"... In PAGE 54: ... Table9 : Conditional Probabilities of Financial Crises Value of Composite Index Probability of Crisis 0 - 0.6 .... ..."

"... In PAGE 27: ... TABLE2 . MEASURES OF LOSSES FOLLOWING BANKING CRISIS (as a sub sample of financial crises - 17 countries and 18 crises) Countries Date of the currency crisis Permnt Loss Transit.... ..."

"... In PAGE 28: ...ANNEX TABLE1 . MEASURES OF LOSSES FOLLOWING FINANCIAL CRISES Countries Date of the crisis Permnt Loss Transit.... ..."

"... In PAGE 12: ... Several Ghanaian bank managers merely appear to have rotated assignments, muting any signal regarding performance. Turning to an evaluation of the restructuring attempts, Table6 shows how countries scored on the criteria developed elsewhere (Caprio and Klingebiel) for assessing the success of responses to insolvency. Briefly, countries received a mark if financial depth rose after the crisis to levels above that seen in the pre-crisis period; if real... In PAGE 50: ...Table6 : Analysis of Crisis Countries 46 (increasing= 1) (between 0-2. 5% 1) (-5 to 10%= 1) (yes=0, no= 1) BENIN 1988-90 increased in early 1990s 1 0 1 1 3 CAMEROON 1987- recovered from 1987 dip, but fell 0 0 1 ongoing I again in 1992-93 CAR 1980s amp; 1991- slowly dips during the 1980s, 0 0 0 0 0 increased in 1993 CHAD 1980s amp; 1990s has decreased ever since 1987 0 1 0 0 1 CONGO 1980s amp; 1991- peaked from 1986-88 and 1992, 0 0 1 0 1 now falling from these marks COTE D apos;IVOIRE 1988-91 decreased in 1989 fairly steady 0 1 0 1 2 ever since ERII-REA 1993 NA I I GHANA 1982-89 decreased from late 1970s to 0 0 1 0 1 1984, increased from 1985-89 fell again in 1990, increased in 1992, slight drop in 1993 GUINEA 1985, 1994-95 fairly steady 0 0 0 0 0 KENYA 1985-89, 1992- increased in 1990s 1 0 0 0 1 95 MADAGASCAR 1988 dipped from 1984-88, increased I 1 0 0 2 since then MAURITANIA 1984, 1988-93 steady, slight increase I 0 1 0 2 MOZAMBIQUE 1987- falling from mid-I980sto early 0 1 0 ongoing I 1990s NIGERIA 1990s decreased during the 1980s, I 0 0 ongoing I increasing slightly since 1990 SENEGAL 1988-91 gradually falling since late 1970s, 0 0 1 1 2 early 1980s TANZANIA 1987, 1995 increasing since 1989 1 0 0 0 1 TOGO 1993- steady during early 1980s, falling 0 1 ongoing I since 1986 UGANDA 1994 increased from 1992 to 1993, lack 0 1 1 ongoing 2 data ZAIRE 1991-92 increased during 1990s 1 0 NA 1 2 BULGARIA - increased in 1980s, but decreased 0 0 ongoing 0 in early 1990s ESTONIA 1992 decreasing 0 1 0 ongoing I HUNGARY 1991-95 dipped from 1988-90, increased I I I ongoing 3 and steady since then LATVIA 1995 increased in 1993-94 1 0 1 ongoing 2 POLAND 1990s increased sharply in 1989, then 1 0 1 ongoing 2 fell sharply in 1990- fairly steady, increasing since ROMANIA 1990s peaked in 1990, falling since then 0 1 0 1 2 RUSSIA 1995 Falling from 1991-94 0 0 0 ongoing 0 SLOVENIA 1990s increasing from 1992-95 1 0 ongoing I .... ..."

"... In PAGE 18: ... Table 3, which shows averages of bilateral country correlations within regions, confirms that financial market interdependence is significantly higher among regional markets. Two results, however, stand out from Table3 : first, controlling for global and country-specific factors often raises the degree of financial interdependence; and second, the residual correlations are particularly high among Southeast and East Asian markets. This suggests that financial integration contagion is stronger both within regions and in particular in Southeast and East Asia.... In PAGE 19: ... In contrast, the rankings for the two financial contagion variables correspond much more closely with the list of countries that became the main victims of either the Latin American crisis or the Asian crisis, thus suggesting that the crises were more likely to have spread through financial interdependence. For instance, South Asian markets have a low degree of financial interdependence with Southeast and East Asia ( Table3 ) despite having a relatively high degree of real integration with that region (Table 1). Thus the lack of their financial market integration and financial openness may offer an explanation as to why contagion did not hit South Asia during the 1997-98 Asian financial crisis.... ..."

"... In PAGE 13: ...Similarly, reforms aimed at further developing the financial sector broadly construed, or making it more efficient, may be bunched with liberalizations and help reduce consumption growth volatility. In this section, we add variables that control for either macro-reforms (Table 3) or financial development ( Table4 ) to our main regres- sions and investigate the impact on the liberalization effect. Of course, it is conceivable that these other variables spuriously account for some of the liberalization effect.... In PAGE 14: ... Given that portfolios worldwide are still very much biased towards the home market, an efficient domes- tic financial sector may be more important to smooth aggregate shocks than the ability to share risk internationally by investing in foreign equities. Hence, Table4 shows results with two variables controlling for financial development added to the regression. The first is a standard measure of banking development (private credit to GDP, see for example, King and Levine (1993)), which is available for all samples, and the second is stock market turnover, a standard measure for stock market development (see for example, Levine and Zervos (1998)).... ..."

"... In PAGE 8: ...lows to Latin American countries while Eichengreen, et. al.(1996b) find evidence that knowing there is a crisis elsewhere increases the probability of a domestic currency crisis. Table2 provides the dates of financial liberalization, the beginning and peak of the banking crisis, and the date of the balance of payments crisis that was nearest to the beginning of the banking crisis. 5 By selecting the nearest currency crisis, whether it predates or follows the beginning of the banking crisis, we allow the data to reveal what the temporal patterns are.... In PAGE 9: ... If, instead, the peak of the banking crisis is used as the conditioning piece of information, no valuable information is gained; indeed, the conditional probability is 22 percent and below the unconditional. This result follows from the fact that a more common pattern (see Table2 ) appears to be... In PAGE 13: ... For currency crises we focus on the 18-month period before and after the crisis. Unlike balance of payments crises, in which reserves are lost abruptly and currency pegs abandoned, banking crises are protracted affairs which tend to come in waves and, hence, the depth of the crisis is seldom reached at the first sign of outbreak (see Table2 ). For this reason, we widen the window and focus on the 18 months before the onset of the crisis, a 18-month arbitrarily chosen crisis period, and the 18 months post-crisis period.... ..."

"... In PAGE 36: ... Finally, changes in net property, plant, and equipment reveal changes in investment policy. Table8 reports results from tests examining the effects of each debt type on different aspects of operating performance. These regressions also include industry and country dummy variables, but the coefficients are not reported to conserve space.... In PAGE 36: ... However, the negative coefficient for SLC debt (significant at the 10% level) is consistent with the financial distress hypothesis and the prior evidence suggesting that SLC debt had a differential adverse effect on performance. The second column of Table8 repeats the analysis using changes in operating margin around the crisis as the dependent variable. In this case, the only significant relation is the weak negative effect of SLC debt (again consistent with the prior findings).... In PAGE 36: ... In this case, the only significant relation is the weak negative effect of SLC debt (again consistent with the prior findings). Results in the third column of Table8 reveal a positive relation between FC debt and changes in net PPE, our proxy for changes in investment. This finding also supports the theory of Bris and Koskinen (2002) which predicts that prior to a depreciation exporting firms will underinvest due to the debt overhang problem, but once a depreciation occurs these highly levered firms will undertake the foregone investments.... In PAGE 36: ... This finding also supports the theory of Bris and Koskinen (2002) which predicts that prior to a depreciation exporting firms will underinvest due to the debt overhang problem, but once a depreciation occurs these highly levered firms will undertake the foregone investments. Overall, the results in Table8 are consistent with findings of Andrade and Kaplan (1998) that financial distress costs are relatively small even for highly levered firms. To make sure the results presented in this section are robust we also try other specifications not reported here.... In PAGE 38: ... 46 All of the coefficients on the debt variables retain their signs and significance with the following exceptions: In column (1) of Table 7 the magnitude of the coefficient for SLC debt is reduced so the difference from the other debt levels is significant only at the 10% level. In column (1) Table8 , the coefficient on SLC debt is no longer significant. In column (2) of Table 8, the coefficient on NLC debt becomes significantly negative at the 10% level, and the coefficient on SLC debt changes from being ... In PAGE 38: ... In column (1) Table 8, the coefficient on SLC debt is no longer significant. In column (2) of Table8 , the coefficient on NLC debt becomes significantly negative at the 10% level, and the coefficient on SLC debt changes from being ... ..."

"... In PAGE 11: ... However, fundamental macroeconomic factors remained significant in generating aggregate risk, and the majority of the explained variation in deposits was accounted for by bank fundamentals, indicating the importance of bank soundness in depositors apos; decisions. Table4 summarizes the effect of the tequila period on the financial system. It should be noted that between December and May, the system lost $8bn or 18% of deposits and the Central Bank lost some $5bn or 30% of international reserves.... ..."

"... In PAGE 13: ...eriods. This criterion deletes around 5% of observations across the different sub-periods. We also have deleted observations if the real growth in value added exceeds 100%, which excludes only a small number of cases. Table3 presents the summary statistics of some variables that indicate changes in real sector and financial sector activity for both the pre-crisis and during crisis periods. When comparing the summary statistics of the pre-crisis and crisis periods, we find the following crisis characteristics.... ..."