Skill and Luck in Hedge Fund "Diversification

We propose that when existing firms must raise capital to fund a new venture ("diversify"), they are more likely to diversify when they are both lucky and skillful. Firms consider diversifying when they experience extraordinarily strong performance because investors are more willing to invest in a firm's new venture when the firm has performed well historically. However, when returns on new investments reveal information about firm ability, lucky but low-ability firms are less likely to diversify compared to lucky and skillful firms. We test the theory by observing the revealed behavior and performance of profit maximizing firms using a large panel dataset on the global hedge fund industry. Our findings show that firms are more likely to diversify when they experience positive risk- adjusted excess returns, yet legacy funds' returns fall significantly following diversification. We interpret the unconditional decline in returns following diversification as evidence that firms time the launch of new funds around peak historical returns. However, legacy fund returns in diversified firms fall less than do returns in a control sample of firms that are matched based on past performance and all other observable characteristics. Furthermore, new funds in diversified firms generate positive excess returns and overall returns are higher for diversified firms than in focused firms. We interpret positive conditional performance of diversified firms as revelatory evidence that firms with greater skill diversify while those with less skill choose to remain focused. Why do firms diversify? Agency cost theorists emphasize the role of private managerial incentives (Jensen and Meckling 1974), while strategists are more likely to invoke synergies that arise from firm-specific ability (Teece 1980). Although these theories are often considered separately, together they capture the tension at the heart of diversification decisions - investors want managers to take advantage of unique firm capabilities, but they are wary of managers' private incentives. This paper takes a step toward integrating agency theory and firm-specific ability in the context of diversification by proposing and testing a model that takes both perspectives seriously. Our theoretical construct builds on the first order prediction of agency theory - firms will diversify when managers benefit from doing so - while accounting for the existence of heterogeneous firm capabilities, which implies that more capable firms are more likely to diversify. The model makes sharp predictions about the patterns of firm performance following diversification when firms must raise capital to fund a new venture that we test using a large panel dataset on the global hedge fund industry. As in Campa and Kedia (2002) and Villalonga (2004) we control for firm heterogeneity by matching diversified firms to