Style And Skill: Hedge Funds, Mutual Funds, And Momentum
University of California, Los Angeles (UCLA) – Finance Area; Yale University – International Center for Finance; National Bureau of Economic Research (NBER)
George Washington University – Department of Finance
Federal Reserve Board
George Mason University – Finance Area
January 6, 2016
Hedge funds tend to buy stocks with low past returns, while mutual funds tend to be trend followers. The nearly two-thirds of hedge funds that follow contrarian strategies outperform their risk- and characteristic-adjusted benchmarks by 2.4% per year. Hedge funds that follow momentum strategies do not outperform their benchmarks, irrespective of whether these benchmarks control for momentum. By contrast, most mutual funds follow momentum strategies; their managers exploit the momentum anomaly but lack trading skill once we control for the effect of momentum on stock returns. The most profitable trades of contrarian hedge funds are purchases of stocks sold by momentum mutual funds. The superior performance of contrarian hedge funds is persistent and arises from strategies that are more complex than purchasing stocks with low past returns.
Style And Skill: Hedge Funds, Mutual Funds, And Momentum – Introduction
The high returns from buying assets that are past winners and selling past losers is perhaps the premier asset pricing anomaly in the literature.1 “Momentum,” as it is frequently called, has persisted since its discovery,2 and is present across multiple asset classes3 and in most countries.4 Notably, the strategy is particularly simple to implement and quite popular, particularly among institutional investors. For example, Grinblatt, Titman, and Wermers (1995) and Carhart (1997) show that mutual funds tend to follow momentum. Grinblatt and Keloharju (2000) find that Finnish institutions, mostly mutual funds, also follow momentum.
Momentum’s popularity with investors has been attributed to the strategy’s profitability and to fund managers’ incentives to trade with the crowd.5 However, little evidence exists on whether momentum is also favored by hedge funds|a $2.5 trillion industry run by what many regard as the most sophisticated and most incentivized managers.
Using 13F institutional stockholdings, we analyze the style and performance of 589 mutual fund, 1,342 hedge fund, and 2,894 other institutional fund managers from 1998 to 2012.
We document that hedge fund managers are not momentum investors. For almost 2/3 of hedge fund managers, stock purchases tend to be contrarian, although their tendency to sell recent winners is less pronounced. The invesment style of contrarian hedge funds is persistent and highly profitable. About 80% of hedge funds that are contrarian in the first half of the sample period are also contrarian in the second half. Despite the documented profitability of momentum, contrarian hedge funds exhibit the best performance. Their quarterly portfolio rebalancing generates a significantly positive alpha, outperforming both mutual funds and the approximately 1/3 of hedge fund managers that can be classified as having momentum-driven strategies. Contrarian hedge funds’ success derives from their managers’ superior stock-picking skills|the ability to pick the right losers among stocks with similar characteristics. We find performance persistence only among outperforming hedge funds|there is no evidence of performance persistence among underperforming hedge funds and outperforming or underperforming mutual funds and other institutional investors.
Consistent with prior studies, we find that about 2/3 of mutual fund managers follow momentum strategies, both for purchases and sales. The investment style of these funds is persistent: Conditional on their survival, 76% of mutual funds that follow momentum strategies in the first half of the sample period also follow the same strategies in the second half. In contrast to hedge funds, momentum mutual funds outperform contrarian mutual funds. However, once we control for the return enhancing effects of momentum, differences in performance between the two styles of mutual fund management disappear.
The style measure we use, first proposed by Grinblatt, Titman, and Wermers (1995) [GTW], is based on the vector product of the fund’s portfolio weight changes with past stock returns. When stocks with relatively high past returns witness portfolio weight increases and those with relatively low past returns generate portfolio weight decreases, this product is positive|indicating a momentum strategy. Our measure of fund performance is the analogous product, but with future returns replacing past returns. Performance thus measures the ability of the fund’s portfolio weight changes to predict future returns, following Grinblatt and Titman (1993) [GT]. An equivalent perspective is that the GT measure evaluates active management by comparing fund returns to a benchmark return that would be earned if the fund maintained its prior stockholdings.
See full PDF below.