Correlation Risk in the Context of Market Turbulences during the COVID-19 Pandemic and BCBS Stress Testing Principles

Abstract

Correlations play an important role in the risk management of banks. Changes of correlation are an important element of an adverse (stress) scenario in the BCBS framework. The purpose of this paper is to show how correlation is plagued by a number of issues that include volatility, directionality and autocorrelation. To that end, we analyze to which degree directionality and autocorrelation of correlation remove diversification benefits when they are most needed, i.e. in a crisis, and how autocorrelation amplifies correlation and makes it more persistent, creating the illusion of stable, reliable correlation levels. Furthermore, we exemplify changing correlations during the COVID-19 pandemic looking at a number of different markets. Our results suggest that prudent bank risk management should be cautious when calibrating its risk models to historical correlation levels. Market price-based stress tests should include various levels of assumed correlation as inputs to the statistical models used to assess a bank’s viability. We propose how banking supervisors and macro prudential authorities should challenge banks’ correlation assumptions and assess the rigor of applying them.