Reserve Requirements, Bank Runs, and Optimal Policies in Small Open Economies

Abstract

Abstract: This paper rationalizes as the outcome of an optimal policy decision the pattern of reserve requirements and other macroeconomic variables in the aftermath of a bank run. The paper develops a general equilibrium model that departs from the standard small open economy (SOE) model in three dimensions: (i) capital mobility is not perfect, (ii) there exists a costly banking system, and (iii) there is an externality affecting individual banks' decisions. The results suggest that the path of reserve requirements would depend on the type of shock that the economy receives and the effect that this shock produces on the interest rate. Interestingly, the size of the risk premium will affect the reaction of the economy to the shock. It is also shown that the dynamic adjustment will be slightly different for permanent and temporary shocks, and it will also depend on the access that the economy has to foreign funds. JEL classification: E44, E58, F30, F41, G28 Key words: reserve requirements, optimal policy, financial crisis, emerging markets 1. Introduction Saving for a rainy day is well understood in all economic fields. If regulators want to apply this policy to the amount of liquidity that banks should hold, they will say accumulate liquidity in normal times and use it in bad times, which in principle seems intuitive and reasonable. But, the evidence does not support this conventional wisdom. In fact, it suggests that the policymaker flips a coin when bad times arrive. This paper shows that resorting to the liquidity accumulated during the good days will depend on the underlying shock and the state of the economy at the time of the shock. Also, it shows that the same shock will generate a cloudy day under some circumstances and will not under others. (1) Therefore, the policymaker's reaction could be understood as an optimal response to different situations, instead as of flipping coins. Liquidity provision is key in the management of systemic banking crises, especially in emerging economies, where the lack of systemic liquidity can exacerbate problems. On one hand, without enough support, a liquidity crisis could lead to a solvency crisis and/or a credit crunch in the economy. On the other hand, the overuse of rediscounts or repos to provide liquidity assistance could lead to excessive money printing, and could weaken the peg in a fixed exchange rate regime or generate inflationary pressures. There is evidence that shows that during the 90's several central banks provided systemic liquidity to finance the bank runs generated in the financial crises. (2) In general, the liquidity support includes all the funds provided to the system through rediscounts, repos, contingent contracts, and the reduction of reserve requirements. (3) This last instrument is the focus of the paper. The literature on the optimality of maintaining a stock of liquid reserves is broad (Baltensperger (1974), Santomero (1984), Cothren and Waud (1994), Stein (1995), Agenor et al. (2000)). Also, there are some works that analyze the use of reserve requirements from a macroeconomic standpoint: as a countercyclical tool (Edwards and Vegh (1997), Calvo et al. (1993)), as a mechanism to enforce capital requirements (Fernandez and Guidotti (1996)), or as an instrument that helps to collect the inflation tax (Englud and Svensson (1988)). There are some other papers that study optimal policies under financial crises (Lahiri and Vegh (2003); Rebelo and Vegh (2001), Ganapolsky (2003)). (4) But, in spite of the importance of reserve requirements as a liquidity provider, there is no work that studies its optimal use in a financial crisis. The purpose of this paper is to provide a framework to understand the behavior of reserve requirements as the outcome of an optimal policy decision. In particular, it rationalizes the pattern of reserve requirements and other macroeconomic variables in the aftermath of a bank run. …