Interaction Effects in the Cross-Section of Country and Industry Returns

Abstract

We comprehensively examine the interaction effects in the cross-sectional predictability of country and industry returns. Using nearly five decades of data from 68 countries, we construct all possible double-sorted portfolios based on 44 portfolio characteristics and uncover numerous significant interactions. An out-of-sample value-weighted strategy that selects the top long-short country (industry) interactions generates a monthly World CAPM alpha of 0.33% (0.62%) with a Sharpe ratio of 0.58 (0.75). The strongest interactions stem from implementing momentum and technical analysis signals in small and illiquid countries or industries. Furthermore, the return patterns mainly emanate from frontier and weakly integrated markets—highlighting the role of market frictions and segmentation in the occurrence of abnormal returns. Consistent with these interpretations, the interactions decline over time as global markets mature and become more integrated.