نمایی منحصر به فرد از استفاده از مشتقات صندوق تامین : حفاظت و یا گمانه زنی؟

We study the common equity and equity option positions of hedge fund investment advisors over the 1999–2006 period. We find that hedge funds' stock positions predict future returns and that option positions predict both volatility and returns on the underlying stock. A quarterly tracking portfolio of stocks based on publicly observable hedge fund option holdings earns abnormal returns of 1.55% through the end of the quarter. Net of fees, hedge funds using options deliver higher benchmark-adjusted portfolio returns and lower risk than nonusers. The results suggest that hedge fund positions reflect significant timing and selectivity skill.

How derivatives are used by investment managers is still largely an open question in the finance literature. The few available empirical studies highlight nonspeculative uses. Koski and Pontiff (1999) find that derivatives are used by mutual funds to reduce fluctuations in portfolio risk, especially systematic risk. In another study of US mutual funds, Deli and Varma (2002, p. 97) conclude that “[the] primary benefits associated with … derivatives are the potential to economize on trading costs, costs of liquidity-motivated trading, and the opportunity costs of holding cash.” These activities neither imply nor are implied by managerial informed trading about stock fundamentals. In this paper, we directly examine whether derivatives also play a speculative role in institutional portfolios by studying the common stock and equity option holdings of a large sample of hedge fund investment advisors over the 1999–2006 period. The hedge fund industry provides an attractive setting in which to study speculative motives for holding derivatives. Hedge funds are different from mutual funds because they are largely unregulated and can implement diverse trading strategies using several types of securities. Hedge funds could, therefore, use derivatives broadly to earn higher returns on information production. In contrast, mutual funds must comply with the Investment Company Act of 1940, and its provisions make using derivatives difficult in practice. Many mutual funds also voluntarily adopt outright prohibitions on the holdings of individual equity options.1 Therefore, a hedge fund setting will better enable detection of speculative motives if they exist. Restrictions on derivatives might be burdensome for all managers but would be the most inhibiting for any subset of managers with the capacity for informed trading. To the extent those best informed are attracted to the less restricted hedge fund industry, our approach is well suited to study the role of derivatives in informed trading. Equity options are an obvious potential vehicle for exploiting volatility timing, that is superior knowledge about stocks' volatility. Calls and puts are also a high leverage channel through which an investor can profit from selectivity skill, that is information about the direction of the underlying stock price (see, e.g., Black and Scholes, 1973 and Merton, 1973; and Cox and Rubinstein, 1985). We report several new empirical findings. First, we examine the well-publicized case of the Nasdaq technology bubble. Brunnermeier and Nagel (2004) report that hedge funds heavily invest in high-priced tech stocks over the period 1999–2000. Our sample confirms this pattern in stock holdings, but we also find that the technology sector constitutes nearly half of the total notional value underlying aggregate hedge fund put holdings. Thus, as volatility increases over the summer of 2000, the hedge funds effectively win doubly, from both price direction and volatility. Second, we undertake a comprehensive investigation into volatility timing ability as revealed by hedge funds' holdings of options over the period 1999–2006. Whereas, in general, Black and Scholes implied volatilities overestimate subsequent realized volatilities, we show a clear pattern in which hedge funds' nondirectional option strategies (e.g., protective puts and straddles) are associated with an attenuation or outright reversal of this effect. For example, we estimate that the difference between realized and implied volatility is −3.84% per month among securities for which no advisors hold corresponding option positions. In stark contrast, this difference is +5.36% when all advisors using the security do so as part of a nondirectional option strategy. When hedge funds report holding options, volatilities tend to increase. Third, we test the selectivity skill revealed by advisors' stock and option holdings through measurement of subsequent abnormal returns in the underlying stocks. We find that stock positions predict future returns, especially when held in focused portfolios that contain relatively few stock positions. In addition, we find that call and put option positions reflect strong selectivity skill. Specifically, a call-minus-put portfolio that buys (sells) stocks underlying call (put) holdings earns average abnormal returns of 1.62% per month over the three months following each quarter end. Fourth, we analyze how quickly the directional information contained in option holdings is reflected in security prices. Typically, a significant lag exists between quarter end and the public filing date. We exploit this feature of the data and partition the sample depending upon whether or not the holdings disclosures are yet publicly observable. A portfolio that buys (sells) stocks underlying call (put) holdings the day after the filing date earns average abnormal returns of 1.55% through the end of the quarter. These returns are calculated before any transaction cost but are based upon publicly available disclosure information. Therefore, this evidence would potentially qualify as a rejection of the joint hypothesis of semi-strong form market efficiency and the benchmark employed. Finally, we examine whether the apparent informed character of hedge funds' option holdings contributes to the success of their constituent investors. We find that option usage is associated with significantly lower after-fee return volatility and higher Sharpe ratio. Moreover, hedge funds deliver higher benchmark-adjusted portfolio returns and lower market risk during quarters that immediately precede and follow greater reported option usage. For example, we estimate that an increase in directional put positions from 0% to 10% is associated with a 35 basis point increase in monthly excess returns. Taken together, our findings suggest that hedge funds use option holdings to profit from volatility timing information and selectivity skill, and these rents are largely passed through to investors in the form of after-fee returns. Our findings broaden those of two recent studies of hedge funds' common stock holdings: Griffin and Xu (2009) and Brunnermeier and Nagel (2004). Our data are collected from original US Securities and Exchange Commission (SEC) filings instead of the commercially available Thomson Financial and CDA/Spectrum database that omits the call and put holdings disclosures. Griffin and Xu (2009) find that hedge fund stock positions are not predictive of future benchmark-adjusted stock returns and, therefore, call into question existing evidence that hedge funds generate alpha (see, e.g, Kosowski, Naik, and Teo, 2007; and Agarwal and Naik, 2004). Unlike option positions, however, many advisors consistently report hundreds of different stocks being held in their 13F filings. This could make it difficult to detect stock selection skill because very broad holdings of stocks might be less likely to be based on significant firm-specific information. Our findings suggest that informed trading by hedge funds is both statistically and economically significant in holdings that plausibly reflect concentrated bets, such as option positions and stocks held in focused portfolios. Our findings also significantly extend evidence suggesting that option market data are informative about the probability distribution of future stock prices. While the option data used in existing studies are market volume aggregates that include both uninformed and informed trades, our sample includes only holdings likely to be informed.2Pan and Poteshman (2006, p. 873) conclude that the predictability of stock prices by option market trading activity is due to “valuable private information in the option volume rather than an inefficiency across the stock and option market.” In contrast, our findings suggest that profitable trading strategies in both the stock and options markets can be identified using only publicly observable information contained in filings of Form 13F. The remainder of the paper is organized as follows. Section 2 describes the data. Section 3 discusses the methodology and empirical results on volatility. Section 4 considers option positions and stock returns. Section 5 addresses hedge funds' portfolio performance. Section 6 concludes.

We decipher eight years of required disclosures by 250 hedge fund managers to directly study whether derivatives play a speculative role in professional investment management. Existing research finds no speculative role for derivatives in mutual fund management, but instead mainly a hedging role. A speculative role for derivatives is more likely to be discovered in the hedge fund industry, to the extent that its managers are both less constrained and more able to generate valuable information than mutual fund managers. We find evidence of two forms of managerial ability that might be revealed by options holdings. Nondirectional strategies, such as straddle positions and protective puts, reveal volatility timing ability because they are followed by significantly higher than normal volatility for the underlying stock. Directional option holdings, such as simple calls and puts with no accompanying underlying stock holdings reported, reflect significant selectivity skill. Buying stocks in which hedge funds take call option positions and holding for one quarter gives a portfolio having 10.68% per year alpha. Buying stocks in which hedge funds take put option positions and holding for one quarter gives a portfolio exhibiting −8.76% per year alpha. We do not claim these are achievable returns for other investors because they exclude transaction costs and ignore the average 45-day reporting lag following the end of each quarter. Nonetheless, a stock portfolio tracking directional put holdings earns benchmark-adjusted returns of −1.16%, on average, in the third month following the quarter end. That third month begins two weeks after the 45-day reporting deadline. Another feasible tracking portfolio, one entering positions only after actual filing dates, produces stark differences in call-based and put-based stock returns over the remainder of the quarter, 1.77% or 1.55%, depending upon the benchmark. We also find evidence of significant selectivity skill from stock positions that plausibly reflect concentrated bets. Detecting predictive power from all stock positions is made difficult, however, by noise from sector or index tracking (or other broadly diversified stock holding) managers. Because very few 13F filings contain many option positions, the same concern would not apply as strongly to detecting predictive power of option holdings. The numbers here emphatically point out the informed character of hedge fund trades, but they do not directly measure hedge funds' investors' performance. Our analysis of after-fee portfolio returns reveals that, compared with nonusers, option users manage portfolios that are larger, have lower return standard deviation, higher Sharpe ratios, and higher excess returns relative to a style benchmark. Overall, the evidence highlights the option market as a useful tool for allowing managers to exploit information. Finally, our results are relevant to potential regulatory changes to the disclosure requirements of investment managers, especially hedge funds. For example, expanding the scope of portfolio disclosure to include other derivative securities, such as credit default swaps, would potentially increase the public stock of information about stock fundamentals, subject to countervailing free-riding threats to a manager's incentive to produce information.