Do Institutions Pay to Play? Turnover of Institutional Ownership and Stock Returns *

We document a new type of holding cost faced by active institutions. Specifically, we show that stocks in which institutions trade more tend to be overvalued and subsequently experience relatively low returns. Because most institutions do not take short positions, on average they have long positions in the stocks they trade. As a result, the low subsequent returns on actively traded stocks impose a holding cost on institutions that is due to overvaluation. Our estimates suggest that this cost can be large. Based on ten portfolios, we find that stocks with the highest turnover of institutional ownership earn 8.9 % lower subsequent one-year returns than stocks with the lowest turnover of institutional ownership. We find that active trading by institutions entails holding costs only when institutions trade against individuals and not when they trade against other institutions. We show that institutions face higher expected holding costs when they trade in stocks that are characterized by high disagreement among investors (as measured by high total share turnover and high return volatility) and in stocks with high relative valuations (as measured by low book-to-market). In additional tests we find that an increase in institutions ’ trading opportunities (as measured by an increase in their holdings) is associated with an increase in the expected holding costs due to overvaluation. We do not find evidence that the relatively poor performance o