Investor regret and cryptocurrency returns

We examine whether a return-based proxy for investor regret predicts cross-sectional cryptocurrency returns. We construct REG as the shortfall of a coin’s weekly return relative to the best-performing coin in its market-capitalization group. The equal-weighted and value-weighted high-minus-low REG portfolio spreads have weekly cryptocurrency three-factor alphas of -1.20% and -0.63%, respectively. The REG-return relation remains after controlling for established cryptocurrency characteristics, persists over longer horizons without reversal, and is strongest for recent outcomes and narrowly defined peer groups. It does not vary systematically with size, illiquidity, or idiosyncratic volatility. Our findings contribute to the literature on empirical asset pricing by documenting that peer-relative return shortfalls predict subsequent cryptocurrency returns.