Scientific organisers: Aaron Tornell (UCLA) and Helge Berger (CESifo).
"Do we really know that currency crises are macroeconomic?" - Andrew Rose's keynote address got the workshop started by addressing one of the more fundamental questions in the field. The issue is of tremendous importance for policy makers for instance in the emerging market and transition countries that were exposed to the currency and financial crises during the 1990s. After all, if macroeconomic fundamentals were indeed irrelevant, what policy guidelines are there to be followed? Critically surveying the most recent literature (including many of his own papers and some of those economists present), Rose's answer was pointedly negative in the tradition of the Rogoff-Meese result: "macroeconomic phenomena", he concluded, "are reasonably unhelpful in forecasting crises." He suggested taking a closer look at the microeconomic structure of currency markets to improve the rather poor track record of the profession in crises forecasting. Not everyone in the audience was as pessimistic regarding the link of macroeconomic fundamentals to currency crises as Rose, and so the discussion was lively.
In the first of a series of presentations focusing on the major players in recent currency and financial crises, Olivier Jeanne investigated the role of the IMF as an international lender of last resort. His empirical analysis shows that there is only a limited transfer component in IMF lending for international bailouts in the post-war period. In fact, many of the recent bailouts granted by the IMF have been repaid. This means that bailouts are ultimately shouldered by national tax payers, not the IMF. Notwithstanding this fact, international bailouts might provoke moral hazard because they provide the liquidity for domestic bailouts. Conditioning the size of IMF bailouts on domestic regulatory reforms might help to counteract this effect. Alexander Karmann in his discussion warned against underestimating the risk for the international taxpayer associated with IMF lending, as financial and currency crises fostered by moral hazard might be contagious, ultimately necessitating even more international bailouts.
Hans Gersbach's paper, relating to Olivier Jeanne's policy recommendation, focused on financial regulators. Comparing different policy options, he suggested that in case of banking crises a "bail out the big banks"-policy in combination with some constructive ambiguity would be optimal. This strategy builds on the so-called funds concentration effect, i.e., the fact that a partial bailout allows the remaining banks less costly access to new depositors. Serving as discussant, Craig Burnside asked, however, whether the introduction of hedging possibilities that would allow bank's to insure against liquidity risks would lead to different policy conclusions. During the discussion some participants pointed to the possible endogeneity of bank size as a potential problem for discriminatory regulatory policies.
Moving on to his own presentation, Craig Burnside turned his attention to the fiscal implications of currency and financial crises. Adapting a first generation model to capture some of the stylized facts of such "twin crises" for the Korean and Mexican case, he linked different financing strategies of the high fiscal costs of such crises to their consequences for inflation and exchange rates. According to Burnside, both Mexico and Korea have not duly provisioned for the cost of their bailout. He also noted that Mexico has so far paid the cost through seignorage, while Korea has implemented a fiscal reform. Hans Gersbach in his discussion asked for a more precise formulation of monetary policy in the model and pointed out that this might lead to multiple equilibria.
Shifting the focus of the discussion to the regulation of international capital flows, Helge Berger asked whether the common notion that widespread deregulation was behind the recent demise of fixed exchange rate regimes was indeed correct. He argued that, if policy makers take into account the consequences of financial deregulation for the stability of exchange rate regimes, the variation in capital control measures should be seen as endogenous to the exchange rate regime decision instead. Empirical evidence supports this view. Aaron Tornell in his discussion inquired about the precise nature of the capital control measures analysed and discussed the example of Malaysia, which has regulated international financial flows during the Asian crisis.
Gaston Gelos used a very comprehensive dataset of mutual funds to investigate the claim that the observed volatility of capital movements during times of currency and financial crises is due to the herding behaviour of institutional investors. While there is some evidence of herding patterns, the overall picture is not one of "panic". He also reported that there is no evidence of a greater reliance on feedback trading rules during times of crisis and more tranquil periods. Dalia Marin stirred the discussion by asking whether Gelos' results implied that herding and its possible real consequences would be unimportant for policy makers.
Turning from the major players to the crisis phenomenon as such, Romain Ranciere presented an attempt to reconcile the contradiction between two empirical findings relating the growth of the financial sector and real activity. His results show that the positive correlation in the long-run identified by the growth literature can coexist with mostly negative effects in the short-run stressed in the financial crises literature. In his discussion Chandima Mendis suggested supporting this finding by using a more complete growth model also including investment variables. For his own presentation Chandima Mendis discussed the role of exchange rate regimes for the impact of internal and external shocks on the likelihood of banking crises. His empirical results suggest that while fixed exchange rate regimes lower the probability that internal shocks lead to a banking crisis, pegging regimes increase the probability of external shocks to trigger a banking crisis. Andrew Rose, joining a lively debate, cautioned against treating currency unions as a fixed exchange rate regime in such empirical exercises, while discussant Romain Ranciere pointed to the lack of a theoretical consensus regarding the interaction of the exchange rate regime and the occurrence of financial crises.
A model that provides an explanation of banking and currency crises that is indeed independent of the exchange rate regime was presented by Aaron Tornell. He gave an analysis of the behaviour of the tradable and nontradable sectors in an economy with credit market imperfections and bailout guarantees. The interaction of these distortions helps to explain the high growth rates in nontradables, the appreciation of the real exchange rate and the balance sheet/currency-mismatch often observed before a "twin crisis". As Olivier Jeanne stressed in his discussion, the model also allows an evaluation of the trade-off between the benefits of the existence of a domestic safety net for the banking sector and the ensuing moral hazard problems. In the last paper presented in the workshop, Dalia Marin described how the coexistence of high yielding government bonds and barter trade held the Russian economy in a "banking development trap" that only the financial crisis in 1998 could resolve. The crisis gave banks an incentive to lend to the private sector and firms reason to return to the banking sector for their liquidity needs. The escape from the banking development trap adds to the explanation of the rapid recovery of the Russian economy after the crises. In his discussion Gaston Gelos wondered whether the empirical part of Marin's paper could be improved by a more detailed analysis of the bank lending behaviour.
HB/AT (23 July 2001)
For more information on the workshop, see the following pages:
RSS