Understanding Financial Crises Franklin Allen and Douglas Gale Clarendon Lectures in Finance June 9-11, 2003 1 Lecture 2 Currency Crises Franklin Allen University of Pennsylvania June 10, 2003 http://finance.wharton.upenn.edu/~allenf/ 2 Introduction Major theme of the banking crises literature Central bank/government intervention is necessary to prevent crises From 1945-1971 banking crises were eliminated but currency crises were not 3 Many of the currency crises were due to inconsistent government macroeconomic policies Explanations of currency crises are based on government mismanagement Contrasts with banking literature where central banks/government are the solution not the problem 4 First generation models Krugman (1979) and Flood and Garber (1984) show how a fixed exchange rate plus a government budget deficit leads to a currency crisis Designed to explain currency crises like that in Mexico 1973-82 5 Salant and Henderson (1978): Model to understand government attempts to peg the price of gold Market Solution: Earn r on gold holdings P(t) = P(0) ert Ln P(t) = Ln P(0) + rt 6 Ln P(t) Ln Pc Ln P(0) T t 7 If the government pegs price at P*, what does the price path look like? Ln P(t) Ln Pc Ln P* T t Can’t be an equilibrium because of arbitrage opportunity 8 Ln P(t) Ln Pc Ln P* T’ T t Equilibrium: Peg until T’ then there is a run on reserves and the peg is abandoned 9 Krugman (1979) realized that the model could be used to explain currency crises Government is running a fiscal deficit It can fix the exchange rate and temporarily fund the deficit from its foreign exchange reserves 10 There is an exchange rate over time such that the “inflation tax” covers the deficit Ln S(t) Ln S* T’ t Equilibrium has predictable run on reserves and abandonment of peg 11 Problems with first generation models Timing of currency crises is very unpredictable There are often jumps in exchange rates Government actions to eliminate deficits? E.g. ERM crisis of 1992 when the pound and the lira dropped out of the mechanism 12 Second generation models Obstfeld (1996): Extent government is prepared to fight the speculators is endogenous. This can lead to multiple equilibria. There are three agents A government that sells reserves to fix it currency’s exchange rate Two private holders of domestic currency who can continue to hold it or who can sell it to the government for foreign currency 13 Each trader has reserves of 6 Transactions costs of trading are 1 If the government runs out of reserves it is forced to devalue by 50 percent 14 High Reserve Game: Gov. Reserves = 20 Trader 1 Hold Sell Trader 2 Hold 0,0 -1,0 Sell 0,-1 -1,-1 There is no devaluation because gov. doesn’t run out of reserves. If either trader sells they bear the transaction costs. The unique equilibrium is (0, 0) 15 Low Reserve Game: Gov. Reserves = 6 Trader 1 Hold Sell Trader 2 Hold 0,0 2,0 Sell 0,2 0.5,0.5 Either trader can force the government to run out of reserves The unique equilibrium is (0.5, 0.5) 16 Medium Reserve Game: Gov. Reserves = 10 Trader 1 Hold Sell Trader 2 Hold 0,0 -1,0 Sell 0,-1 1.5,1.5 Both traders need to sell for a devaluation to occur Multiple equilibria (0.5, 0.5) and (1.5,1.5) 17 Equilibrium selection Sunspots – doesn’t really deal with issue Morris and Shin (1998) approach Arbitrarily small lack of common knowledge about fundamentals can lead to unique equilibrium 18 With common knowledge about fundamentals e.g. currency reserves C Unique Peg fails CL Multiple CU Unique Peg holds 19 With lack of common knowledge Unique C* Peg fails Unique Peg holds Major advance over sunspots Empirical evidence? 20 Twin Crises Kaminsky and Reinhart (1999) have investigated joint occurrence of currency and banking crises In the 1970’s when financial systems were highly regulated currency crises were not accompanied by banking crises After the financial liberalizations that occurred in the 1980’s currency crises and banking crises have become intertwined 21 The usual sequence is that banking sector problems are followed by a currency crisis and this further exacerbates the banking crisis Kaminsky and Reinhart find that the twin crises are related to weak economic fundamentals - crises when fundamentals are sound are rare Important to develop theoretical models of twin crises 22 Panic-based twin crises Chang and Velasco (2000a, b) have a multiple equilibrium model like Diamond and Dybvig (1983) Chang and Velasco introduce money as an argument in the utility function and a central bank controls the ratio of currency to consumption 23 Banking and currency crises are “sunspot phenomena” Different exchange rate regimes correspond to different rules for regulating the currency-consumption ratio Policy aim is to reduce parameter space where “bad equilibrium” exists 24 Fundamental-based twin crises Allen and Gale (2000) extends Allen and Gale (1998) to allow for international lending and borrowing Risk neutral international debt markets Consider small country with risky domestic assets 25 Banks Use deposit contracts with investors subject to early/late liquidity shocks Can borrow and lend using the international debt markets Domestic versus dollar loans 26 Domestic currency debt Risk sharing achieved through: Bank liabilities Deposit contracts Large amount of domestic currency bonds Bank assets Domestic risky assets Large amount of foreign currency bonds 27 Government adjusts exchange rate so the value of banks’ foreign assets allows them to avoid banking crisis and costly liquidation Risk neutral international (domestic currency) bond holders bear most of the risk while domestic depositors bear little risk If portfolios large enough all risk transferred to international market 28 Viable system of international risk sharing for developed countries whose banks can borrow in domestic currency Many emerging countries’ banks cannot borrow in domestic currency because of the fear of inflation – they must borrow using dollar-denominated debt 29 Dollar-denominated debt The benefits that a central bank and international bond market can bring are reduced Dollarization: The central bank may no longer be Dollar debts and domestic currency deposits: It may able to prevent financial crises and inefficient liquidation of assets not be possible to share risk with the international bond market 30 Policy Implications Is the IMF important as lender of last resort like a domestic central bank (Krugman (1998) and Fischer (1999) OR It misallocates resources because it interferes with markets (Friedman (1998) and Schwartz (1998)? 31 Framework above allows these issues to be addressed Case 1: Flexible Exchange rates and Foreign Debt in Domestic Currency – No IMF needed Case 2: Foreign Borrowing Denominated in Foreign Currency – IMF needed to prevent banking crises with costly liquidation and contagion 32 Conclusions When is government the problem and when is it the solution? The importance of twin crises Interaction of exchange rate policies and bank portfolios in avoiding crises and ensuring risk sharing 33