Welcome to GEFRI


Mission and Approach DC Global Center Institute Objectives Research Focus Working Papers Executive EducationProgram Execution Global Finance Conference, April 2006Nanotech Conference,
Fall 2006

Managing Bank Risk

By: Ted Barnhill and Liliana Schumacher


This paper focuses on bank risk in countries that have abandoned their domestic currency and adopted the American dollar as a means of payment or in countries in which the dollar is a defacto currency, together with a domestic currency. In such economies Central Banks have a capacity to provide liquidity is constrained by their capacity to acquire dollars.  Banks that operate in these economies thus face a risk profile that is significantly different from banks that operate in an economy where the central bank has the capacity to create unlimited amounts of domestic currency. We discuss this risk profile in detail. We also argue that, since much of the bank risk imposed by dollarization is systemic, banks may have little or no incentives to make provisions for the scenarios in which losses materialize. Under these circumstances, bank regulation may introduce liquidity and capital requirements in order to minimize the costs associated with a panic. The Basel capital accord does not provide any guidance for the management of bank risk in a dollarized economy. Given assumptions regarding the probability of a systemic financial system crisis, we develop a constrained optimization model for calculating the optimal amounts of capital and liquid assets banks should hold to maximize bank value.  We compare and contrast the optimal decisions for banks with and without regulatory restrictions on bank capital and liquidity.  The paper also provides insights into the potential financial system wide costs to of liquidity crises in a dollarized economy, an issue that is usually overlooked in the dollarization debate. 


We have in hand an initial computer model for undertaking the above analysis.  Current efforts focus on gathering and analyzing information from a number of countries to appropriately estimate the model parameters.