بحران بانکی و رژیم نرخ ارز: آیا یک لینک وجود دارد؟

This paper investigates the linkages between banking crises and exchange rate regimes, using a comprehensive cross-country dataset for the last two decades. The paper examines whether the choice of exchange rate regime affects the likelihood, cost, and duration of banking crises. Empirical results seem to indicate that adopting a fixed exchange rate diminishes the likelihood of banking crises among developing countries. However, once crises occur, the real costs associated with them appear to be larger in countries with fixed exchange rates. The duration of crises does not seem to be robustly affected by the exchange rate policy

Following the recent financial crises in Asia, Brazil, and Russia, the debate over the appropriate exchange rate regime has once again taken center stage. Since the work of Mundell (1961), an extensive literature has developed examining the links between the exchange rate regime and macroeconomic performance.1 However, until recently, this literature had largely ignored the implications of the exchange rate regime for financial stability. Similarly, most studies on the determinants of banking crises have focused primarily on the role of the macroeconomic, external, and regulatory environments.2 Recently, a number of studies (see Chang and Velasco, 1998, Eichengreen and Hausmann, 1999, Eichengreen and Rose, 1998 and Hausmann et al., 1999; Velasco and Cespedes, 1999) have begun to discuss—primarily at the theoretical level—the potential links between the exchange rate regime and financial stability. With the exception of Eichengreen and Arteta (2000) and Eichengreen and Rose (1998), this issue remains largely unexplored at the empirical level.3 Furthermore, the existing empirical papers focus exclusively on developing countries and ignore indirect channels through which the exchange rate regime may affect the likelihood of banking crises, beyond the impact of external shocks. Also, these studies are silent on the question of how, if at all, exchange rate regimes affect the cost and duration of crises. This paper attempts to fill some of the gaps in the empirical literature on the links between exchange rate policies and banking crises. Using a comprehensive data set including developed and developing countries for the period 1980–1997, we examine whether the choice of exchange rate regime affects the likelihood, cost, and duration of banking crises. Regarding the likelihood of crises, we test the validity of some of the indirect channels recently discussed in the literature linking the exchange rate regime to the probability of banking crises. In particular, we examine (a) whether unhedged liabilities increase the likelihood of banking crises under fixed exchange rate regimes, (b) whether the extent to which broad money is backed by reserves affects the likelihood of banking crises caused by runs under pegged regimes; (c) whether the liquidity of the banking system reduces the negative repercussions of the lack of a lender of last resort under fixed exchange rate regimes; and finally (d) whether the impact of external shocks on the probability of banking crises varies under different exchange rate regimes. Also, we extend the existing empirical work in a number of ways. Firstly, we consider a larger number of banking crises that affected not only developing, but also developed countries. Secondly, we study whether the results on the role of the exchange rate regime vary when we use alternative measures of de facto exchange rate flexibility (or lack thereof). Finally, to verify the robustness of our results, we control for the potential endogeneity of the exchange rate regime.4 In the estimations of the cost and duration of crises, we analyze the impact of the exchange rate regime, controlling for macroeconomic and financial factors affecting these variables. We construct three measures of the cost of crises. Firstly, we use estimates of the fiscal cost of crises. These are related to the clean-up cost of banking crises faced by the government. Secondly, given the end date of crises as determined by studies that document the chronology of these events, such as Caprio and Klingebiel (1999), we measure the cost of crises to the economy as the foregone output growth that resulted from the episodes of banking distress. Thirdly, we allow the end date of crises to be determined as the period when growth goes back to the pre-crises trend and we measure the cost of crises as the differential between the growth rate before and during crises. Finally, we estimate hazard models to study the impact of the currency regime on the duration of crises, where the end date of crises is determined by the dates identified in Caprio and Klingebiel (1999), in some estimations, and given by the time it takes the economy to recover to its pre-crises trend growth, in others. The remainder of this paper is structured as follows. Section 2 discusses the literature and arguments relating exchange rate policies to banking crises. Section 3 describes the empirical methodology and the data used in this study. Section 4 presents the empirical results. Finally, Section 5 concludes.

The choice of exchange rate regime is probably one of the most important macro-economic policy decisions, especially for developing countries. A number of studies have tried to ferret out the influence of exchange rate arrangements on economic performance. This strand of the literature, however, has paid little attention to the link between the exchange rate regime and financial stability. This study attempted to fill this void in the literature by empirically investigating the impact of the exchange rate regime on the likelihood, cost, and duration of banking crises, in a large sample of developed and developing countries over the period 1980–1997. The main conclusion that emerges from our study is that exchange rate stability appears to reduce the probability of banking crises, particularly in developing countries. This finding seems to be robust to various specifications, different criteria for identifying banking crisis periods, using measures of de facto exchange rate flexibility, and even controlling for the possible endogeneity of the exchange rate regime. Lack of exchange rate flexibility appears to increase the cost of crises, however, these results are not as robust. On the other hand, we find no consistent evidence that the exchange rate regime affects the duration of crises. Table 7 provides a summary of the main results in the paper.A number of factors could explain our empirical results. Firstly, in the context of the modern literature on exchange rate regimes, which underscores the existence of important trade-offs between credibility and flexibility, our findings suggest that the credibility associated with fixed exchange rates may also help to promote financial stability. Since the fixed exchange regime reduces both erratic and discretionary policy making, it would also decrease the occurrence of domestic shocks that frequently produce banking crises. Secondly, the findings also agree with those who argue that flexible exchange rates are no longer a useful shock absorber for real shocks. By contrast, they can now be a transmitter of financial shocks, thereby adversely affecting productive economic activities. More precisely, countries with flexible exchange rates, except those with well-developed and sophisticated markets, are likely to experience a surge in the volatility of the real value of domestic assets due to the increased capital mobility. Excessive fluctuations in the real value of domestic assets may, in turn, thwart financial stability. Thirdly, it is also possible that making exchange rate stability a priority provides additional incentives for authorities to undertake macroeconomic and institutional changes that not only allows them to maintain their exchange rate commitment, but also promote financial stability. For example, the lack of a lender of last resort and the knowledge that domestic credit expansion may cause the peg to collapse may compel bank managers and supervisors to improve their prudential standards (Eichengreen, 2002). Fourthly, the fact that the real cost of crises tends to be higher in countries with more rigid exchange rate regimes may be attributed to: (i) lending-based consumption booms, which usually take place under fixed exchange rate regimes and bring sharp contraction in economic activity, when they evaporate and (ii) inconsistency between injecting the much needed liquidity to the banking system—particularly in the absence of close substitutes for bank loans—and the exchange rate regime in place.21 A final word on financial soundness, exchange rate stability, and the choice of exchange rate regime. This paper emphasizes the role of exchange rate stability in averting banking crises. Countries can attain exchange rate stability by pursuing an active policy to fix the exchange rate or to minimize its movements. However, it is important to recognize that governments’ success in achieving exchange rate stability is not only determined by an effective and consistent design of its monetary and fiscal policy, but also depends significantly on the extent to which the economy is subject to large and frequent shocks. Under extreme volatility, it will be hard for any government, no matter how well intentioned, to maintain a stable exchange rate. Similarly, faced with small and infrequent shocks, even a country that is de jure floating its exchange rate may experience de facto exchange rate stability.