More than l00 episodes of bank insolvencies have been documented by Caprio and Klingebiel (1996) for the period from 1980 to 1995 - and this does not include the Asian crisis. About 75 percent of these were classified as major insolvencies with potential systemic implications. The overwhelming majority occurred in developing or transition countries but eleven major episodes were recorded for industrialised countries including the Scandinavian crises in the late 1980s (Norway) and the early 1990s (Finland and Sweden), the Savings and Loans crisis in the USA during the late 1980s, the Crédit Lyonnais case in l994-95, and Japan in the 1990s. These are the most recent episodes of stress where large losses led to insolvencies. The transfer payments from governments in the wake of these crises are often very large. The costs of the Savings and Loans clean up in the USA has been estimated at 3.2 percent of GDP. This figure is actually low even by industrialised country standards. In the Finnish, Norwegian and Swedish crises, the transfers to banks amounted to 8.1 percent, 3.6 percent and 4.1 percent of GDP, respectively. In developing countries the costs have been between 10 and 20 percent of GDP in many cases. The two most expensive cases were Argentina and Chile in the early 1980s. In both countries, the transfers associated with the banking crises exceeded 40 percent of GDP (Caprio and Klingebiel, l996). These transfer costs do not include the potentially more serious effects of banking crises on output and employment. Thus, an important objective of supervision and regulation of financial institutions is to prevent the occurrence of crises that burden the tax-payers, while providing incentives and conditions for efficiency in the provision of financial services.