Reliable Quantification and Efficient Estimation of Credit Risk

Abstract

The recent crisis in the global financial markets requires a critical review of current regulatory practice. Corporate governance is a key issue since improper due diligence and myopic incentive schemes for employees contributed largely to the credit bubble and the ensuing recession. A better oversight by regulatory authorities and institutional changes of financial firms pose political and economic challenges in their own right. In addition, substantial efforts are required to devise efficient quantitative methods that allow one to measure financial risks more reliably in the future. These tools should be able to detect extreme scenarios that are very unlikely to occur but whose impact may be dramatic as illustrated by the recent liquidity crisis of Lehman Brothers, Merrill Lynch, Fannie Mae, Freddy Mac, AIG, and others. We report here a novel Monte Carlo (MC) approach for the direct and efficient computation of an important class of financial portfolio risk measures, known as Shortfall Risk (SR). Unlike the current industry standard Value-at-Risk (VaR), convex risk measures such as SR are sensitive to the tails of loss distributions and provide financial institutions with incentives to diversify properly.