How Is This Research Useful to Practitioners?
Against the backdrop of the current surge in low-cost passive investing, the authors’ research sheds new light on the crucial role active management plays in determining market prices. In contrast to the focus of the extant literature on the cross-sectional properties of fund turnover, the authors develop a model of time-varying profit opportunity. This model predicts a more significant time-series relationship between turnover and benchmark-adjusted future returns because of the timing mismatch of when trading costs are incurred and profits are measured.
For all the actively managed US equity funds studied, a one standard deviation increase in turnover resulted in a 0.66% annualized increase in gross returns. This relationship does not necessarily help investors, because “the active management industry may not provide superior net returns to its investors,” as shown in prior studies. Nonetheless, the authors illustrate that trading skill exists, noting that skillful active managers tend to charge higher fees. Practitioners are likely to find the research useful in understanding when the aggregate stock market is likely to be mispriced, which offers heightened profit opportunities.
The authors study the co-movement of fund turnover by extracting the turnover caused by investor inflows and outflows. Turnover co-moves across active US equity funds through time, and the relationship is stronger for funds with similar characteristics. Funds trade more when sentiment or volatility is high as well as when liquidity is low. The increased turnover predicts improved future performance for the funds, unlike the lower returns experienced by households that trade more often, as shown in previous research.
The authors find that the average turnover for growth funds is consistently much higher than the average turnover for value funds.
How Did the Authors Conduct This Research?
The authors use a dataset that combines CRSP and Morningstar data for the period 1979–2011. The returns of 3,126 actively managed US equity mutual funds are measured against the funds’ Morningstar-designated benchmarks. The authors use the US SEC definition of turnover, which is the ratio of the smaller of a fund’s annual purchases and sales scaled by the fund’s average net asset value over the same period.
The key regression specifies benchmark-adjusted gross returns as the dependent variable and turnover in the previous fiscal year as the independent variable. The authors enhance the power of their statistical inferences by running panel regressions. They initially pool across all funds and later, across four categories: fund size, expense ratio, small-cap versus large-cap style, and value versus growth style. Consistent with the theoretical model’s prediction, a stronger turnover performance relationship exists for funds that are able to trade less-liquid stocks, such as smaller funds and small-cap funds. Funds that charge higher fees also have a strong turnover–performance relationship, while large funds have the weakest relationship.
A logical deduction would be that a small fund focusing on small-cap stocks would have the strongest turnover–performance relationship. The authors did not study this type of fund.
The authors perform multiple tests to ensure robustness, including studying results over more than one year and isolating funds sold by brokers. They check whether funds are exploiting 11 well-known anomalies, such as momentum and return on assets, and find no significant relationships.
Abstractor’s Viewpoint
The authors’ key insight is that even with lower net returns, active management plays an important societal role in the price discovery process. This is consistent with increasing concerns that the current rise in passive investing could be having undesirable consequences in terms of herding and decreased shareholder activism.
That the authors do not directly identify profit opportunities and market mispricing is worth noting. Indeed, they suggest that whatever mispricing higher turnover is attempting to exploit, it does not appear to be due to the 11 well-known anomalies, including accruals-, momentum-, and gross profitability–based anomalies.
The research is likely to be immediately useful to performance attribution specialists, who will now be able to add market opportunity to their various other style categories.