During the recent financial crisis, the use of redemption gates by hedge fundmanagers caught many investors by surprise. Gates allow hedge funds to limit the percentage of fund capital that can be redeemed by investors at any point in time. Hedge funds that raised gates include the large and hitherto successful Citadel, Tudor Investor Corp, Fortress Investment Group, and D.E. Shaw. Fund managers argue that gates protect investors as they permit funds to liquidate in an orderly fashion and avoid selling assets at fire sale prices (Pulvino, 1998; Mitchell, Pedersen, and Pulvino, 2007). Investors contend that fund managers who raised gates, especially those who continue to levy management fees on gated capital, care more about business continuity than about investor protection. Underlying all this are concerns that the hedge fund industry suffers from an asset-liability mismatch. Investors worry that there may be a disparity between the liquidity that hedge funds say they can provide and the liquidity of their underlying assets
Motivated by these events, we study hedge funds that offer favorable redemption terms, i.e., monthly redemptions or better. These funds provide a fertile ground to search for instances when hedge funds over promise in terms of liquidity. We ask the following: How liquid are these liquid hedge funds? Do these hedge funds take on excessive liquidity risk? That is, are they forced to sell assets at fire sale prices in response to investor redemptions? If so, what drives the excessive liquidity risk-taking? To proxy for excessive liquidity risk, we use the impact of investor demand shocks on fund returns. In doing so, we leverage on the Brunnermeier and Pedersen (2009) concept of a loss spiral, a concept motivated by Shleifer and Vishny’s (1992) work on asset fire sales. In a loss spiral, initial losses by speculators precipitate investor redemptions which force speculators to sell assets at fire sale prices, thereby inducing further investor withdrawals. According to Brunnermeier and Pedersen (2009), this interaction between market liquidity (the ease with which assets can be traded) and funding liquidity (the ease with which traders can obtain financing) can explain why liquidity can suddenly dry up, co-moves with the market, and has commonalities across securities. A major channel through which this interaction can occur is via hedge funds. Anecdotal evidence suggests that this channel has become more important as several investment banks have scaled back or wound down their proprietary trading operations following the 2008 financial crisis.
The empirical findings are striking. We show that there exists substantial variation in the liquidity risk of these “liquid” hedge funds. Within this group of funds, the portfolio of funds with high liquidity risk exposure outperforms the portfolio of funds with low liquidity risk exposure by 5.80 percent per year (t-statistic = 2.26). To measure systematic liquidity risk exposure, we use fund beta with respect to the Pástor and Stambaugh (2003) market-wide liquidity measure (henceforth PS measure). The PS measure is particularly suited for gauging liquidity risk as it is based on temporary price changes accompanying order flow. We account for risks that are not directly related to liquidity with the Fung and Hsieh (2004) seven-factor model. We adjust the bond factors from the Fung and Hsieh (2004) model appropriately for duration so that they represent returns on traded portfolios. After adjusting for co-variation with these factors, the spread is 6.11 percent per year (t-statistic = 2.58). The relationship between liquidity risk exposure and fund performance also manifests in cross-sectional regressions. Controlling for other hedge fund characteristics, a one-standard deviation increase in liquidity risk exposure is associated with a 2.20 percent per annum (t-statistic = 2.90) surge in annual returns. These results reinforce those of Sadka (2010) who shows that liquidity risk, as measured by the Sadka (2006) information-driven, permanent variable component of price impact, can explain the cross-sectional variation in hedge fund returns. Since the price impact of asset fire sales, as envisaged by Shleifer and Vishny (1992), is transitory and unrelated to information, we argue that the PS measure is more relevant for our purposes.
The aforementioned results suggest that hedge funds that grant favorable redemption terms differ significantly in terms of their appetites for liquidity risk. Moreover, the rewards for bearing liquidity risk are high. But do these hedge funds take on excessive liquidity risk? We show that liquidity risk exposure parlays into problems for hedge funds when investors deploy and redeem capital. On average, hedge funds that experience high inflows subsequently outperform hedge funds that experience low inflows by 4.79 percent per year (t-statistic = 4.70) after accounting for co-variation with the factors from the Fung and Hsieh (2004) model. These results are robust to adjustments for backfill and incubation bias, fund fees, and thin-trading induced serial correlation in fund returns (Getmansky, Lo, and Makarov, 2004).
Consistent with a fire sale story, liquidity risk amplifies the effects of capital flows on fund returns, both in the cross-section and inter-temporally. Within the fund quintile with the highest exposure to liquidity risk, the flow portfolio abnormal spread is 4.97 percent per year. Conversely, within the fund quintile with the lowest exposure to liquidity risk, the spread is 2.84 percent per year. When the markets are bereft of liquidity, i.e., when the PS measure falls below its 20th percentile level, the flow portfolio abnormal spread is 9.13 percent per year. When markets are flushed with liquidity, i.e., when the PS measure rises above its 80th percentile level, the spread is only -1.48 percent per year. In addition, the spread is particularly large for months that are anecdotally associated with sharp contractions in market liquidity. For example, in August 1998, during the LTCM crisis, the annualized abnormal spread was 24.57 percent. More recently, in March 2008 with the demise of Bear Stearns and in September 2008 with the bankruptcy of Lehman Brothers, the annualized abnormal spreads were 8.57 percent and 6.37 percent, respectively.
