How past market movements affect correlation and volatility

Abstract

The influence of past stock price movements on correlations and volatilities is essential for understanding diversification and contagion in financial markets. We develop a model that makes the influence of past returns, aggregated into driving factors for correlations and volatilities, explicit. Employing information about recent market movements leads to a more realistic model for the behavior of stock returns in a downturn than conventional models. Our approach offers a fresh perspective on the behavior of stock markets, and provides an alternative to the concept of exceedance correlation. For a US investor we find that international diversification in China or the UK remains beneficial in a crisis.