Journal of Financial and Quantitative Analysis

Research Articles

Information, Trading Volume, and International Stock Return Comovements: Evidence from Cross-Listed Stocks

Louis Gagnona1 and G. Andrew Karolyia2

a1 School of Business, Queen’s University, 143 Union St., Kingston, Ontario, K7L 3N6, Canada. lgagnon@business.queensu.ca

a2 Johnson Graduate School of Management, Cornell University, 348 Sage Hall, Ithaca, NY 14853. gak56@cornell.edu

Abstract

We investigate the joint dynamics of returns and trading volume of 556 foreign stocks cross-listed on U.S. markets. Heterogeneous-agent trading models rationalize how trading volume reflects the quality of traders’ information signals and how it helps to disentangle whether returns are associated with portfolio-rebalancing trades or information-motivated trades. Based on these models, we hypothesize that returns in the home (U.S.) market on high-volume days are more likely to continue to spill over into the U.S. (home) market for those cross-listed stocks subject to the risk of greater informed trading. Our empirical evidence provides support for these predictions, which confirms the link between information, trading volume, and international stock return comovements that has eluded previous empirical investigations.

Footnotes

We are grateful for financial support from the Office of Research Services at Queen’s University and from the Dice Center for Financial Economics at Ohio State University. This paper was written while Karolyi was on faculty at the Fisher College of Business at Ohio State University. We express our thanks to Warren Bailey (associate editor and referee) and Stephen Brown (the editor) for excellent comments, and to participants at the 2005 meetings of the Financial Management Association; Tim Kuepfer (Datastream); Rose Liao, Alvaro Taboada, and Caroline Trevithick for their editorial assistance; Simon Lauzon, Roger Loh, and Jonathan Witmer for their research assistance; as well as to IBES for providing the data on analyst earnings forecasts. All remaining errors are our own.

Metrics