Investor behavior under the law of small numbers

I study how investors trade under the law of small numbers, the belief that even a small sample represents the characteristics of the underlying population. These investors expect short-term trends to reverse but long-term trends to continue. Using a simple model, I show that the law of small numbers can explain several well-documented trading phenomena: chasing long-term trends, bucking short-term trends, the disposition effect, and the V-shaped selling propensity. Moreover, I derive and successfully test the model's new predictions, and in doing so, I (1) provide evidence for heterogeneous horizons, (2) highlight how investors' extrapolation horizon and holding period can explain variation in their disposition effect, and (3) show that the V-shaped selling propensity is an aggregate phenomenon driven by separate groups of investors.