Do aggregate analyst recommendations predict market returns in international markets?

This study examines the time-series variation in the predictive value of aggregate analyst recommendation ratings for 42 countries during the period 1994-2017. We find that the information content of local aggregate recommendations in the United States and in other global markets declined significantly in the aftermath of reductions in trading frictions (e.g., decimalization & co-location). These findings lend support to the emerging view that analyst recommendations exhibit a more limited predictive role during recent years amid reductions in trading frictions and increased arbitrage activity. We also study the flow of information from aggregate recommendations across markets and find that a component capturing global information is consistently predictive. Our results reveal that the global information component can be best extracted from U.S. recommendations in the early portion of the sample period. In the latter period, as the dominant role of the U.S. in the global economy decreases, this global component appears to be best measured using recommendations associated with the broader set of non-local, non-U.S. firms.