Aggregated predictability of global anomalies

Abstract

I replicate 102 return signals across 48 countries and find a consistently low success rate under
modern empirical standards. A simple anomaly composite (AC) aggregating similar signals (e.g., B/M and E/P) significantly improves the return predictability of major effects (e.g., value) and shows that their performance highly varies across emerging and developed markets. ACs also challenge single signals on their cross-country evidence and suggest that 1) value premium is best explained by investor overextrapolation, 2) investment effect is driven by mispricing rather than optimal investment, and 3) profitability effect likely reflects both optimal investment and investor sentiment.

Keywords: empirical asset pricing, anomalies, international markets, market efficiency, mispricing, limits-to-arbitrage

JEL Classification: G10, G11, G14, G15