The results suggest that it was those countries with high levels...

The results suggest that it was those countries with high levels of corporate debt that were hardest hit: the corporate leverage parameter estimates consistently have the highest statistics and omission of leverage greatly worsens the fit. The capital inflow measure seems to have little explanatory power it self, suggesting it is not the degree of the credit cutoff that explains the differential output contractions per se. Openness is also negatively correlated with output adjustment. While these regression results are based on a small sample, the results are suggestive and indicate that more cross-country analysis of corporate leverage would provide a basis for crisis prevention policies.

The channels

The contractionary impulse of a cutoff of capital inflows is amplified into a systemic crisis largely by corporate sector balance sheet channels. Initially, the shock is localized in the foreign exchange market, but it is then transmitted to the real sector, and, thereafter, passed on to banks through nonperforming loans. Decapitalized banks curtail their lending, exacerbating the downturn.

Rapid exchange rate depreciation can quickly and onerously increase the foreign debt servicing costs of heavily indebted and unhedged firms (particularly for net importers) and can even threaten their viability, especially if exchange rates overshoot. For the nine crisis A panel regression that would provide more general inference regarding corporate leverage. is precluded by a lack of data. The advantage of the small sample cross section approach employed here is that the results are less prone to within sample structural variance, and it essentially conditions on a single global shock (Berg and Patillo, 1999).

The most important of the overlapping and mutually reinforcing explanations for the crisis found in the literature and thier implications for output are: common external shocks; spillover; tight monetary and fiscal policies (Sachs and Radelet, 1998); loose monetary policy, domestic bank overlending (stressed by Corsetti et al., 1998, Krugman, 1998 and Dooley, 1997); political risks (cf. Roubini et al., 1998); and excessive corporate leverage (emphasized in Krugman, 1999a and 1999b and Kim and Stone (1999).

The standard errors are Newey-West heteroskedasticity and autocorrelation adjusted countries, nominal exchange rates against the U.S. dollars at the trough month fell by an average of nearly 50 percent against the pre crisis peaks (Table 8). The impact of these depreciations will depend very much on the level of corporate foreign indebtedness. For example, a 50 percent depreciation is estimated by Gray (1999) to reduce the equity value of Korean corporations by 9 percent and that of Indonesian corporations by fully 21 percent owing to the heavier foreign debt burden of the latter. Not surprisingly, corporate crisis dynamics go hand in hand with currency crisis (Table 9). By the definition used in IMF (1998) all of the seven nontransition episodes involved a currency crisis. For those that took place in the last half of the 1990s, the lag between the currency crisis and the corporate crisis trough averaged 11 months, shorter than for the earlier crises.