1. "Price and Volume Dynamics in Bubbles," with Jingchi Liao, 2019
We propose a framework based on two ingredients—extrapolative beliefs and the disposition effect—and show that it can generate the sharp rise in both prices and volume observed in many bubbles. We test this framework using novel, account-level data on the 2014–2015 Chinese stock market bubble. The interaction of extrapolative beliefs and the disposition effect explains 30% of the rise in volume. Investors who are both extrapolative and prone to the disposition effect are quick to buy a stock with positive past returns, but also quick to sell it if good returns continue.
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 investor 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.
3. "Positive Feedback Trading and Stock Prices," with Chen Wang, 2019
We show that mutual funds contribute to cross-sectional momentum and excess volatility through positive feedback trading. Stocks held by positive feedback funds exhibit much stronger momentum, almost doubling the returns from a simple momentum strategy. This “enhanced” momentum is robust to alternative measures of positive feedback trading and cannot be explained by other stock characteristics, ex-post firm fundamentals, fund flows, or herding. Moreover, enhanced momentum is almost fully reversed after one quarter, suggesting initial overshooting and subsequent reversal. We argue the most likely explanation is the price pressure from positive feedback trading. Finally, we relate positive feedback trading to mutual fund performance and show that it can positively predict a fund’s return from active management.
Work in Progress
4. "How do Investors Extrapolate? Evidence from Surveys," with Adriana Robertson