Stress testing with fully flexible causal inputs

Propagating causal stress tests on selected risk factors to all the risk drivers is a challenging task. Attilio Meucci relies on entropy pooling to address this issue

The distribution of the key risk drivers in any given market can never be estimated completely correctly. Stress testing becomes the only effective tool to handle estimation risk both in risk management and portfolio management: the base-case risk model is modified manually, and the ensuing profit and loss distribution is calculated and evaluated.

Stress testing has earned an even more prominent place in the industry following the financial crises of the late 2000s. Basel III, the global regulatory standard for banks, advocates tighter limits on risk-based capital. The US Federal Reserve declared in December 2011 that it will conform to Basel III standards. More precisely, “a wide range of measures addressing issues such as capital, liquidity, credit exposure, stress testing, risk management, and early remediation requirements, is mandated by the Dodd-Frank Wall Street Reform and Consumer Protection Act”. Specifically on stress testing, the US Federal Reserve declared that “stress tests of the companies would be conducted annually by the board using three economic and financial market scenarios” (US Federal Reserve System, 2011).

Click here for full paper.