Trace the development of stress testing, and compare the earlier stress tests run by the European Banking Authority (EBA), CCAR and SCAP in terms of methodology.
The idea
Bank stress testing began as an internal risk tool, mostly for trading books. Traders and risk managers applied stylised shocks beside VaR, and the 1996 Basel market risk amendment expected banks to run them. It became a supervisory capital exercise with the 2009 SCAP, after the 2007 to 2009 crisis. The US Supervisory Capital Assessment Program (SCAP, 2009) imposed common baseline and adverse macro scenarios on the 19 largest bank holding companies. It combined bank estimates with supervisors' own estimates, published bank-level results and came with a public capital backstop. CCAR (from 2011) is an annual Federal Reserve exercise that also judges the capital plan, including dividends and buybacks. The EBA EU-wide tests were mainly bank-run, bottom-up under a common methodology, with a static balance sheet. Exam point: separate who models, what is held constant, and what the result is used for.
A practice question
Which statement best describes how stress testing of banks developed up to the post-crisis supervisory exercises?
- It began with supervisors running system-wide macro scenarios in the 1990s, and banks adopted it for internal risk management only later.Why not: Reverses the order: early stress tests were internal sensitivity tests on trading books, and supervisory macro tests arrived with SCAP in 2009.
- It began as an internal trading-book tool beside VaR, and became a supervisory capital exercise after 2009. - correct
- It was invented by the 2009 SCAP, the first time any bank examined its losses under adverse conditions.Why not: Treats SCAP as the origin; banks and the Basel market risk framework used stress tests on trading portfolios well before 2009.
- It has always been a purely statistical exercise in which scenario probabilities are calibrated from VaR models.Why not: Confuses stress testing with VaR; stress scenarios are stylised, judgment-based events meant to complement VaR, not derive from it.
Stress tests started inside banks, mainly for market risk and as a complement to VaR (Basel's 1996 market risk amendment expected them). After the crisis, SCAP (2009) turned them into supervisory, capital-linked, system-wide exercises.
A harder one
A risk manager compares the EBA EU-wide stress tests with the US exercises. Which statement on methodology is most accurate?
- EBA tests were mostly bank-run, bottom-up and static; SCAP and CCAR added supervisors' own estimates. - correct
- EBA tests used only supervisors' top-down models, whereas SCAP and CCAR relied solely on banks' internal models, with no supervisory review.Why not: Reverses the approaches: the EBA exercises were bank-run, and the US exercises added supervisors' own estimates as a cross-check.
- Each EBA bank chose its own scenario, whereas SCAP and CCAR imposed common scenarios on all participants.Why not: EBA tests also imposed common scenarios and a common methodology; bank-specific scenarios were not the difference.
- The EBA tests allowed full dynamic balance-sheet management, so none of the three exercises held balance sheets constant.Why not: The EBA tests used a static balance sheet assumption, so this misstates their methodology.
The EBA exercises were constrained bottom-up: banks computed results with a common methodology, a common scenario and a static balance sheet, with supervisory challenge. SCAP and CCAR added the supervisors' independent estimates of losses and revenues.
The hardest tier
Before 2008, a bank's stress test usually meant a trader's shock to a few market factors, for example equities down 20 percent. Relative to the post-crisis supervisory exercises, which description of that earlier practice is most accurate?
- Comprehensive, covering revenues, credit losses and capital over a multi-year macroeconomic scenario set by supervisors.Why not: Describes the post-crisis supervisory exercises, not the earlier trading-book sensitivity tests.
- Calibrated to a stated confidence level so that each shock had a known probability, as with VaR.Why not: Confuses stress tests with VaR; early shocks were chosen by judgment, with no associated probability.
- Common across banks, allowing supervisors to compare results across institutions.Why not: Early shocks were chosen by each bank; common scenarios arrived with SCAP, CCAR and the EBA exercises.
- Partial, market-risk focused, bank-specific, with no capital hurdle and no public disclosure. - correct
Early practice was partial (single or few factors), centred on the trading book, set by each firm and not tied to a capital hurdle. Post-crisis exercises are comprehensive, common-scenario, capital-linked and published.
In the app: 6 practice questions on this outcome, spaced review that brings it back before you forget it.
Start free →