Flows, Price Pressure, and Hedge Fund Returns

The authors studied how capital flows affect hedge fund returns and found that funds with high inflows outperform funds with high outflows during the month of the flows. This immediate reaction, combined with feedback trading, gives rise to a cycle: Flows exert price pressure, this effect on returns induces more flows, and these flows cause further price pressure. The cycle is so strong that it takes two years for a full return reversal, and it contributes to the observed persistence in hedge fund performance. The impact of flows on returns has clear implications for performance evaluation: One-third of estimated hedge fund alphas are due to flows.Over the past two decades, hedge funds have experienced large inflows of capital. Academic literature shows that fund flows result in an uninformed demand shift, which may affect asset prices. In this study, we examined the effect that capital flows have on hedge fund returns. Our results are consistent with a mechanism whereby funds respond to flows by scaling their portfolios up or down, rather than diversifying. These trades have a contemporaneous price impact on the funds’ underlying assets, leading to an effect on fund-level returns. Reversal of the initial flow-induced price pressure is delayed by the price impact exerted by further performance-chasing flows. The sequential nature of hedge fund flows and returns—monthly flows are submitted before learning the concurrent return—allows us to quantify the contemporaneous effect that flows have on fund-level returns.Our article contains four key results. First, hedge fund returns exhibit statistically and economically significant flow-induced price pressure: Funds that received high inflows outperformed funds experiencing large outflows during the month of the flows. This effect is present in calendar-time portfolios sorted on flows, in fund-level time-series regressions, and in cross-sectional regressions. When we sorted hedge funds into five flow portfolios, the high-flow funds outperformed the low-flow funds by 0.96% per month, which indicates that hedge fund flows are sizable enough in relation to the liquidity of the underlying assets for flows to have a significant price impact.To obtain our second key result, we traced the cumulative outperformance of the initial high-flow (high-inflow) hedge funds over the low-flow (high-outflow) funds to examine the long-term impact of the flows. The cumulative outperformance rose gradually for the first 10 months following the flows, after which it started to slowly revert. A full reversal of the initial price impact took a total of about 24 months. This delayed onset of reversal can be explained by a flow–return cycle: Flows exert price pressure, this effect on returns induces more flows, and these flows cause further price pressure.Our third result is that the flow–return cycle contributes significantly to the observed performance persistence in hedge funds. If both returns and flows depend on past returns, we should observe a relation between the two even in the absence of a causal link. We showed that this is not the case; both past returns and contemporaneous flows have a significant positive impact on hedge fund returns. We also constructed two investment strategies—one for the flow impact and the other for momentum—and compared their returns. The alpha of the flow impact strategy is both positive and significant. Further, even when adding the momentum returns to the model, the flow impact alpha remains positive and significant. We thus concluded that the two effects coexist and, if anything, flow impact is a driver of performance persistence, and not vice versa.Finally, our fourth set of results deals with the implications of flow impact for hedge fund performance attribution. If the price impact exerted by flows is not perfectly correlated with the risk factors used, a part of what is interpreted to be pure manager skill may be driven by flows. To our knowledge, we are the first to quantify the flow effect on estimated alphas. In practice, we augmented the Fung–Hsieh seven-factor model with a flow impact factor. The estimates of hedge fund alphas fell by one-third when including the flow impact factor, further underscoring the importance of assessing flow impacts when measuring the sources of hedge fund returns.