Preliminary. Please do not cite without permission.

This paper assesses empirically the effect of disclosure on bank stability. In doing so it offers a new approach for assessing the marginal effect of structural factors on the likelihood of crises that involves panel-data techniques applied to bank-level data. Our dataset covers more than 500 banks in 32 different countries over the years 1994-2000. Using this dataset we assess the likelihood of a bank experiencing a dramatic fall in its stock price in any given year and relate the likelihood of such a “bank crisis ” for any particular bank and in any particular year to both the existence of a deposit insurance scheme and to bank transparency, while controlling for macro- as well as bank-level factors that might also play a role in affecting the likelihood of such an event. We find evidence that bank transparency reduces the likelihood of bank crises, while the effect of deposit insurance depends on the particular design features of the regime