Share these talks and lectures with your colleaguesInvite colleagues
The art and science of stress testing
Although stress tests are intended to complement the use of risk management models such as value at risk (VaR) and potential exposure models, sometimes these risk management models can be used to generate more empirically relevant stress tests. In this article, four case studies are discussed in which it is illustrated how to use risk management models in the formulation of stress tests. These examples also illustrate how risk management models may be used to formulate stress tests as well as quantify the largest risks inherent in a firm's portfolio. The paper discusses how models might be used to calibrate a stress test of a financial variable when there is very little empirical evidence on how the variable has behaved historically, and it also suggests how models can be used to calibrate and formulate stress tests that include the mitigating effects of dynamic hedging, using a credit valuation adjustment model as an example. The emphasis throughout is on the use of practical techniques that employ risk management models already in use.
The full article is available to institutions that have subscribed to the journal
Greg Hopper is a managing director at Goldman Sachs and oversees the development of counterparty credit risk models and methodologies. Prior to joining Goldman Sachs, he was an executive director at Morgan Stanley and a Federal Reserve economist.
CitationHopper, Greg (2013, December 1). The art and science of stress testing. In the Journal of Risk Management in Financial Institutions, Volume 7, Issue 1.