Variations in Overnight Return Correlation of U S Stocks Explained by Liquidity

Abstract

This article shows that the overnight price reversal phenomenon namely stock overpricing at daily market open followed by price reversal of individual stocks can be effectively explained by liquidity I measure the degree of price reversal in terms of “overnight (return) correlation ” which is the Pearson correlation coefficient of daytime and overnight returns Overnight correlations of all U S common stocks from 2004 to 2013 are grouped into deciles using liquidity measures I find the composite liquidity measure best captures the variations in overnight correlation at up to 51 6% of the overall range The most illiquid decile of stocks bears the lowest average overnight correlation down to -0 367 while that of the most liquid decile is very close to zero Moreover volatility and skewness of daily returns bring only small and non-monotonic impact to overnight correlation The above findings show that the liquidity of individual stocks must be taken into account for trading strategies in order to better manage overnight risk In particular investors must avoid trading illiquid stocks at market open in the same direction of overnight returns

Date of Award	2014 Aug 27
Original language	English
Supervisor	Min-Hsien Chiang (Supervisor)