We develop a general approach for stress testing correlations in stock and credit portfolios. Using Bayesian variable selection methods, we build a sparse factor structure, linking individual names or stocks with country and industry factors. We specify a parametric form of the correlation matrix, where correlations of stock returns are represented as a function of the country and industry factors. Regular calibration yields a distribution of economically meaningful stress scenarios on the factors, which can then be translated into stressed correlations. The method also lends itself as a reverse tress testing framework: using e.g. the Mahalanobis distance on the joint risk factor distribution, allows to infer worst-case correlation scenarios. We give examples of stress tests on a large portfolio of European and North American stocks.
Recorded at 9.9.2021 / 6th European COST Conference on Artificial Intelligence in Industry and Finance - Afternoon Finance Session