Periodic dynamic conditional correlations between stock markets in Europe and the US

Abstract

This study extends the dynamic conditional correlation model of Engle (2002, Journal of Business and Economic Statistics 20, 339-350) to allow periodic (day-specific) conditional correlations of shocks across international stock markets. The properties of the resulting periodic dynamic conditional correlation (PDCC) model are examined, focusing particularly on stationarity and the implications for unconditional shock correlations. When applied to the intraweek interactions between six developed European stock markets and the United States over 1993-2005, we find very strong evidence of periodic conditional correlations for the shocks. The highest correlations are generally observed on Thursdays, with these sometimes being twice those on Monday or Tuesday. In addition to these PDCC effects, strong day-of-the-week effects are found in mean returns for the French, Italian, and Spanish stock markets, while periodic effects are also present in volatility for all stock markets except Italy.