’s illiquidloan portfolio. The collateral required fordiscount window lending was subsequentlybroadened in the 1 930s.

‘11w history of the pre-Fed era suggests thatthe earls’ limitations on discount window lend-ing were important. (iorton (1989), Calomirisand Gorton (19911, and Calomiris and Schwei-kart (1991) have argued that sudden withdraw-als from the banking system occurred whendepositors received news about the state of theeconomy that was had enough to make themthink that some banks were insolvent. Becausedepositors were uninformed about the incidenceof this disturbance across individual banks (be-cause of depositors’ limited information aboutbank portfolios) all banks’ depositors had anincentive to withdraw funds from their banksuntil they could better ascertain the risks of in-dividual banks. ‘I’hus, relatively small aggregateinsolvency risk could have large costs throughdisintermediation from banks.

Costs associated with hanking panics can moti-~‘ate a more aggressive policy than one requir-ing riskless collateral for all central bank lending.‘l’he central hank could provide loans to thebanking system on illiquid collateral to offsetthe temporary withdrawal of depositors’ funds.‘the rationale for this intervention lies in infor-nmtional externalities caused b panics. Banking

panics create negative externalities among banksand their customers. Banks whose assets havenot declined in value, and theim’ borrowers anddepositors, suffer because of the confusion overwhether they are among the banks holding low-value assets. The banks lose business, the bor-rowers lose access to credit, and the depositorslose interest and pay transaction costs of trans-

ferring funds out of the banking system. Banksand their’ borrowers benefit by keeping thehanking system from shrinking.

If hank credits and deposits play special rolesin financing and clearing transactions, then con-tractions in hank activity will he costly. The dis-count window can he thought of as a way tocoordinate a mutually beneficial decision amongdepositors not to withdraw their deposits during

panics. Removing the private incentive for depo-sitors to withdraw their funds makes all deposi-tom’s better off. While private deposits fall,

public “deposits’’ made through the discountwindow (the indirect assets of the public) rise tocompensate. Open market operations would nothe an adequate substitute policy. Open marketoperations would simply insulate the moneysupply from the reduction in the mone mul-tiplier as hank deposits and hank credit fell;they would not reduce withdrawals from banks.

Thus, one could argue for central hank adop-tion of the following rule fom’ use of a”backup”discount window: Undem- normal circumstances(when there is no general systemic banking pan-ic reducing private deposits in hanks), the cen-tral hank provides no loans to banks. During asystemic crisis, the central bank agrees to pro-vide loans to banks up to the amount of deposi-tor withdrawals (at an interest rate that fairlycompensates the government for the defaultrisk of the average hank). Such crisis loans musthe short-term and paid in full after the crisispasses (which, if history is an~’guide, should beno longer than two months). The governmentmnight increase the interest rate it charges onloans to banks over time to encourage them toassist in resolving the information asymmetrymore quickly (for example, by sharing informa-tion about themselves and one another). Thecentral hank might even charge a fee to banksex post as a function of actual losses, to furtherencourage good banks to hring the crisis to a

2These limitations were eliminated in the 1930s, For a dis-cussion of changing collateral requirements on Fed lend-ing, see Friedman and Schwartz (1963, pp. 190—5). Notethat lending from the Fed, even on riskless collateral, canprovide special assistance to banks (up to the amount oftheir risktess collateral) because the Fed enjoys a specialright to “lump the queue” of debt seniority. By taking thebest assets of the bank as collateral, the Fed effectivelysubordinates existing debt claims. Private creditors wouldnot be able to do so and, thus, would not be able to lendto the bank on the riskless collateral,

34

speedy conclusion.~As deposits return to thebanks, they would use them to repay thegovernmnent loans. Banks that fail to attractdepositors (relative to other banks) as the crisisdraws to a close would he denied continuing ac-cess to credit and would he allowed to fail.

In principle, banks might be able to preventpanics by pooling tesources privately withoutany intervention by the central bank. If thebanking systeni were able to allocate funds toinsure against banking panics by agreeing totreat deposits as a collective liability of all banksduring a systemic crisis (as some groups ofbanks did historically), then, so long as the pub-lic was conhdent of the aggregate solvency ofthe banking system, there would he no threatof systemic hank runs and no need for agovernment-m’un discount window to reduce thecosts of banking panics.4 Kaufman (1991) arguesthat intem’hank markets did not operate effec-tively historically, but that this is no longer thecase. He claims that the existence of themodern federal funds market obviates the needfor the discount window during crises becauseopen market operations, combined with inter-bank transfers, can funnel cash to whicheversolvent banks experience large withdrawals. Ifbanks as a gm-oup are willing to p00

1 theirgovernment security holdings during a crisis,then Fed purchases of securities combined withinterhank transfers to hanks that lack sufficientgovernment securities can keep the banking sys-tem afloat, and possibly prevent runs (if inter-hank insurance is credible cx ante).

Despite the existence of a delivery mechanisrit(the fed funds market), lending among hanksduring a crisis may not occur due to asymmet-ric information. If banks are unahle to regulateand credibly monitor each other’s portfolios andhehavior, they will he reluctant to insure oneanother during a banking panic. Even thoughthe interbank market operates quite well duringnormal times among most banks, it cannot

necessarily he relied upon to protect the bank-ing s\’stem from panics.

The interbank loan market call operate effec-tively so long as banks have adequate informa-tion about and control over each other’s actions.Lending banks must he confident that horrow-ers are not abusing the interbank market tosubsidize excessive risks or provide a bailout toinsider depositor-s of a failed bank. Althoughthis “incentive compatibility” requirement maybe difficult to satisfy, there are many examplesthat show it is possible to do so. Gorton (1985,1989), Calomiris (1989a), Calomiris and Kahn(1990, 1991), and Calomiris and Schweikart(1991) argue that information asymmetry abouthank borrowers and the consequent risk ofpanics prompted cooperative behavior amongbanks historically. Coordination among banks inresponse to panics characterized many coun-tries’ banking systems (notably England’s duringthe Baring Crisis of 1890, and Canada’s repeatedlyduring the 19th and 20th centuries). But in thetinited States, laws limiting hank branching andconsolidation effectively limited interbankcooperation. As the number of U.S. banks andtheir geographical isolation from one anothem in-creased, the feasibility of national cooperationwas undermined. A hank’s cost of monitoringand enforcing cooperative behavior rises withfragmentation, while the benefit to any hankfm-om monitoring and enforcing declmes withthe number of members in the coalition (thebenefit is shared by all).

‘I’hus, tile need for discount window assistanceto banks is magnified by unit hanking laws thatmake private interhank cooperation, lendingand mutual insum’ance infeasible. Absent suchregulations, the potential for costly banking pa~ics would he substantially reduced, and the ex-

pected benefits of discount window protection

of the banking system would hesmaller.~In closing, four points are worth noting. First,

I have not assumed that the government has

3There must be an implied “subsidy” relative to the termsby which private lenders would be willing to lend to thebank, or else government tending cannot prevent runs, Theactuarialy fair government lending will be lower than therates banks would pay in the private market, since govern-ment intervention reduces default risk.

4Calomiris (1990, 1992c) argues that a nationwide branchbanking system would not have experienced aggregate in-solvency risk even during the worst episodes of bankfailure and bank panic,

5See the related discussion of other countries’ experiencesin Bordo (1990) and Calomiris (1992a).

35

superior information regarding individual banksolvency—an alternative justification for govern-ment lending to banks even in noncrisis states.While such an argument can be made (based onthe government’s access to information by vir-tue of its supervisory role), the recent history ofbank failures and losses, and of regulatory agen-cies’ inabilities to anticipate, observe or preventwidespread abuse seems to argue against such apresumption. Kane (1988) argues that regulatorsface distorted incentives to collect and reportinformation about banks. These incentive prob-lems may outweigh regulators’ special channelsof information due to supervisory authority.

Second, discount lending can be motivated byphysical transaction costs that limit interbanklending. Such physical costs mean that openmarket operations will have uneven effects onthe supply of reserves available to differentbanks if the market for reserves is segmented.Although this may have been a legitimate moti-vation for the discount window historically, asKaufman (1991) argues, current interbankreserve transfers are accomplished at little cost.

‘l’hird, I have not addressed the possible roleof the discount window in bailing out a bankingsystem that is insolvent as a whole. Even in aconcentrated, mutually insuring banking system,interbank insurance and lending could neverdeal with enormous adverse asset shocks (thatis, those larger than aggregate bank capital).Partial government deposit insurance (with largedeductibles) for mutually insuring groups of

banks can protect against this unlikely eventbetter than wholesale bailouts through discountwindow “lending” (Calomiris, 1992b).

Fourth, the need for the discount window toprotect the current U.S. banking system fromfinancial panic has been substantially curtailedby deposit insurance,6 Under the current depositinsurance system, discount window interventionwould be largely redundant as protection againstsystemic risk. Insured depositors have little incen-tive to run their banks during a financial crisis.Although deposit accounts in excess of $100,000under current law are not protected (de jure)by government deposit insurance, largerdeposits may be covered if a general run on thebanking system ensued. The FDIC ImprovementAct (FDICIA) of 1991 establishes a formula fordetermining whether a systemic threat warrantsthe coverage of larger-denomination deposits!Fed lending does retain a potentially importantrole in providing implicit protection for the inter-bank clearing system, which is discussed below.~

~TO~TBt%NK LENDING AND THE

HOLE OF THE DISCOUNTIVINDOW

In an economy in which physical costs of in-terbank transfers are small, and interbank coor-dination and mutual insurance, or governmentdeposit insurance, protects the banking systemfrom the risk of panic, there is no additionalneed for the discount window to facilitate the

°ltis beyond the scope of this article to examine all of therelative advantages of government deposit insurance ordiscount lending for stabilizing a fragmented (uncoordinated)banking system. Perhaps the most obvious potential advan-tage of discount window lending is that government inter-vention can be state-contingent. If a bank fails when thereis no systemic panic, the bank’s depositors will not bebailed out by government insurance. This reduces themoral-hazard costs of the government’s “safety net?’ Thisargument for the relative desirability of the discount windowas a means to insure against panics presumes that thecentral bank will not cave in to the political pressures ofspecial interests to bail out banks in noncrisis times. Recentaccusations by the House Banking Committee of inap-propriate lending by the Federal Reserve to insolventbanks cast some doubt on the ability of current institutionsto make and enforce appropriate distinctions regardingwhen banks should have access to the discount window(see Business l4~ek,July 15, 1991, pp. 122—3). Schwartz(1992) argues that the history of the discount window isreplete with such examples. Congress has mandated, andthe Fed has implemented, specific new guidelines that limitFed lending to distressed banks (The American Banker,August 12, 1993, pp. 1—2).

‘tinder 12 U.S.C. § 1823 (c) (4) (C) of FDICIA, for insuranceto be extended to uninsured liabilities of a bank, beginning

in 1995, the FDIC, the Secretary of the Treasury (in consul-tation with the President), and a supernumerary majority ofthe boards of the FDIC and the Federal Reserve, mustagree that not doing so “would have serious adverse ef-fects on economic conditions or financial stability?’ If unin-sured deposits are covered through this provision, theinsurance fund must be reimbursed through emergencyspecial assessments. Because the nation’s largest bankswould end up paying a disproportionate cost of such abailout, they would be expected to lobby against the exten-sion of insurance to uninsured deposits, unless the criteriafor assistance were truly met.

°Theprotection afforded to bank clearing houses is consi-dered in more detail in the conclusion to this article.

35

operation of the banking systeni. But even insuch an environment, problems that atise out-side the banking system may motivate centralbank lending through the discount window. Inparticular, securities markets may be vulnerableto externalities arising from asymmetric infot’-mation. I will argue that these problems may headdressed effectively by channeling fundsthrough banks that horrow from the window,rather than thm’ough diiect lending from thecentral bank to nonbank firms.°The examplethat I will focus on is the commercial papermarket “run” that followed Penn Central’s 1970bankruptcy.

As many r’esearchem’s have stressed, the bank-ing system is particularly vulnerable to confu-sion about the incidence of disturbances fortwo reasons. First, its assets (that is, bank loans)typically are not traded in centralized markets.Thus, it is difficult for an uninformed bankdepositor to keep abreast of the effect of agiven news item on the value of his hank’s as-sets. Second, the fact that banks finance throughlarge quantities of demandable debt allows ner-vous depositors to withdraw from the hankrather than wait to see whether their bank willsurvive or fail.

Although these two attributes that makebanking panics possible—nontraded assets amiddemandable deht—seemn to set the banking sys-tem apart from other’ markets, the banking sys-tem is just an extreme case of a much moregeneral phenomenon. The condition necessaryto generate a costly panic in a debt market isthat the time horizon for rolling over debt isless than the time it takes to make accuratereappraisals of firm-specific risk during episodesof general bad news. Lenders’ lack of informna-tion about the attributes of specific firms mayresult in the pooling of borrowers with com-nion observable characteristics. In such circum-stances, firms will face temporarily high‘‘lemons premia’’ in debt and equity markets,~yhich will increase the cost of finance, and

reduce investment, even for firms whose tt-ue“fundamentals” are unaffected by the had news.

Firms with short-term debts (which must herolled over’ regularly) can he particularly vulner-able to systemic risk and the possibility of arun. A liquidity crisis that would prompt ageneral calling in of debt by creditors couldlead firms with outstanding short-term debt toexperience high costs of debt rollover or assetsale not experienced by other firms.

Furthermore, if intermediaries for particularmarkets (for example, commercial paper deal-er’s) suffer losses from one firm’s issues, theymay be less able to deal in the paper of otherfirms. This, too, can force firms to pay highercosts for funds temporarily in the affected mar-ket, or switch to new, higher-cost sources offunds.

Firms that face liquidity problems in nonhankdebt markets may have difficulty borrowingfrom bankers, too, particularly if they lack exist-ing bank-lending relationships. To the extentthat banks have special information about bor-rowers’ attributes, due to their past involvementwith firms and their’ ongoing monitor-ing of firmcom-pliance with lending covenants, banks maybe able to assist firms when their costs of fundsrise in other credit markets. For firms thatmoved away from reliance on bank credit, how-ever, there may be no strong banking relation-ship to fall hack upon. Assistance from banksfor these firms would be forthcoming only athigher interest rates, which would compensatebanks for the transaction and information costsof drafting emergency lending arrangements. Inparticular, if the hank expects only a temporaryrelationship r~’iththe firm in need (for the dura-tion of the “emergency”), the bank will have tocharge higher interest rates to recoup its fixedcosts over a shorter lending period.

Given the high cost of substituting bank creditfor other cr-edit on short notice, a credit marketrun may force some solvent firms into financialdistress, or simply reduce their ability to investor to lend to other firms. 10 If the social costs ofsuch disruptions to short-term debt markets arelarge, Fed intervention to defuse such crisesmay be warranted. Specifically, the Fed could

°Mishkin(1991a) also argues that asymmetric informationis relevant outside the banking sector, He uses data oninterest rate spreads between risky and riskless debtinstruments to support this view. He finds evidence of anincrease in these spreads (which he interprets as reflectingan increased inability to sort borrowers according to risk)coinciding with or prior to the Penn Central crisis of 1970and the stock market crash of 1987.

10Calomiris, Himmelberg and Wachtel (forthcoming) find thatnonfinancial commercial paper issuers of the 1980s tendedto be net lenders to other firms through accountsreceivable.

37

supply banks with funds at low cost throughthe discount window for the express purpose ofrefinancing maturing short-term debts of firmssuffering from disruption in the short-term dehtmarket. ln a competitive banking system, thissubsidy would be passed on to borrowers andwould mitigate high short-run costs of switchingto banks for credit.

New financial markets may be particularlyvulnerable to negative externalities among firmsor temporary disruptions to market dealers.The lack of data on the risks and liquidity ofnew products, and relatively thin trading, in-creases the likelihood of systemic risk in newmarkets.

In the following section, I consider whetherthe commercial paper market experienced sucha financial crisis in mid-1970, and whether thatcrisis warranted discount window intervention.The commercial paper market of mid-1970 is anespecially interesting case to examine for sixreasons. First, most commercial paper maturedquickly—with an average maturity of under 30days (Stigum, 1983, p. 632). This meant that asudden disinclination by investors to hold com-mercial paper would entail substantial problemsfor firms trying to roll over their commercialpaper debt.

Second, commercial paper was a new andgrowing method of finance during the 1960s.’1

Institutional arrangements for rating and sup-porting commercial paper issues were virtuallynonexistent; thus, information imperfectionswere potentially important.

Third, commercial paper finance originated asa substitute for bank credit. Many firms thathad moved to this matket in the 1960s mayhave curtailed or terminated their’ relationshipswith comnmercial banks (making the disruptionin the supply of paper more costly).

Fourth, during the early years of rapid growthin this market, there was a major shock to thecommercial paper market, namely the failure of

Penn Central in 1970, which was associatedwith substantial contraction of outstandingpaper (that is, a “run”).

Fifth, commercial paper issuers include manyof the economy’s largest firms, and other firmsoften depend upon them for credit (Calomiris,Flimmelberg and Wachtel, forthcoming). in-creases in the cost of funds for this class ofborrower-s thus may have significant second-order effects on the cost of credit for otherfirms.

Finally, the Fed intervened during this crisislargely by encouraging banks to come to thediscount window to finance the payoff of com-mercial paper. Evidence from the Penn Centralcommercial paper crisis of 1970 allows adetailed case study of “information externali-ties,” the potential for a run in markets fortraded short-term debt, and an evaluation ofFed intervention in response to such a crisis.

Penn Central’s Failure anti theLiquidittr C,’isis of’ Mid-1970

The facts surrounding the commercial paperrun following the Penn Central failure are com-monly known (see Schadrack and Breimyer,1970; Maisel, 1973; Timlen, 1977; Brimmer,1989; and Mishkin, 1991a), but some importantdetails are worth reviewing. Along with manyother firms, Penn Central’s financial conditiondeteriorated during the recession of 1969—70.Penn Central was a major issuer of commercialpaper, with more than $84 million outstanding,much of which came due in June, July andAugust of 1970. As Penn Central’s cash flowdeclined, its debt holders and their agents ap-pealed to the federal government for financialassistance, which the Nixon Administration sup-ported.

The Administration proposed a $200 millionloan guarantee to a syndicate of some 70 banks,which wet-c to provide a two-year loan in thatamount. The loan guarantee would he autho-r’ized through a loose interpretation of the

‘1There had been an earlier incarnation of the commercialpaper market that thrived from the 1870s and declined inimportance during the 1920s. Calomiris (1992a) argues thatthis operated effectively as an interbank loan-sale market,moving high-quality borrowers from high credit-cost areasto low credit-cost areas. Consistent with that argument,James (1994) views the demise of this market as the resultof the bank merger wave of the 1920s, which provided analternative means to channel credit through the financialsystem.

38

Defense Production Act. Although there was in-creasing congressional opposition to this plan,as late as Friday, June 19, the Wall Street jour-nal reported that “the opposition doesn’t yet ap-pear strong enough to halt the $200 million loanguarantee.” That article also reported the possi-ble existence of a secret memorandum fromthe Federal Reserve Bank of New York, recom-mending “that the loan be granted, based on aninvestigation that bank is believed to have con-ducted into the credit-worthiness of Penn Cen-tm-al.” Contrary to the Wall Street Journal report,no such memorandum existed, and that sameFriday the Penn Central plan was rejected byCongress. The Nixon Administration then askedthe Federal Reserve Board (through the NewYork Fed) to make a loan to Penn Central tohelp it meet immediate obligations. The NewYork Fed recommended against the loan, and itwas denied. This news forced Penn Central’sbankruptcy on Sunday, June 21.

The surprising news of the unwillingness ofCongress and the Fed to prop up Penn Centralcreated widespread concern over the weekendthat the Penn Central failure would have reper-cussions elsewhere in the economy, particularlyfor other firms that had large outstanding com-mercial paper issues. It is not easy to explainthis concern without invoking an “informationexternality” of some form. That is, one needs toexplain why the bad news about Penn Centralwould raise doubts about other firms.

The bad news about Penn Central on June 19had two parts. First, pm’ior to that date, the WallStreet Journal reported that the New York Fedhad made a favorable audit of Penn Central’sunderlying financial strength. After Friday,quite the opposite was known. The ieaction ofthe market, as reported in the press, was that ifPenn Central’s financial state could so rapidlyand unexpectedly have turned sour in the previ-ous year, what other “blue chip” commercialpaper issuers might be in the same position?‘t’his concern was fueled by the fact that theincome t’eductions during the recession of1969—70, which potentially affected many firms,were not known at the firm level with any pre-cision at the time. ‘those concerns about otherfirms began to he voiced even before the revela-tion of the New York Fed’s audit. For examnple,a lead article in the Journal on June 12 queried:“How many other U.S. corporations are so short

of cash that they may soon find themselvessimilarly unable to pay their bills?” Until themarketplace could assess the extent to whichPenn Central’s financial position was the resultof idiosyncratic shocks and mismnanagement, asopposed to a signal of a common problem likelyto be faced by many firms, Penn Central’sfailure would cast doubt on the financial posi-tion of other firms.

The second element of general bad newsrevolved around the fate of Penn Central andits creditors. It became clear that, whatever itsunderlying condition, the government would notguarantee Penn Central’s debt and that, there-fore, Penn Central’s creditors faced the possibili-ty of substantial losses. The incidence of losseson the firm’s commercial paper was unknown,but it was rumored that ownership was quiteconcentrated. For example, on June 1 the Jour-nal reported that Morgan Guaranty owned oracted as “agent” for nearly $84 million in PennCentral’s commercial paper. According to FederalReserve data on holdings of commercialpaper, in early 1970 nonfinancial corporationsowned about 74 percent of outstanding paper.12

The June 12 Journal article cited above alsoasked: “If even one major corporation shouldbecome insolvent, would its failure bring downother cash-short companies because the failingcompany couldn’t pay its hills? Could that, inturn, intensify the present severe strain on thecash resources of banks and corporations into aliquidity crisis, draining the flow of money andcredit and plunging the nation into a depres-sion?” While this “domino” scenario of economy-wide depression may seem a bit farfetched, itwould have been less farfetched to imagine thatone or two major commercial paper issuers(who may have been creditors of Penn Central)might also find it hard to repay their debts.

Thus, lack of infor-mation about the effects ofthe recession on other firms (which Penn Cen-tral’s failure indicated might be large), and aboutthe identities of Penn Central’s creditors andtheir creditors in turn, could have producedlegitimate, rational concern about rolling over’the commercial paper of other firms at pre-existing termns. ‘I’he commercial paper marketwas especially vulner-able to these sorts of doubtsbecause it was a fast-growing new financial mar-ket, as shown in Figure 1. From 1956 to 1966,

125ee Schadrack and Breimyer (1970, p. 283).

—o

-‘

~H~

CI~

A’O

c_

~~

~’~

’t’D

~0

-~

-~

~~&

~

~ ~~

—8

.nA.~

~t’D

~~

I~t

~z:.

~

A’~

t~

-j:1

’—

—c

—-+

~-cS

—,

-,~

~~r~

-’~

~z-5

~a:.

—o~

~’~

t-~

tZ_O

~~

:4~

~;_

~~

fl~

~O.~

rtm

i~r’

2.-~

rt~

~~

Ett

~_

~~

~‘:

~9

H2

2~

,~

-n—

‘—C

”~

-~-

jT=

E-2

EE

~8

~~~

:-~

~~“

o-~~

2‘~

—~

~‘2

~a

~rc

-~~

O~

’——

.A

’’t~

~

~~.

a~

~2

A’A’

——

~?—

~—

1-

‘-~

-,

—C’

~;

%-

ft~-

8z

~a’~

i5.‘r

~5

~2—

~‘o

n~

a—~‘

——

=~

Oz0

_.~

A’~

~r°c0

~’+

—.

—©

‘~-~

,,,C

D;4

0C

t 9j~

ca~

_.~

O~

-~

Ct~

-~

art

aA’

‘l©

a~

-oS

—,-a

C—

-‘—

or,

.,

~~

~2-

~a-

~~

Ca

‘tt

A’C

~-j-

+-n

‘I_

Ct

—C

-~‘

C’-

~A’

~ ‘C~

~5

d~

~2

~~

a’C

—’’tR

C’

~!C

CQ

~a

~E

‘ia

’.,

~‘t~

A’‘~

a

—S

:~P

PP

00

00

0

aw

Om

pPEo~5

0°3”3

c~ o

m—

’-h

—a2,

-‘~

0(n

o, -a 0 0 C U, 0) a. (a

C.,0

2.

H e 003

CDr CD

a0

CD

)C

sIn

(0

0~ 1°’

a’,

‘DC V

0) In so

0 5 5 CD

-s(I) a C a 0 U) C A, a C) 0) a 3 -s -4 0 C) a

-5~

-s (0 (31 C,) (31 Cl, U)

—I

(8 (0 a) -s 0) Cd,

0) (31 a) -‘I

a) (0 -s Co -‘I 0

40

ket pricing had been too optimistic in the 1960s.His (post-crisis) study of Penn Central’s financialposition in the 1960s concluded that there wasmunch to be learned from the Penn Central col-lapse about the need for greater caution in valu-ing commercial paper: “A careful financialanalyst might well have recommended.. againstthe purchase of Penn Central commercial papera year or more before the events of May andJune 1970.”~Murray accounted for the poorcx ante evaluation of risk by the fact that so“many new faces appeared in that market forlarge sums at the time and Penn Central washardly noticed as an unusual case.”

Schadrack and Breimyer (1970) provide a simi-lar perspective. They claim that before the PennCentral failure, “the confusion of corporate sizewith liquidity tended to mask some deteriora-tion during [the late 1960s] of the quality ofcommercial paper outstanding. - the fact that anumber of firms in the market by 1970 hadvery high debt-to-equity ratios and/or’ incomeflows of dubious quality (some conglomerate,franchising and equipment leasing companies,for example) suggests such a detem’ioration inthe quality of outstanding paper.” They alsoargue that, in addition to the concern aboutother commercial paper borrowem’s brought onby the failure of Penn Central, the bank’s failureraised concern about some of the major broker-age houses, which acted as dealem’s and pur-chasers in the market. Commercial papem’ dealersmaintain open positions in the paper they selleithem’ as part of an under-writing arrangement,or through a commitment to maintain a secon-dary market in the paper (Stigum, 1983, p. 640).The threat of a liquidity crisis for firms andtheir dealers led to a collapse of demand for thedebt instruments of others. ‘l’hese fears fueledthe flight to cash. Schadrack and llr’eimyer(1970) also argue that the crisis led to refinedmethods of pricing commercial paper, which isconsistent with Murray’s view that there wasroom for improvement. In particular, after thePenn Central crisis they found a wider disper-sion of rates for dealer-placed paper’, which

they interpreted as the result of “greater inves-tor selectivity.” Also, they noted a persistentshift toward bank CDs and Treasury bills.

As Mishkin (1991a) and Schadrack and Brei-myer (1970) point out, the spread between com-mercial paper and Treasury bills widened duringand after the crisis. This widening seems toreflect a persistent revision in the evaluation ofcommercial paper risk. Schadrack and Breimyer(1970) report that in November 1970 the dealerpaper rate averaged 103 basis points above theTreasury bill rate, compared to previous~~spreadsof roughly half that amount. A similar pattern isvisible in Table 1, which reports the federalfunds rate, three-month Treasury bill yields, thediscount rate, and the four-to-six-month primecommercial paper rate before, during arid afterthe crisis.

The “flight to quality,” visible in the decliningyields of Treasury hills and rising short-termspreads, is also visible in long-term yields andspreads, shown in Table 2. From June 20 toJune 27, Treasury bond yields fell as corporatebond yields rose. The spread between theTreasury bonds and the Aaa corporates reacheda peak on July 11. Interestingly, the spread be-tween Aaa- and Baa-rated bonds was essentiallyconstant during the crisis, but rose afterwards.‘t’his is consistent with the view that during thecrisis, increased riskiness was attributed to allsecurities, hut that, after the crisis, investom’swere better able to sort firms into risk categories.

Concerns about the financial condition ofcommercial paper issuers and dealers provedunwarranted ex post (since no other commer-cial paper issuers defaulted), hut seem to havebeen important cx ante, as evidenced by move-ments in the stock market and commercialpaper market. Firms, especially those with largeoutstanding debt, saw large stock price declinesin the first three days of the crisis. During thattime, the Dow Jones Industrial Average lost 28points (a fall of roughly 4 percent). Chrysler,General Motors and IBM all saw large losses asrumors circulated that they’ faced risks of beingunable to meet their debts (Wall Street Journal,

‘4See Murray (1971). Whitford (1993) applied Altman’s (1968)“z-score” model to Penn Central’s accounts as of Decem-ber 1969, and found a remarkably ow z-score of 0.135. Alt-man had found that no healthy firms had z-scores of below1.81 and no bankrupt firms had a score above 2.99.

“See Schadrack and Breimyer (1970, p. 289).

rcnrfl a, ,1,CC~CoC.;,r 11AI’~é (‘IC CT CC

41

Table 1

Selected Yields and Interest Rates3-month Treasury Discount 4-6 month prime

Date bill yield Federal funds rate rate commercial paper

1970January 7.89°/. 9 04% 6.00% 8 55%February 6.88 8A1 600 850March 616 7-45 600 8.03Apri 659 8.43 600 8.00May 7.00 764 600 813Junel 682 784 600 813

2 676 798 600 8th3 6.71 780 600 8254 651 721 6.00 825

July 1 6.46 7 23 6.00 8 387 6.62 f34 600 8353 646 7.59 600 8.254 634 716 600 8.35

August 6 25 6.34 6 00 7 70September 5 80 6 05 6.00 7 20Oclobor 584 611 600 663November 499 5 16 5.85 5.75December 4 83 4 82 5 52 5 75

NOTES: Data are all ow-at-month. excepl for June and July. wh.cb are reported end-of-week Trea-sury bill and commercial paper yields are quoted or June 6. 13 20 and 27 and July 4. 11.IS and 25 Federal funds rates ae ‘or June 3. 10. 17 and 24. and July 1. 8. 15 and 22

SOURCES Board of Governors of the Fedpral Reserve System f1976~Table 12.SB. Taoic 12 66.and Fable 12 78: Federal Reserve Bank of St Lobis

June 23-23, 1970, “Abreast of the Market”). Busi-ness Week quoted one stock market analyst assaying that “investors think that any company...with. - debt is going bankrupt” (June 27, p. 42).

Perhaps the best indicator’ of the extent ofthese fears is the contraction in the volume ofcommercial paper’ outstanding from late June tomid-July. Total outstanding nonbank comniercialpaper fell from $32 billion on June 24 to $29billion on July 15, with $2.3 billion of that declinein the first week of the crisis (see Figure 2).

Interestingly, commercial paper rates showedlittle change during the crisis, although thespread between paper rates and other moneymarket r’ates did widen. i’he reason for this wasthe speedy reaction of the Federal Reserve tothe failure of Penn Central. Luckily, it occurredover a weekend, which gave the Fed time topm’epare for the opening of financial markets onMonday. The Fed pursued four courses ofaction.

The Fed’s Discount Window PolicyDuring the Crisis

First, the Fed contacted member banks andnotified them that “as they made loans to enabletheir customers to pay off maturing commercialpaper amid thus needed more reserves, the Fed-eral Reserve discount window would be availa-ble.”°’Ihe meaning of “available” is of paramountimportance. The Federal Reserve let memberbanks know that if they borrowed at the dis-count window for purposes of making loans tocommercial paper issuers, they would be able todo so without incurring any costs other thanthe discount rate. The Fed was informed bybanks when their discount horr’owing resultedfrom financing commercial paper rollovers, andthe total amount of such discount borrowing to-taled some $500 million in the weeks immediate-ly following Penn Central (Melton, 1985, p. 158).Beyond the amount lent through the discountwindow, access to the window for’ commercial

16See Treiber (1970, p. 16).

42

Table 2Long-Term Yields and Spreads

Long-term Spread between Spread betweengovernment Aaa corporate Baa corporate Aaa and government Baa and Aaa

Dale’ bonds2 bonds3 bonds3 bonds corporate bonds

I 375Jaruary 684

1k /9Vu 88”t~ 107 090

February 625 783 877 1 58 0 9CMa-ch 633 792 8 Sb 1 ~9 0 74Apr:! 670 783 67~ 1 13 0.91Vay 721 821 9.10 100 089June 6 700 830 913 130 083

3 709 842 918 153 07620 705 8.bb 926 150 07127 689 860 936 171 076

July 4 673 860 941 187 081it 656 855 944 199 089

8 661 849 939 88 09025 654 840 938 186 098

August 673 813 9.47 140 134Septerroor 6 52 8 06 9 32 1 5-4 1 2Octooe’ 665 807 934 142 127Novemoer b97 802 937 2.05 135Doccrrber 6.05 51 902 1 46 1 31

‘All data ar~end-ui monin. rless ~lherwiscindicated-‘Marurty varies‘Ray-c by Mood-, sSOURCE Boaro of GoL-ernors of the Feoera- Reservp Svrrem f197b~ Thble 12.1’S

paper rollover’s gave

assurance to the financial markets that the liquidityessential ro their operation woutri he preserved. Ifpanicky investors refused to renew their’ holdingsof minotci ‘cia I paper, profe ‘ring Ti—ca sui—v hi!Is,hank deposits— anything!“—Cnstead, their extreme

pi.efer’ence for- safe my woo Id i in t he atlow ed tocontribute to widespread insolvency, Once ever’v-

nderstnod rtiat, there was little reason for

panic (Melton, 1985, p. 158).

Fed encouragement to use the discount win-dow to finance the payoff of commercial paperwas associated with reduced costs of horr’owingfrom the Fed, even though the discount rate re-mained unchanged. Normally, the costs of bor-rowing from the discount window include thediscount rate and a nonpecuniarv ‘‘hassle’’ cost.That is, the Fed does not want to encourageabuse of the privilege of borrowing from thediscount window and banks that may he seenas abusing the privilege run the risk of E:~xamni-nation and regulatory sanctions. ‘this penaltyexplains the positix’e difference between the fed

funds rate and the discount rate. If there werenO penalty, banks would be indifferent betweenborrowing from other banks and the Fed’s dis-count window. In this case, the two rates wouldbe identical. In the presence of a nonpecuniarycost of borrowing From the Fed, as long as bor-r’owings are positive, the fed funds r-ate will behigher than the discount rate since, on the mar-gin, banks will be indifferent between pavingthe fed funds rate in the interhank market andhor’rowing fr’oni the Fed (which entails a dis-count rate cost and a hassle cost).

Figure 3 provides a simple illustration of thesimultaneous determination of the feder-al fundsrate and borrowed reserves, which is helpful inanalyzing the effect of discount window lendingduring the Penn Central crisis. Reserve demandis shown as a negative function of the federalfunds rate. The position of the demand sched-ule varies with loan demand, reserve require-ments, and the demand for excess reserves. ‘l’heFed determines the amount of nonhorrowed

tzrflrasr przQrm,r name flc ciT i niridi

43

Figure 2Commercial Paper and Business LoansJune-August 1970

Billions of dollars33

32

31

30

29

28

* Including business loans sold to affiliates.

Source: Schadrack and Breimyer (1970), Chart V.

Billions of dollars

94

93

92

91

90

89

reserves through its open market operations.Borrowed reserve costs are given by an upwardsloping schedule, which sums a constant pecuni-ary cost (the discount rate) with an increasingnonpecuniar’y hassle cost. The more reservesthat are borrowed, the more the Fed is liable topenalize borrowing. Figure 3 illustr’ates equilibr’i-um in the reserve market for June 17 and July15, 1970, using actual data on the discount rate(which remained at 6 percent throughout thecrisis), nonborrowed reserves, borrowed reservesand the federal funds r’ate. Assuming equilibri-um in the reserve market, we can identify shiftsbetween these two days in reserve demand (asbank loans rose to compensate for the contrac-tion in commercial paper) and in r-eserve sup-ply.- ‘I’he reserve supply function shifted inslightly (nonhorrowed reserves fell due to in-cr-eased currency demand, which was onlypartly offset by open market operations) androtated downward as the Fed reduced its non-pecuniary penalty for borrowing.

The downward rotation of the borrowedreserve supply function illustrates how the Fed-eral Reserve lowered the nonpecuniarv cost ofborrowing from the discount window duringthe crisis. Other evidence on the composition ofbank lending, bank borrowings from the Fed,and the different rates charged to differenttypes of hank customers suggests that thereductions in nonpecuniary costs wer’e linked(as the quotation above suggests it was) to in-

direct subsidies for commercial paper rollovers.That is, it seems that loans to member banksfor this purpose were granted a special “subsi-dy” by the Fed (in the form of lower, or possi-bly zero, nonpecuniary costs).

Consistent with this account, the compositionof member bank borrowings changed duringthe crisis. As of June 24, large commercialbanks (primarily money-center’ banks) accountedfor omil~’75 percent of borrowing from the led.The trebling of member bank hot-rowing from

31017241 8 1522295121926June July August

44

Figure 3Shifts in the Reserve MarketJune 17-July 15

7807-59

6*

June 24 to July 15 was due to an increase inmoney-center borrowing, as one would expect ifit was earmarked for commercial paper payoff.As shown in Table 3, total borrowed reservesrose by 51.196 billion, while borrowed reservesof large commercial banks rose 81.224 billion.These same banks were the only ones that sawa large growth in loans to businesses andfinance companies during the crisis. Loans in-creased by 52.3 billion from June 24 to July 15,almost an exact offset of the amount by whichcommercial paper was reduced during this peri-od. This rise of 2.6 percent in total loans forthis group of banks was highly unusual. ‘I’heaverage rate of increase for the preceding fouryear’s during this period of the year had been0.03 percent, and the highest rate of growth inthe preceding four years had been 0.25 percentin 1968.

Finally, there is weak evidence that large bor-rowers from money-center banks as of August1970 (which would have included former com-met-cial paper issuer’s) received loans on rela-

tively favor-able terms. Available data on averageloan interest rates for the first two weeks ofMay and August 1970 by size of borrower andregion show that large, short-ter’m borrowers inNortheastern financial center-s exper’ienced thesmallest increase in lending rates over this peri-od (although differences are small). As Table 4shows, the largest classes of borrowers in NewYork City actually saw slight reductions in aver-age loan interest rates.

Other Fed Reactions to the (]rists

The discount window announcement tar-geting assistance to commercial paper’ issuerswas only the first of the Fed’s four policyresponses to the crisis. On Tuesday, June 23,the Fed suspended regulation Q ceilings onlarge-denomination hank CUs. ‘I’his alloweda Flood of money into the commercial banks,so that maturing commercial paper could hedirectly recycled through CDs, which financedbank loans to formner issuers. As shown inTable 3, from June 24 to July 15, large negotia-

Federal funds rate

Reserves supplyJune17

Reserves supplyJuly15

Reserves demandJuly 15

* Discount rate

June17

Borrowings - July 15 Total reserves

45

Table 3Banking System Changes During the Penn Central Crisis

Loans tobusiness

Borrowed reserves and finance Large negotiable U S. governmentFederal funds Total of large companies CDs at large securities heldrate minus borrowed commercial by large commercial by Federal

Date1 discount rate reserves banks commercial banks banks2 Reserve Banks

1970January 304 81,07 $ 807 $83,423 $10,444 655 568February 2 41 873 522 83549 10.839 55749March 145 1594 ‘1334 83903 11795 55621April 243 926 680 84122 13,022 56085May 164 979 675 83,265 12,984 57115June 3 164 1335 1,063 83545 12964 57698

0 198 834 624 83811 12956 5755217 1 80 459 273 85,785 12,741 57 82324 1 21 840 613 85331 12,949 57,005

July 1 1.23 923 871 87,212 14,118 57,7148 1.34 1598 1402 8716 15199 57,671

15 159 2036 1837 87590 15,980 5883922 115 1216 1044 87472 16311 58138

August 034 4044 941 86067 20,157 59618September 005 852 788 88426 22,227 60055October 0 11 418 341 86,514 23,546 69,283November —089 1 144 1,098 88,385 25201 61,209December 070 252 224 89 30 26075 60632

All data are end-of-month unless otherwise shown Dollar amounts are in millions2These are the sum of commercial and industrial loans by large commercial banks, and loans to personal and salesfinance companies, etcSOURCES Table I and Board of Governors of the Fede al Reserve System (1376) Table 4 18, Table 10.10

ble GUs at large commercial banks increased loans, directl~or indirectly, to “worthy” hor-from $13 billion to $16 billion, and the growth rowers who were other wise unable to securecontinued, with GUs of large banks in excess of credit. The Fed never made such loans because526 billion by the year’s end.” its other policies proved sufficient to contain

the i-un on commercial paper, but it is clear-The third policy intervention by the Fed was that the Fed was willing to provide direct lend-

open market operations. From June 17 to July ing if banks had been unwilling to make ap-is, total U.S. government securities held by the propriate loans for commercial paper rollovers.Fed increased fi-om 5.57.8 billion to $58.8 billion. In his statement to Congress on July 23, theAs noted above, however, open market opera- Chairman of the Board of Governors, Arthurtions were not sufficient to maintain the stock Burns, made this commitment clear. He viewedof nonborrowed reserves, given the increased the discount window as the key to preventing ademand for currency hy the public. Thus, bor- liquidity crisis, and saw direct lending by therowed reserves were relied upon as the primary Fed to firms in need, if necessary, as an ap-vehicle for’ expanding reserwes dum-ing the crisis. propriate fail-safe measure:

The Fed was also prepared to use “standby ci’eclit demands On the banking system at lam-ge

procedures” so that, if necessary, it could make can be accommodated h~’open mat-ket npen-ations,

“An unintended cost ot Regulation 0 was that it removed an“automatic stabilizer” from the tinancial system by makingit less attractive for investors to hold bank debt at times ofcrisis in other markets.

46

Table 4

Average Loan Rates onShort-Term Loans

Other Northeastern

New York City financial centers

Loan amount May August May August

Alt sizes 8 24% 824% 886% 8 89%$ 1,000 9,000 9.05 907 923 9.41

10,000- 99000 891 $95 934 942100,000-499,000 853 859 90 301500,000-999 000 8 31 823 8 72 8681 million and over 813 812 845 849

SOURCE Board of Governors of the Federal ReserveSystem (t976~Table I2SA

while the needs of individual banks can be metthrough the discount window...we have found,also, that minoi- adaptations of conventional mone-tary tools can provide solutions to special financialproblems.. it was made clear that the discountwindow would he made available to assist banksin meeting the needs of businesses unable tn rollover maturing commercial paper, and memberbank borrowings for this purpose subsequentlyhave risen. These conventional tools aie but-tressed with standby procedures to permit theFederal !~esemx’eto make funds ayailable to credit-worthy borrowers facing unusual liquidity needsthrough ‘conduit toans’-~-thatis, loans to a mem-ber hank to provide [minds needed for lending toa qualified hnrrower...Further-more, the FederalReserve could~undei-unusual and exigent cir-cumstances—utitize the limited power granted bx’the Federal Reserve Act to make direct loans tobusiness firms on the security of Government obli-gations or other eligible paper, provided the boi--i-ower is creditworthy but unable to secure creditfrom other sources.”

Here, Burns explicitly allows for Fed loansbacked by commercial paper or other eligiblecollateral.

in dealing with the Penn Central crisis, theFed did not simply focus on controlling themoney supply om an interest rate, which itcould have done easily through open mai-ketoperations. Rather the Fed coaxed deposits intobanks by relaxing Regtrlation Q ceilings, and

used the discount window to encourage banksto make loans to customers experiencing distress—especially cotnmercial paper issuers. The logicof the Fed’s combined approach was that mone-tary aggregates, bank cm-edit and assistance tothe commercial paper’ market could he targetedindependently by using three instruments.Relaxation of Regulation Q, rather- than expan-sionary open market operations, allowed hankcredit growth without (narrow) money growth.The discount window was directed toward thespecial difficulties in the commercial paper mar-ket. The Fed left open the possibility of lendingdirectly to firms in need if they were turneddown by bankers.

Evaluating Discount WIndow PolicyDuring the Crisis

it is not self-evident that the Fed’s policyresponse was correct. Schwartz (1992) has ar-gued that the Penn Central crisis was not a“real” financial crisis and that discount lendingserved no useful purpose. Of course, the ab-sence of a financial collapse in mid-1970 mayhave been attributable to Fed intervention itself,a possibility Schwartz does not take into ac-count. But even if Schwartz is too qtmick todismiss the potential seriousness of the PennCentral crisis—particularly given the evidenceon yield-spread movements and contraction ofthe volume of commercial paper—that does notprove that the discount window was a neces-sary instrutnent for dealing with the crisis. Ifthe failure of Penn Central increased doubtsabout the solvency of all firms in the economy,then a temporary expansion of open marketoperations or a Regulation Q m’elaxation—to in-crease the supply of credit available to all bor-rowers through relatively informed financialintermediaries—would have been a desirableresponse to an economy-wide need for liquidity,and there would have been no need to use thediscount window

On the other hand, if the crisis inyolyed aspecial reappraisal of the creditworthiness ofcommercial paper isstmers and commercial paperdealers in particular, and a reassessment of thedesirability of lending through the commem’cialpapet- market, then increasing the supply ofloanabie funds from banks may not have beenas effective as targeting temporary assistance (ashort-run suhsidy for bank loans to comnmercial

“See Burns (1970, pp. 624—5).

47

paper issuer’s) using the discount window as ameans to smooth issuers’ costs of rollover.” Inthis case, open market operations or- RegulationC) relaxation would have been a blunt instru-ment for- dealing with a run on commercialpaper per se, while discount window subsidiesfor the payoff of commercial paper would havepr’ovided targeted assistance without affect-ing monetary aggregates or interest rates onall traded assets. If some combination of aneconomy-wide reassessment of firms and a com-mercial paper run characterized the crisis, thenpolicy could have combined an aggregate in-crease in open market operations or RegulationQ relaxation with targeted assistance to com-memcial paper issuers.

Thus, to assess the desirability of the use ofthe discount window during the crisis, onemust examnine the incidence of the crisis acrossfirms. Was it purely an economy-wide phenome-non or did it pose a special threat to commer-cial paper issuers?

An Event Sturh~of the Penn Central

Crisis

To investigate the extent to which the PennCentral crisis posed a special threat to commer-cial paper issuers, I examine data on firmns’ahnormal stock returns during the crisis. Didfirms with outstanding commercial paper sufferabnormal negative returns relative to otherfirms duritig the onset of the crisis, and werethose negative retum’ns reversed by Fed inter’-vention? To answer this question, i combineCRSP data on daily stock retumns with Com-pustat data on annual income and balance sheetvariables for- nonfinancial compoiations to mea-

sure cross-sectional differences in abnormalreturns ovem’ var-ious dates, and to link them tofirm financial characteristics mneasured at thebeginning of 1970. I employ standar’d measuresof abnormal returns, using residuals from fore-casts of market returns hased on estimates offirms’ betas (from a 100-day pre-sample period)and the aggr’egate contemporaneous movementsin the market.

Specifically, consider- a standard model offirms’ stock returns, which decomposes returnsinto systematic and idiosyncratic factor’s:

(1) B~= a + h-B + e~,

where B measures returns, i indexes firms, Idenotes the date, and a and h are parameters tobe estimated. The error’ term e mneasures abnor-mal returns—the firm-specific, idiosyncraticdaily return at each date—or, in other words,the part of the stock return that is not fome-castable using the simultaneous aggregateretum’n for- the market and the firmns’ estimatedcorrelations with the market (h). Each firm’s bis estimated using observations on daily stockreturns for 100 trading days pm’ioi to the event(in this case, June 12).

Cumulative abnormal m’etur’ns over any “win-dow” are the accumulation of abnormal returnsfor each of the dates included in the window.Cumulative retur-ns generated from the aboveforecasting equation are “standardized” suchthat they can be inter-pm’eted to have heendrawn from a unit normal distribution.20 Thisadjustment m’esults in a cross-section of stan-dam’dized cumulative abnormal returns (SCARs)for each firm in the sample over- the eventwindow -

“The moral hazard costs of government pass-thnoughs wereminimal, since the banks, not the government, bore thedefault risk on the loans. This statement requires somequalification. It the pool ot borrowers laced large aggregatedetault risk, then bank failures might have resulted fromthe loans, in which case the government would have bornesome of the losses. Moreover, it some banks had been onthe brink of failure, they might have been willing to makesubsidized loans to the riskiest firms, thus concentratingoverall default risk and making the government’s indirectdefault risk greater. The central assumptions underlying myclaim that the government’s share of the risk was small arethat banks were not on the verge ot failure at the time, andthat the average quality of the commercial paper borrowerspool was high. The relaxation of Regulation 0 ceilings onCDs was also helpful in limiting the government’s risk,since it limited the amount of borrowing from the Fed. COsalso provided a natural vehicle for financing fixed-termcommercial paper, and did so without affecting the moneysupply.

20For details, see Wall and Peterson (1990).

4$

The event windows are defined as June 12-June 22 and June 23-July 9. Early concernsahout commercial paper issuers reported in theWall Street ,Journal date fromn June 12. June 22is the date aftem’ which Fed intervention shouldhave imnpr-oved the position of commercial paperissuers. By the second week of July, the con-traction in outstanding commercial paper beganto be reversed.

‘l’he goal of the event study is to examinewhether (likel ) commercial paper issuerssuffered abnormal negative stock returns dur-ing the Penn Cenim-al crisis, and whether- Fed in-tervention m’eversed those costs to commercial

paper issuer’s, after- controlling for other- meas-tires of cross-sectional differences among firms.‘to control for- other influences that wotrld nothave been specific to the commercial papermarket, I add a variety of balance sheet and in-come statemnent variables taken fr’om the Janu-ar-v financial reports of these nonfinancial firms.All firm halance sheet and income data ar-cmeasured as of the beginning of i970.z Thecontrol variables incltrded am-c: the ratio of debtto assets; the ratio of short-termn debt to assets;the size of the firm (market value of capital);the t’atio of net income to market value of capi-tal; the ratio of inventories to sales; and the

squares of each of these variables. These t’aria-bles are included to contr-ol for the possihilitvthat the share prices of firms with high ex-posur-e to macroeconomic shocks (fir-ms withhigh levem’age, or with large financing needsrelative to sales) niav have m’esponded mor-estrongly to economic news, irrespective ofwhether or not they were commer’cial paper’isstiers. For example, if Penn Central’s failureincreased the cost of debt for- all firms, thenleverage ratios or inventor-to-sales ratios wouldidentify ci-oss-sectional differences in SCARs.

tsolating the effect on SCARs of reliance onthe commercial paper- market is not stm’aightfor-ward, since data on otrtstanding commercial

paper issues of firmns are not availahle for thisperiod. ‘l’he regular’ reporting of commercial

paper ratings was largely a consequence of thePenn Central crisis. Standard and Poor’s beganpublishing some commnercial pitper’ m-alings in

The Bond Outlook in July 1970, but these r’at-

ings wet-c for only a handful of issuers, most ofwhich were financial firms. Moodv~sIndustrialManual and other’ similar publications, whichtoday provide some data on commercial papem-issues by firms, did not provide such data in1970. Outstanding commercial paper cannot lieinferred by looking at firms’ reported balancesheets. Commemcial paper can appear in firmbalance sheets either as long-tem’mn tir short-termdebt. Although it is usually included in short-term debt, even in that case it cannot he sepa-rated fm’om other short-term deht (loans frombanks, finance companies, amid so on). TimeBoard of Governors of the Federal Reserve Sys-tern did not collect fit-m-level data on issuei-s,only on aggregate amounts of outstanding is-stres, based on dealers’ reports. Despite searchesof various publications by the rating agencies, Ihave heen unable to uncover any comprehen-sive listing of fit-ms which issued commercialpaper in 1970.

Given the lack of data identifying issuers, Iuse bond ratings to sort firms according towhether they were likely to have issued corn-mercial paper in 1970. In the 1970s, commercialpaper issuance was usually restr-icted to thefirms with the highest bond ratings (Standardand Poor’s, 1979, p. 47). 1-laying a AA or ~~A1\rating in 1970 is likely to be the best proxy forthe likelihood of being a commem’cial paper is-suer. Eight of the 11 nonfinancial firms whoseratings wer’e published in Standar-d and Poor’sflood Outlook in 1970 and 1971 were m’ated AAor AAA (the remnainder were A-rated). Also, datafrom latet- years indicate a close relationship he-tween high bond ratings and commercial papem-access. Standard and Poor’s first comprehensivelisting of rated commercial paper issuers, TheCommercial Paper Ratings Guide, was publishedin 1978. Of the 90 nonfinancial fitmns that hadAA or AAA bond ratings in 1970, 64 were issu-ing commercial liaper in 1978. Of the 146 non-financial firms listed in Compustat with AA om’AAA bond ratings in 1978, 93 were commercialpaper issuers. In 1978, 94 of the 207 A-ratednonfinancial fir-ms in Comptmstat wem’e commer-cial paper issuers, and only 43 firms with bondratings below A issued commer’cial paper (all ofthese were firms with BBB or BB ratings). Usingthe AA rating as our cutoff, therefore, seems

2lThis was dictated by the superior data available on theannual Compustat tape. Quarterly Compustat data for thisperiod are often incomplete.

49

advisable. Based on available data, it seemsreasonable to assume that a majority of AA orAAA nonfinancial firms were commercial paperissuers in 1970, and that a much smaller per-centage of remaining firms were issuers.~~Thetotal number of mionfinancial firms in our sam-ple (that is, those without missing observations,and covered by both CRSP and Compustat in1970) is 1,482. Of these, 90 had bond ratings ofAA or AAA.

If commercial paper issuers experienced aspecial problem during the crisis, and if Fed in-tervention reversed the strain on issuers, thecoefficient on the high-rating indicator variableshould be negative during the onset of the crisisand positive after Fed intervention. The use ofAA or AAA bond ratings as an indicator of acommercial paper issuer provides a “conserva-tive” measure of the problems in the commer-cial paper market, for three reasons. First,measurement error (the existence of someA-rated commercial paper issuers, and of non-issuers with AA or AAA ratings) biases thecoefficients on the high-rating indicator variabletoward zero. Second, the excluded A-rated com-mercial paper issuers likely would have ex-perienced the largest adverse effects of thecrisis, since their debt was riskier to begin with.‘I’hird, the flight to quality during a financialcrisis should produce a relative improvement inthe value of high-rated firms, which would im-ply positive effects on AA and AAA firms, aftercontrolling for other firm characteristics, duringthe onset of the crisis.

Table 5 reports regression results for SCARsfor two windows around the Penn Centralcrisis—Jumie 12 to June 22, and June 23 to July923 It is important to emphasize three pointsbefom’e reviewing Table 5. First, coefficients onthe control variables in this regression must beinterpreted cautiously. For example, while rela-tively high leverage m’atios may have created

problems for firms during the crisis, high debtratios may themselves have been associatedwith firm attributes (like creditworthiness) thathelped firms weather the crisis better (and ledto relatively higher stock values). Thus, it is notpossible to infer “structural” relationships fromthese cross-sectional findings. The main point ofincluding the control vam’iables is to separate theeffect of commercial paper issuance pet- se fromfactors unrelated to commercial paper issuance.Second, the abnormal returns measures arepurged of cross-sectional differences in firms’betas that might be correlated with the variousregressors. For example, higher debt ratiosmight be associated with lower returns cross-sectionally because leverage increases a firm’sbeta. But, by construction, the abnormal returnsused in Table 5 are uncorrelated with the firm’sbeta. Third, squared terms were added for allregressors, but they do not affect the directionof the results. in no case does a squared termmore than offset the linear effect of the samevariable when both coefficients are evaluated atthe mean of the regressor (given in Table 6).The direction of association between SCAR andany regressor is that of the linear effect.

‘The results reported in Table 5 indicate thatthe ratio of debt to assets and the ratio of in-come to net worth (both measured at the begin-ning of the year) may have been associated withmore negative returns cross-sectionally duringthe onset of the crisis. Firm size per se had noeffect on returns in the presence of squaredterms for debt ratios. For the period after June22, the total debt ratio and the profit ratio areassociated with a positive effect on returns,indicating a reversal of the stock price move-ments dum’ing the period prior to Fed inter-vention. The inventory-to-sales ratio and theshort-term debt-to-assets ratio are both nega-tively associated with abnormal returns afterJune 22.

221t is less clear whether the data on A-rated firms in 1978 isrepresentative of A-rated firms in 1970. From 1970 to 1978,market analysts argue that the growth in commercial paperissuers brought more firms with lower ratings (A or EBB)into the market; thus, it might not be appropriate to as-sume that 1970 saw the same high proportion of A-ratedfirms issuing paper as in 1978 (45 percent). For purposesof constructing an indicator variable, given the uncertaintyabout the number of issuers with A ratings in 1970, it isbest to exclude A-rated firms from the group of likely is-suers because A-mated firms are a small fraction of totalfirms with ratings below AA, but a large fraction of AA orAAA firms.

237he results reported below are not sensitive to whetherJune 22—which arguably could have been included in thesecond window—is included or excluded from either win-dow. The results of the first period are driven by pre-June22 returns, and the results of the second window aredriven by post-June 22 returns.

50

Tabl 5

Event Study Regçes ion R suits forStandardized Abnormal 13e urns(standarçt ~ IP p~eflth*Ø~),.

611218 6/2 /7~ 6/23/7~- SITU(1) ~(3J (4)

tnte~cept 0~S1 042 5 39(9> 1 iSP) 4

Detlt(A8s 051 068 G~ OS

(0-551 (0’5&} (060> (060)(W&sq~ 06 00 00 028

p74) 01-4 p 0$>

1 080 4 198S OR

(TOM q 51 S 08

san> ~30) I~o~we toG 0 2 *29

2> a) 0(MVE)-sq OS tOt 00 0*~

(01 IG~1 0) )MMVE OS $82 IS 1

) (0 } ( ) -asM/MVE$q 022 80 84 0 0

40 040 1) 47)1PNISM-E 0-89 q 1 55

040) (81 4044(tN~$AlESsq 004 0~04 078 080

041) 41) (045 4045)

Ma MA 030 0640 }‘ 018)

Ad ft-sq red 0*40 0042 0080 0089

\fter controlling for observed balance sheetand income characteristics firms w mlii ~ ~ orAAA bond ratmngs expem’menced significant, negathe abnormnal returns during the onset of thecrisis and signific rnt, positive m’etur n after Fedmntem ~ention. The addition of thi mdii ator t an-able increases the adjusted H squared in bothregressions ‘the evidence prot rdes stmpport fomthe notion that in addition to the economy-wideliqumditv crisis durmng the Penn ( entmal crisiscomniercial paper issuers faced a special prob-1cm. this, in turn, lends support to the argu-ment that discount wmndoti subsidizatmon oflending mnay haxe been useful mn targeting

assistance to the commercial paper market.Thus, the Fed may have been correct to dividepolicy into two components: Regulation Q relax-ation to provide liquidity to all firms throughbanks, and discount window lending to targetsubsidized assistance to commercial paper is-suer’s to offset the special (lisorder in that mar-ket. That is not to say Fed policy achieved theright mix. For example, negative returns forfirms with high inventory-to-sales ratios or’ highshort-term debt after June 22 may indicate thatcredit supply was too tight overall.

Changes in the commercial .PaperMarket After the Crisis

The commercial paper market changed as aresult of the Penn Central cmtsis. In addition toincreased diligence in evaluating cm-edit risk, twoother changes have reduced the possibility of asimilam’ problem in the future. First, in Augustof 1970, the Fed passed a regulation to restrictthe growth of bank commercial paper. Bankpaper would lie treated, for reserve require-ment purposes, the same way as demand omtime deposits, depending on the mattmritv of the

paper. i’his eliminated the advantages of off-balance sheet financing through hank comtner--cial papem and led to the contm-actiomi of hankpaper. This had little effect on banks or on thegrowth of the commemcial paper’ mam’ket, whichhas been robust (Post, 19921. It simply propelledhamiks toward relying on negotiable CDs (virtual-ly identical to conimercial papet-) as an alterna-tive source of funds.

Of greater imnportance were institutionalchanges in the way commertral paper is mar-keted. First, mating agencies made finer distinc-tions in their ratings of commercial ~iaper issues(Stigum, 1983, p. 637). An important element inthe rating hecame evidence of commercial bankbackup arrangements hehind commercial palierprograms. Commercial hank support for com-mercial paper- programs was a private innova-tion. After, and largely as a result of PennCentm’al, commercial paper issuers increasinglysought “hurricane insurance” in the form ofbackup loan commitments (Stigum, 1983, pp.633-4; Standard and Poor’s, 1979, p. 47). Most ofthese loan commitments (m-oughly 85 percent in1989) are not credit guarantees to commercialpaper’ holders, but rather promises for as-sistance during a general liquidity crisis if theborrower remains creditworthy (Calomiris,1989b). Within a few yeam’s of the Penn Central

51

Table 6Means, Standard Deviations, and Correlations AmongRegressors

Standard

Mean deviations Correlations (p-values in parentheses)

STDIA ‘VlVE Nl/MVE iS AA-

0/A 028 0.21 052 011 028 005 011ti 000 (0 000) 10.000) (0 05) (0 000)

STO/A 007 009 —- —028 0.06 026 —0.0810 000( to 031 0 00W 10 003

MVE 112 1.6 -- — —003 —013 0.32tO 221- to ooo; 0 00W

NUMVE 0.17 013 — . 013 007(0 000) (0 004)

I/S 017 013 —. . -- 01910. 000)

AA~ 006 023 — — —.

crisis, backup lines were almost always 100 per-cent of outstanding issues, except for large, top-rated, highly liquid issuers like GMAC or largeconmmetcial banks. These loan commitmentswere issued by banks for the same reason bankassistance had been relied on during the PennCentral crisis: Banks have access to the discountwindow and believe that they can rely on theFed (which maintains no official policy in thisregard) to temporarily suspend normal non-pecuniary discount window penalties to grantlending subsidies during an emergency. Institu-tionalizing Fed discount window protectionthrough explicit bank loan commitments, onecould amgue, reduces the time to pmocess cm’editrollovet during a ctisis. Furthermore, the exis-tence of clear’ comnmitmermts to lend duming acrisis may itself reduce the threat of a generalliquidity squeeze and, thus, make crises lesslikely.

Currently, the use of backup lines of bankcredit, “backed” by access to the discount win-dow, has virtually eliminated risk of anotherPenn Central cm-isis in the commercial papemmarket. But this does not imply an end to them’ole played by the discount window. ‘l’he pm’o-tection offered through backup lines of creditdepends on banks’ potential access to fundsthrough the discount window.

EVALUATING OTHER POSSIBLE

FED INTERVENTIONS

Thus far, I have argued that both economy-wide policy (open mnarket opetations and Regu-lation Q relaxation) and targeted discount lend-ing may have been desirable interventionsduring the Penn Central crisis. But the Fed waswilling to go beyond these interventions, ifnecessary, as Chairman Burns’ comments citedabove indicate. Was the Fed might to haveprovided for the possibility of dimect lending tofirms, or should it have been willing to relyonly on the discount window and open marketoperations? Was the Fed right to have allowedPenn Central to fail in the first place?

The Fed’s decision not to prevent the failureof Penn Centm’al is easy to defend. The successof the capitalist system requires that firms face

<‘hard” budget constraints. As reformers inEastern Europe and the Soviet Union have beensaying for years, protecting lange corporationsfiom bankruptcy through assistance from thestate imnposes lam-ge costs on more successfulgrowing enterprises. More fundamentally, allow-ing corporations to fail is what encourages themto succeed. it is worth emphasizing that thepublic policy rationale for insulating financialmarkets from temporary information externali-

52

ties during panics does not in any way justifybailing out discernably insolvent institutions.

With regard to the other question—whetherdirect Fed lending to corporations is ever justifi-able—it is also hand to justify intervention. AsMishkin (1991b) notes, it is better to decentral-ize the decision over who receives how much,and place it in the hands of relatively informedbankers who have incentives to avoid makingbad loans. If banks had been unwilling tofinance the payoff of the commercial paper ofcertain firms, even on highly subsidized terms,that would have indicated the likely insolvencyof those individual issuers.~~Discount window’protection should not be used to save individualfirms which are viewed as insolvent by theircreditors. Of course, creditors are not alwaysright, but pam-t of the rationale for corporate re-organization under bankruptcy law (increasinglypopular since the 1978 changes in the bankrupt-cy code) is to minimize unnecessary costs of li-quidating defaulting firms who turn out to besolvent. Given the availability of the reorganiza-tion option, it may be best for the governmentto allow private markets to decide whether in-dividual corporate borrowers am-c viable.

COULD A SIMILAR CRISISHAPPEN TODAY?

Although I have argued that the possibility ofanother Penn Central crisis today in the com-

mercial paper market is remote, in other newand growing financial markets the potential fora crisis similar to Penn Central may loom larg-er.25 For example, within the banking system alarge overdraft default in the Clearing House In-terhank Payments System (CHIPS) might lead toa general run of uninsured liabilities of CHIPSmembers, due to problems of unraveling whichbanks stood to lose from the default. Subsidizedlending to CHIPS members might be warrantedto prevent a panic.2” The Fed is cognizant of itspotential role in assisting banks in the event of asystemic crisis in the payments system, and itregulates the payments system accordingly. Likemany other central banks, the Fed limits over-drafts of bank accounts with the central bankand requires private bank clearing systems tolimit overdrafts among their members. Suchlimits include collateral requirements, quantitylimits on overdrafts, and pre-established loss-sharing arrangements. These regulations aremeant to ensure that the potential ptotection af-forded by the Fed is not abused.

It is also conceivable that discount window in-tervention could be used to target assistance tomarkets for financial derivatives. In the swapmarket, for example, if a major swap providerbecame insolvent, its counterparties, and thirdparties who have contracted with those coun-terparties, could experience unpredictablechanges in their market misk exposures and,consequently, in their default risks. Because ofthe intermelatedness of the vamious positions and

24sf course, the Fed could have done even more to en-courage banks to make pass-throughs than it did duringPenn Central by making its subsidy larger. The subsidythat the government can grant is potentially very large. Bylowering the discount rate to zero and discriminating inimposing nonpecuniary penalties across banks (for exam-ple, charging a zero hassle cost to banks borrowing for tar-geted pass-throughs and a prohibitive rate on otherborrowing), the subsidy can be increased to the level ofthe equilibrium fed funds rate without affecting monetarycontrol.

t5Gorton and Pennacchi (1992) argue that there is no evi-dence for “contagion” among commercial paper issuers orfinance companies. They examined the failures of severalissuers and finance companies and found that a failure didnot lead investors in securities markets to lower the priceof other issuers’ or finance companies’ securities, ceterisparibus. It is premature, however, to interpret this as evi-dence that issuers or finance companies are now immuneto panics, or more broadly, that financial technology hasimproved so much that intermediaries are not potentiallyvulnerable to panics. Gorton and Pennacchi’s sample ofevents is small, and the events they examine may simplyhave been transparently idiosyncratic (unlike, for example,the Penn Central crisis). It is possible that events unlikethose in their sample could produce panics.

260f course, the discount window is not the only way to dealwith such a problem. Alternatively, deposit insurance couldbe extended to the CHIPS clearinghouse as a whole. Forexample, the government could offer insurance to CHIPSwith a large deductible, with the liability for the deductibleshared by all clearing members.

ncg,c~n ~flMW nc CT fl~flC

53

uncertainty as to which swap contracts will sur-vive the crisis, it might be difficult for counter-parties to gauge their true exposure to mnarketrisk. ‘this could produce a flight to cash by allparties. Furthermore, a reversal of market opin-ion about the reliability of swaps as hedgingdevices could suddenly affect the market’s per-ception of firms with lam-ge swap positions. Inthis case, temporary disruptions to the supplyof credit to certain classes of firms could con-ceivably result. These problems could motivatediscount window subsidies as in the Penn Cen-tral crisis.

The lesson in this dismal scenario is not thatswaps are a bad idea. They offer real long-runsystemic risk reduction as a low-cost vehicle tohedge interest rate risk. But reaping the advan-tages of this and other financial innovationsrequires a period of learning about how tomeasure and control the risks created by newfinancial instruments. The existence of the dis-count window provides a safety valve to protectthe financial system from growing pains like theones it suffered in 1970. Recent financial inno-vations in derivatives and asset secunitizationmay have increased the need for the discountwindow as an instrument of public policy. Itsrole is not just to protect the banking systemfrom systemic runs on commercial banks (in-deed, it may have little importance here in thepresence of deposit insurance); rather, its role isto effect occasional, contingent and focusedcredit subsidies to particular markets throughbanks during moments of tempom-ary disruption,like that of the Penn Central crisis.

Another example of a potential application ofthe discount window is a run on a futuresclearing house. Individual clearing membersstand between all contracting parties and theclearing house provides mutual insurance amongall members against default. To limit the risk ofdefault by clearing members, clearing housesimpose reserve r-equirements in the form ofcash or Treasury bills on open positions andfrequently monitor those positions. Still, it isconceivable that a very large, sudden price drop(say, in the stock market) might bankrupt acleam-ing member with a large open position andconceivably threaten (he clearing house. ‘i’hiscould cause a run on the futures market asholders of contracts, wary of the credibility ofthe clearing house’s solvency, try to sell theircontracts. ‘i’his could amplify the losses to theclearing house and legitimize the initial con-

cerns that prompted the run, leading to furthercashing-in of positions. if the clearing housewere to fail, many hedges would disappear withits demise, increasing the risk of many financialclaims and causing confusion about the inci-dence of the increased risk in ways that mightprovoke a general liquidity crisis.

The Fed could reduce the chance of a run ona futures clearing house and the negative exter-nalities attendant to such a run by agreeingtemporarily to lend through the discount win-dow without penalty to banks making loans toclearing house members, and could even lowerthe discount rate if necessary to encourage suchsubsidies. Indeed, this seems a reasonablecharacterization of the Fed’s response to con-cerns about futures markets posed by the stockmarket collapse of October 1987.

There is a more difficult policy question,however, that so far has not been addressed.if banks are unwilling to lend to a clearinghouse—even on highly subsidized terms—shouldthe Fed let the clearing house fail? On onehand, ad hoc direct lending by the Fed runs therisk of encouraging lax self-regulation withinthe clearing house. On the other hand, thefinancial disruption from a clearing housefailure might generate substantial negativeexternalities in the financial system. it mightbe desirable for the Fed to decide whether itwould stand behind the liabilities of failed fu-tures clearing houses. lf so, the Fed should con-sider whether existing pmivate risk-managementdevices (like margin rules) are adequate. If not,it might recommend changes to the CommodityFutures Trading Commission, which regulatesthese exchanges. As the volume of derivativetransactions expands, so does the need to de-velop policies for dealing with possible systemicrisks related to these markets.

Identifying a potential benefit from a “backup”discount window does not justify the currentform of the discount window. There may be nobenefit from Fed lending to banks during nor-mal times, and as Schwartz and others have ar-gued, such lending may be costly. There alsoremains the risk that government agencies willabuse even a “reformed” discount window bydefining noncm’ises as crises to make loans to fa-vor-ed parties. The evidence presented in thispapem, therefore, does not prove that the dis-count window has been a net benefit as a poli-cy tool, only that it has the potential to providebenefits as well as costs.

54

REFERENCES

Altman, Edward I. ‘Financial Ratios, Discriminant Analysisand the Prediction of Corporate Bankruptcy;’ Journal ofFinance (September 1968), pp. 589—609.

Board of Governors of the Federal Reserve System. Bankingand Monetary Statistics: 1941-1970. Board of Governors ofthe Federal Reserve System, 1976.

Bordo, Michael D. “The Lender of Last Resort: AlternativeViews and Historical Experience7 Federal Reserve Bank ofRichmond Economic Review (January/February 1990), pp.18—29.

Brimmer, Andrew F. “Distinguished Lecture on Economics inGovernment: Central Banking and Systemic Risks inCapital Markets;’ Journal of Economic Perspectives(spring 1989), pp. 3—16.

Burns, Arthur F. “Statement to Congress;’ Federal ReserveBulletin (August 1970), pp. 619—26.

Business Woek. “Can Stocks Shake the Liquidity Jitters?”June 27, 1970, p. 42.

Calomiris, Charles W. “Regulation, Industrial Structure, andInstability in U.S. Banking: An Historical Perspective;’ inMichael Klausner, and Lawrence J. White, eds., StructuralChange in Banking. Business One Irwin, 1992a, pp. 19—116.

_______- “Getting the Incentives Right in the CurrentDeposit-Insurance System: Successesfrom the Pre-FDICEra;’ in James R. Barth, and R. Dan Brumbaugh, Jr., eds.,The Reform of Federal Deposit Insurance: Disciplining theGovernmentand Protecting Taxpayers. Harper Business,1992b, pp. 13—35.

_______ - “Do ‘Vulnerable’ Economies Need Deposit Insur-ance?: Lessons from U.S. Agriculture in the 1920s,” in Phil-lip L. Brock, ed., If Texas Wore Chile: A Primer on BankingReform. CS Press, 1992c, pp. 237—314.

_______- “Is Deposit Insurance Necessary? A Historical Per-spective;’ Journal of Economic History (June 1990), pp.283—95.

_______- “Deposit Insurance: Lessons from the Record,”Federal Reserve Bank of Chicago Economic Perspectives(May/June 1989a), pp. 10—30.

_______- “The Motivations for Loan Commitments BackingCommercial Paper;’ Journal of Banking and Finance (May1989b), pp. 271—7.

_______ and Gary Gorton. “The Origins of Banking Panics:Models, Facts, and Bank Regulation;’ in R. Glenn Hub-bard, ed., Financial Markets and Financial Crises. Universityof Chicago Press, 1991, pp. 109—74.

_______ Charles P. Himmelberg, and Paul Wachtel. “Com-mercial Paper and Corporate Finance: A MicroeconomicPerspective;’ Carnegie-Rochester Series (forthcoming,1994).

_______ and R. Glenn Hubbard. “Price Flexibility, CreditAvailability, and Economic Fluctuations: Evidence from theUnited States, 1894-1909;’ Quarterly Journal of Economics(August 1989), pp. 429—52.

_______ and Charles M. Kahn. “The Role of DemandableDebt in Structuring Optimal Banking Arrangements;’ TheAmerican Economic Review (June 1991), pp. 497—513.

_______ and ______- “The Efficiency of Cooperative Inter-bank Relations: The Suffolk System;’ University of IllinoisWorking Paper (1990).

_______ and Larry Schweikart. “The Panic of 1857: Origins,Transmission, and Containment;’ Journal of Economic His-tory (December 1991), pp. 807—34.

Cummins, Claudia. “Fed Proposes Curbing Access to Ad-vances by Troubled Banks,” The American Banker (August12, 1993), p. 2.

Diamond, Douglas W., and Philip H. Dybvig. “Bank Runs,Deposit Insurance, and Liquidity;’ Journal of Political Econ-omy (June 1983), pp. 401—19.

Friedman, Milton, and Anna J. Schwartz. A Monetary Historyof the United States: 1867-1960. Princeton University Press,1963.

Goodfriend, Marvin, and Robert 0. King. “Financial Deregu-lation, Monetary Policy, and Central Banking;’ in William S.Haraf, and Rose Marie Kushmeider, eds., RestructuringBanking and Financial Services in America. American En-terprise Institute, 1988, pp. 216—53.

Gorton, Gary. “Self-Regulating Bank Coalitions;’ University ofPennsylvania Working Paper (March 1989).

_______- “Clearing Houses and the Origin of Central Bank-ing in the US” Joumal of Economic History (June 1985),pp. 277—83.

_______ and George Pennacchi. “Money Market Funds andFinance Companies: Are They the Banks of the Future?;’in Michael Klausner, and Lawrence J. White, eds., Structur-al Change in Banking. Business One Irwin, 1992,pp. 173—214.

James, John A. “The Rise and Fall of the Commercial PaperMarket, 1900-19307 Paper presented at Anglo-AmericanFinance: Financial Markets and Institutions in 20th-CenturyNorth America and the U.K., a conference held at NewYork University, New York, NY., December 10, 1993.

Janssen, Richard F., and Charles N. Stabler. “Empty Coffers:Cash Shortage Causes Worry that Big Firms Could Col-lapse;’ Wall Street Journal, June 12, 1970, p. 1.

Kane, Edward J. “How Incentive-Incompatible Deposit Insur-ance Funds Fail;’ Ohio State University Working Paper Na88-35 (1988).

Kaufman, George 0. “Lender of Last Resort, Too Large toFail, and Deposit-Insurance Reform;’ in James R. Barth,and R. Dan Brumbaugh, Jr., eds., The Reform of FederalDeposit Insurance: Disciplining the Government and Protect-ing the Taxpayer Harper Business, 1992, pp. 246—St

_______- “Lender of Last Resort: A Contemporary Perspec-tive;’ Journal of Financial Services Research (October 1991),pp. 95—110.

Maisel, Sherman J. Managing the DolIar Norton & Co., 1973.

McLean, David. “Abreast of the Market;’ Wall Street Journal,June 25, 1970, p. 29.

- “Abreast of the Market;’ Wall Street Journal,June 24, 1970, p. 31.

______- “Abreast of the Market;’ Wall Street Joumal,June 23, 1970, p. 33.

Melton, William C. Inside the Fed: Making Monetary PolicyDow Jones-Irwin, 1985.

Miron, Jeffrey A. “Financial Panics, the Seasonality of theNominal Interest Rate, and the Founding of the Fed;’ TheAmerican Economic Review (March 1986), pp. 125—40.

Mishkin, Frederic S. “Asymmetric Information and FinancialCrises: A Historical Perspective;’ in R. Glenn Hubbard,ed., Financial Markets and Financial Crises. University ofChicago Press, 1991a, pp. 69—108.

“Anatomy of a Financial Crisis;’ NBER WorkingPaper No. 3934 (December 1991b).

Murray, Roger F. “Lessons for Financial Analysis: The PennCentral Debacle;’ Journal of Commercial Bank Lending(December 1971), pp. 42—9.

55

Post, Mitchell A. “The Evolution of the U.S. CommercialPaper Market Since 1980,” Federal Reserve Bulletin(December 1992), pp. 879—91.

Schadrack, Frederick C., and Frederick S. Breimyer. “RecentDevelopments in the Commercial Paper Market,” FederalReserve Bank of New York Monthly Review (December1970), pp. 280—91 -

Schwartz, Anna J. “The Misuse of the Fed’s Discount Win-dow;’ this Review (September/October 1992), pp. 58—69.

Standard and Poor’s. Standard and Poor’s Ratings Guide.McGraw-Hill, 1979.

_______- Commercial Paper Ratings Guide. Standard andPoor’s Corp., February 1978.

_______ The Bond Outlook. Standard and Poor’s Corp.,various issues.

Stigum, Marcia. The Money Market. Dow Jones-Irwin, 1983.

Timlen, Thomas M. “Commercial Paper — Penn Central andOthers;’ in Edward J. Altman, and Arnold W. Sametz, eds.,Financial Crises: Institutions and Markets in a Fragile En-vironment. John Wiley and Sons, 1977.

Treiber, William F. “Problems of Financial Community UnderConstant Scrutiny7 The American Banker (October 13,1970), p. 16.

Wall, Larry D., and David R. Peterson. “The Effect of Con-tinental Illinois’ Failure on the Financial Performance ofOther Banks;’ Journal of Monetary Economics (August1990), pp. 77—99.

Wall StreetJoumal. “Patman Asks to See Pentagon’s Dossieron Penn Central Before Any Aid is Given,” June 19, 1970,p. 5.

___________- “Three Outside Directors Quit Penn Central,Rail Unit; Bank Conflict May Affect Others;’ June 15, 1970,p.7.

White, Eugene N - The Regulation and Reform of the AmericanBanking System, 1900-1929. Princeton University Press,1983.

Whitford, David. “Financial Ratios and the Prediction ofBankruptcy The Case of Penn Central7 University ofIllinois Working Paper (1993).

Yang, Catherine, and Mike McNamee. “Reform, Or A Crack-down On Banking?,” Business Week (July 15, 1991),pp. 122—3.