**********

Not since the Great Depression has so much of the world faced widespread banking problems, with 112 episodes of systemic banking crises in 93 countries since the late 1970s (Caprio and Klingebiel 1999). These crises have imposed significant economic and fiscal costs on the countries involved; Honohan and Klingebiel (2000) find the average direct costs of banking collapses to be equal to 12.8 percent of GDP, with many countries' direct costs substantially exceeding this percentage.

Problems in the banking sector extend well beyond the fiscal cost to taxpayers, for a number of reasons. First, many firms do not have significant access to nonbank sources of external finance. Second, most firms have relied on financing from domestic banks, with bank relationships being highly valued and frequently including cross-shareholding or inclusion of bank representatives on the firm's board of directors. Third, most domestic banks in a given country have had similar portfolio exposures, so that banking problems have tended to affect the entire banking sector, rather than being idiosyncratic and affecting only a few individual banks. Thus, a major domestic shock can impair the solvency of a country's entire banking industry, leaving a country with no (or few) healthy major banks.

Such a sharp deterioration in the health of a country's banking sector forces the government to make a stark choice. On the one hand, bank regulators can undertake strict enforcement of bank regulations that will result in the widespread closure of insolvent banks. This can ensure the safety and soundness of the banks that do survive, but bank closures can be quite expensive for taxpayers, and the cost of the ensuing credit crunch can be substantial for individual firms and for the overall performance of the macroeconomy. While the early closure of insolvent banks can stop the flow of red ink and contain the cost to the government of recapitalizing the banking system, at least in the short run, the increased macroeconomic costs associated with lost GDP have the potential to more than offset any cost savings, as weakened and failed firms cut production and employment. This path becomes even more problematic for policymakers if it leads to destabilization of the economy and political unrest. The alternative for bank regulators is to follow a policy of forbearance, allowing insolvent banks (and firms) to continue operating. Such a policy may limit the severity of any credit crunch, but it may also increase the ultimate cost to the government of recapitalizing the banking system. This will be particularly true if the moral hazard problem leads insolvent banks to take risky bets in a gamble for resurrection.

Bank regulators in many countries in Asia and Latin America have been focused on triage for their banking sector, and many banking reforms have, by necessity, been a pragmatic reaction to evolving domestic economic problems. Some countries have initiated major reforms, such as enhanced disclosure in financial statements, measures to improve transparency, and enhanced regulatory oversight. However, the sequence of measures taken has frequently had a pattern of two steps forward and one step back, as bank regulators have sometimes retreated from their initial supervisory and regulatory reforms in an attempt to satisfy political constraints and placate a populace resentful of squandered funds and the huge potential tax liabilities caused by banking problems.

In response to the difficulties associated with reforming domestic bank supervision during a banking crisis, a number of countries have, in effect, imported their bank supervision by encouraging greater penetration of domestic markets by foreign banks. While foreign banks are subject to supervision by the host country, they also are supervised by their home country supervisor, which frequently provides more oversight and requires greater disclosure than traditionally has been the case in many emerging markets.

A decision to open up domestic banking markets to foreign competition can provide important potential benefits for the host country, but it is not without significant risks. Among the benefits of opening domestic markets to foreign bank entry are the importation of new management and information technologies to improve banking services, the provision of a new source of funds to recapitalize a troubled banking sector, the provision of an alternative "safe haven" within the country that can reduce the volume of domestic funds that flow offshore during a financial crisis, and the presence of deep-pocket, well-capitalized (foreign) banks that can continue lending following a major adverse shock that substantially weakens the domestic banking sector. Arguments against allowing the entry of foreign banks into domestic markets usually include concerns that the competition from foreign firms will weaken domestic banks, that local regulatory and monetary authorities will have a diminished ability to alter bank behavio r, that adverse shocks to foreign banks that are external to the host country may be destabilizing insofar as they adversely affect the banks' behavior in the host country, and that foreign banks will not serve as a stabilizing influence by providing additional credit during a crisis in the host country.

Of course, many firms still have access to credit from foreign banks, even if those banks do not have a local presence. Much cross-border lending occurs through offices in a bank's home country (or even one of its subsidiaries located in a third country), with no subsidiary (or even branch presence) located in the country in which the borrowing firm is headquartered. While retail banking requires brick and mortar points of contact with customers, wholesale banking requires a much smaller investment. For example, banks with no physical presence in a country can lend substantial volumes of funds to firms and governmental entities of that country through project finance and loan participations. However, the composition of borrowers will differ, depending on whether a foreign bank has a physical presence in a country or makes all of its loans from offshore. For example, offshore lending would tend to benefit multinational firms and the larger, more well-established firms in a country. On the other hand, a local r etail-banking network would rely in part on funding from local depositors and also typically make credit available to small and mid-sized firms, as well as individual consumers. Thus, the distribution of the potential benefits across economic agents in a country that can arise from the provision of intermediary services by foreign banks will differ, depending on whether the banks have a physical presence in that country and on the nature of their operations.

A key question that has important implications for the extent and nature of any benefits or costs of foreign bank activity is how foreign banks behave during a banking crisis. Will foreign banks fill the void left by weakened domestic banks whose lending capacity is reduced by a large domestic shock, or will foreign banks retreat in the face of emerging problems in the host economy? Will strong foreign banks with a local presence serve as a safe haven for domestic depositors that is a viable alternative to moving their deposits offshore? In that case, a flight to quality by depositors associated with a crisis (or the threat of a crisis) will not produce as severe a drain of funds from the country's banking system. Furthermore, if foreign banks do expand market share, how do they choose to expand? Do they increase offshore lending from offices in the home country, expand lending from their branches or subsidiaries located in the host country, or make acquisitions of existing banks in the host country?

This paper focuses on the Latin American experience with foreign bank penetration, the response of foreign banks during a crisis, and the implications for bank supervision of having globally active banks with significant stakes in the domestic banking sector of emerging economies. Focusing on Latin America has several advantages. First, individual Latin American countries have adopted different strategies toward foreign banks. Argentina has become particularly open to foreign banks and Mexico is beginning to be more open, while countries such as Brazil and Ecuador have been somewhat reluctant to open their banking markets. Second, the series of shocks that buffeted Latin America during the 1990s provides an opportunity to examine how foreign banks respond to both banking and currency crises that potentially have significant effects on the domestic economy.