Skip to Main Content
Access your saved content
In the wake of the 2008 financial crisis and in line with new regulations from the Federal Reserve Board (FRB), banks are having to upgrade their stress-testing capabilities.
Accenture and SAS Financial Services group have developed a six-step stress-testing methodology to support them in this effort.
Large financial institutions in Europe and the United States have been following rigorous, regulator-mandated stress-testing programs since 2009. Even institutions not subject to required prescriptive examinations are more likely than ever to perform some form of stress testing. However, recent economic calamities have demonstrated that the depth and rigor of existing stress-testing processes are often insufficient.
These shortcomings, combined with new liquidity requirements and rules from the FRB, should inspire many banks to upgrade their stress-testing capabilities. In December 2011, for example, the FRB finalized its provision that bank holding companies with consolidated assets of $50 billion or more submit annual capital plans. Rigorous stress testing is an integral part of these requirements.
We designed our stress-testing methodology to help banks of all sizes understand the possible impacts of FRB-mandated stress testing.
Based on our observations of banks' crisis responses, and our examinations of the Supervisory Capital Assessment Program (SCAP), Comprehensive Capital Adequacy Review (CCAR), European Banking Authority (EBA) and Bank for International Settlements (BIS) programs, Accenture and SAS formulated a six-step approach to stress testing:
Develop generic scenarios. Whether or not they use history to help develop scenarios, stress-testing teams must develop scenarios that represent a robust indicator of the bank's specific vulnerabilities.
Identify relevant macroeconomic factors. The next step involves translating the generic scenarios into quantitative macroeconomic factors that can be used as initial inputs to the segment-specific models.
Segment the portfolio. Portfolio segmentation must be granular enough to incorporate the most relevant characteristics of each asset class, and the use to which the results will be put have to be clearly articulated.
Localize macro factors into the segmented portfolio. The econometric variables generated in Step 2 must be translated into a series of factor inputs.
Run scenarios. While this methodology allows an institution to leverage many existing default, loss and revenue models, the typical horizon for stress-testing initiatives tends to be longer than most modeling activities.
Aggregate results and analyze. Because any stress test results are likely to undergo scrutiny, an institution must be able to drill down into the constituent segments to identify the institution's particular sensitivities and perform coherence analyses on the results.
Bill Spinardis executive director, Risk Management and key offerings lead North America, is based in Washington D.C. With more than 25 years of experience in enterprise risk management, corporate governance, and risk quantification, his broad insight helps organizations implement enterprise wide risk management solutions. Through the integration of an organization’s risk management and decision-making processes, Spinard guides Fortune 500 and large-scale non-profit organizations in implementing sustainable, practical risk management solutions that drive their efforts to become high-performance organizations.
Carsten Heiliger is a senior risk consultant in the Americas Risk Practice at SAS. He focuses on stress testing, loss forecasting, and credit scoring. Prior to joining SAS, he worked at a large regional US bank where he managed the loss forecasting and stress testing of the wholesale credit portfolio. In that role, his team also supported due diligence activity on potential acquisition targets, the allowance for loan and lease losses, and goodwill forecasting. He has also worked at a multinational commercial bank in Frankfurt, Germany, and at a global investment bank in New York. Heiliger has a bachelor’s degree in computer science, specializing in artificial intelligence, and a master’s degree in international business.
March 26, 2012
Skip Footer Links