There is no support for the view that currency movements were consistently important determinants of the performance of banks in the crisis countries.
When in 1997 economies across Asia were suddenly knocked flying from their pedestals as the prima donnas of international development, many a learned analysis was quick to finger the banking system as the locus of the problem. Most notably, several critics argued that banks suffered steep losses--and thus contributed significantly to the crisis--because they had invested in a way that brazenly ignored the possibility of a sharp currency depreciation. This has lead some to call for requiring banks to do their borrowing and lending in the same currency.
But according to a recent examination of the situation, it may be that banks did not play an outstanding role in the crisis after all. And while in many Asian countries bank investors suffered mightily--during the crisis a dollar sunk into the Korean bank index dwindled to about 14.7 cents--for the most part, losses appear to be linked to the general economic slump, not to a bank-specific impact of the exchange rate drop.
In Banks, the IMF and the Asian Crisis (NBER Working Paper No. 7361), Bong-Chan Kho and René Stulz assert that an examination of bank performance during the period indicates that "currency returns do not seem to contribute to the poor performance of East Asian banks except for Indonesia and the Philippines" and that, aside from the exceptions noted, devaluations did not hurt banks "beyond their overall impact on the economy." In other words, they find "nothing unique about the exposure of banks to exchange rates."
"From our analysis, there is no support for the view that currency movements were consistently important determinants of the performance of banks in the crisis countries once one takes into account the stock market returns in these countries," the authors conclude. "Given the many statements made about the importance of currency exposures for the East Asian banks, our results are surprising."
Kho and Stulz reach their conclusion by comparing the performance of bank shares to broader market indexes for individual countries from January 15, 1997 to July 15, 1998, a period, they noted, that "includes all the important events of the Asian crisis." They assume that such market data "captures the common effects of exchange rate shocks across industries." Still, they consider that this notion might be flawed or "biased," that instead of having a broad impact the currency drop actually slammed banks so hard that their peculiar misfortunes skewed the overall market index, just as one bad stock can drag down an otherwise sound portfolio.
However, they feel confident that this was not the case. For example, in Thailand, where bank performance plays a large role in calculating the country's market index, one would have expected to see the now infamous collapse of the baht, and its subsequent deleterious effect on the national economy, to produce a particularly bad set of numbers for Thai banks. Instead, the opposite is true.
Kho and Stulz discover that "in the case of Thailand, we find that banks benefited from depreciation of the local currency so that the bias discussed here (the influence of bank performance on the market index) would make our results even more surprising." In other words, in Thailand, bank performance during the crisis made the index look better, not worse.
Also rejected by Kho and Stulz is the general notion that a wave of crisis-induced currency devaluations in Asia hurt U.S. banks. They note that even during a particularly turbulent five day period, Chase Manhattan earned a 5.09 percent return in excess of the return predicted by general stock market movement and that a dollar invested in a U.S. bank index at the start of the crisis would have been worth $1.73 at the end.
Finally, Kho and Stulz look at the crisis-induced bailout orchestrated by the International Monetary Fund (IMF) and dismiss an oft-stated justification for such rescues: that they benefit banks in general, not just those with extensive investments in the affected countries. The authors note that their data indicate IMF actions simply made it more likely that banks that loaned a lot of money in Asia would be repaid. They found no evidence that the bailout produced widespread benefits "by somehow reducing systemic risk."
The authors believe that the direct effect, or lack thereof, of the Asia currency crisis on banks could prompt further study to learn more about why the conventional wisdom seems to have missed the mark. For example, they note that banks might have "hedged" their bets more astutely than many observers think. They also wonder whether "the market expected the currency losses to be offset by bailouts."
-- Matthew Davis