NBER WORKING PAPER SERIESFINANCIAL CRISES AND LIQUIDITY SHOCKS:A BANK-RUN PERSPECTIVEGuillermo A. CalvoWorking Paper 15425http://www.nber.org/papers/w15425NATIONAL BUREAU OF ECONOMIC RESEARCH1050 Massachusetts AvenueCambridge, MA 02138October 2009I am grateful to Alejandro Izquierdo, Ivan Khotulev and Enrique Mendoza for useful comments. Theviews expressed herein are those of the author(s) and do not necessarily reflect the views of the NationalBureau of Economic Research.© 2009 by Guillermo A. Calvo. All rights reserved. Short sections of text, not to exceed two paragraphs,may be quoted without explicit permission provided that full credit, including © notice, is given tothe source.FINANCIAL CRISES AND LIQUIDITY SHOCKS: A Bank-Run PerspectiveGuillermo A. CalvoNBER Working Paper No. 15425October 2009JEL No. E5,E58,F41,G2ABSTRACTThis note is motivated by trying to understand the macroeconomic implications of assuming that periodsof financial bonanza and turmoil are driven by financial innovation and collapse in line with the “bankrun” literature of the Diamond-Dybvig (1983) variety. Bypassing a host of important but, for the presentpurposes, secondary details the note assumes that the initial effects of financial innovation and crashcan be summarized by a parameter that determines the “liquidity” or “moneyness” of land or capital.This simplification helps to shed light on some issues that are at the center of the policy debate. Inparticular, one can show that preventing price deflation is not enough to offset asset meltdown. Furthermore,lower policy interest rates increase asset prices and steady-state output which, however, gets reversedas liquidity is destroyed. An interesting result is that, in the neighborhood of a first-best capital allocation,an increase in the moneyness of capital may lower the welfare of the representative individual, evenif the higher liquidity of capital is sustainable and, hence, not destroyed by future crash. Moreover,an extension of the basic model supports the conjecture that low policy interest rates may have givenincentives to the development of “shadow banking.”Guillermo A. CalvoColumbia UniversitySchool of International and Public Affairs420 West 118th St, Room 1303BMC3332New York, NY 10027and [email protected]. Introduction Few would deny that we are living in extraordinary times. Where opinions start to differ is about what is behind current events. The popular press, not surprisingly, stresses fraud and crony capitalism, while better informed observers blame inadequate financial supervision. Both have a share of the truth. However, these opinions may miss a central fact, namely, that financial systems are, as a general rule, prone to eventual collapse if they don’t have a “lender of last resort,” LOLR. Diamond and Dybvig (1983) made that point in terms of a simple and elegant banking model (for a recent exposition, see Diamond (2007)). Thus, under this optic the present financial troubles could be traced to the existence of a “shadow” banking system that developed without enough supervision and, critically, without a LOLR. This is the main conjecture that inspires the present note.1 A second conjecture is that the shadow system succeeded in “printing money” through devices like collateralized debt obligations, CDOs. This is formalized by assuming that financial innovation succeeded in making land or capital more “liquid,” e.g., by making them more useful as a means of payment. These assumptions or conjectures are then employed to trace the general equilibrium implications of an increase or collapse in land’s or capital’s liquidity – the central comparative-dynamic exercise – in a rational expectations setup.2 There are many models in which capital market imperfections contribute to make business cycle fluctuations deeper and more persistent (see, for example, Bernanke and Gertler (1989) and Kiyotaki and Moore (1997)). This literature stresses agency problems 1 See Brunnermeier (2008) for a highly didactic exposition of the Subprime crisis that supports the view that the financial collapse has Diamond-Dybvig characteristics. 2 Rational expectations may look like an awkward vessel to encapsulate the discussion. However, in this paper liquidity crises are taken as exogenous and unanticipated. Thus, our agents will be rational but ignorant of the central driving force, i.e., the sudden demise of liquidity-enhancing financial instruments.2in which first-best credit allocation is not achieved because of, for instance, asymmetric information of debtors and creditors. A financial crisis of the kind we are experiencing could be modeled in that context by assuming that there is a sudden deterioration in agency problems. There is no doubt that the Subprime crisis itself has contributed to worsening the creditworthiness of individuals and firms but it would be hard to argue that the initial kick was given by a sudden deterioration of agency problems. Financial institutions increasingly engaged in non-transparent activities but this did not occur overnight; these activities took years to reach pre-crisis characteristics. Thus, available models fail to meet a central analytical challenge for rationalizing the current crisis, namely, developing a case in which behavior discontinuity could be justified in an intuitively plausible manner.3 The challenge is better met by the model discussed in this paper, because the Diamond-Dybvig bank-run story which underlies the model displays equilibrium multiplicity (in absence of a LOLR).4 A financial crisis could erupt just because people expect that other people expect that financial instruments issued by shadow banks will be bereft of liquidity. This equilibrium fragility is not obtained by superimposing an ad-hoc assumption to the original model (as it would be the case in the asymmetric-information approach),5 but as a result of an essential characteristic of the underlying bank-run model. No real shock is needed for crisis to happen in this context. Thus, the model provides a rationale for the fact that the triggers behind big-time financial crises in the 30s and late 80s – and even the Subprime crisis’ global 3 This challenge is also faced by the literature on Sudden Stops in emerging capital markets (see Calvo (2007). The challenge is less critical, however, because the initial kick took place outside most of emerging