Do Sophisticated Investors Believe in the Law of Small Numbers?

Believers in the law of small numbers tend to overinfer the outcome of a random process after a small series of observations. They believe that small samples replicate the probability distribution properties of the population. We provide empirical evidence indicating that investors are mistakenly driven by this psychological bias when hiring or firing a fund manager after a successful (or losing) performance streak. Using quarterly data between 1994 and 2000 of 752 hedge funds, we analyze actual money flows into and out of hedge funds and their relationship with the length of the streak. We first show that persistence patterns have a predictive ability of future relative performance of a manager: the longer the winner streak, the larger the probability for a fund to remain a winner. Investors, in turn, appear to be aware of quality dispersion across managers and respond by following a momentum strategy: the longer the winning (losing) streak, the more likely they will invest in (divest from) that fund. Yet, we find that investors place excessive weight in the managers' track record as a criterion for decision. Our model shows that the length of the streak has an economically and statistically significant impact on money flows beyond rationally expected performance, which confirms a ldquo;hot-handrdquo; bias driving to a large extent momentum investing. Apparently, even sophisticated investors exhibit psychological biases that may have adverse effects on their wealth