Hedge funds are a magnet for attention, not all of it positive. But instead of being a villain in the recent financial crisis, it turns out that hedge funds were as much a victim as other segments of the market. Hedge funds are often marketed to investors as good tools to increase returns and reduce risk. The reduced risk results from the low correlation between hedge fund returns and other assets in investors’ diversified portfolios. But in the financial tumult of the past three years, hedge funds have surprised investors with their poor performance.
Hedge funds fared worse than some investors expected during the financial crisis, in part because their risk is often underestimated by traditional risk analysis like other instruments of “alternative” assets. As a result, their historical risk-adjusted performance is often overestimated. Thus, hedge funds can be riskier than they seem on first examination, and their returns may not match the promise of their reputation.
Underestimating the downside
That is the assessment of Robert Korajczyk, a professor of finance at the Kellogg School of Management, and co-authors Gregory Connor, a professor at the National University of Ireland, and Lisa R. Goldberg, an executive director at MSCI Barra. They recently published Portfolio Risk Analysis, an overview of financial risk modeling with an emphasis on practical applications, empirical reality, and historical perspective. In one chapter the authors examine the pitfalls in risk analysis performed by managers of alternative assets, such as hedge funds.
“Our work essentially conveys that many hedge funds are riskier than many investors believe,” Korajczyk says. “We wanted to alert investors to biases in risk estimates and show them alternative ways to estimate fund risk that eliminate those biases…By having better risk estimates, investors will be able to structure their portfolios in a way that exposes them to fewer surprises driven by poor quality risk measures.”
Korajczyk and his co-authors point to two factors that lead to underestimating hedge fund risk. “The first factor is that these types of funds often invest in illiquid assets whose quoted prices update more slowly than more liquid assets,” Korajczyk says. “Therefore, risk in the long run is higher than short-horizon measures imply. The second factor is that these types of funds often have asymmetric payoffs. That is, their downside risk is higher than their upside potential. Standard risk measures assume symmetric exposures, which underestimate downside risk and overestimate upside potential.”
Betting on the downside
The Government Accountability Office defines hedge funds as “pooled investment vehicles that are privately managed and often engage in active trading of various types of securities and commodity futures and options contracts.” Hedge funds are exempt from some federal regulations governing other investment vehicles such as mutual funds.
In these loosely regulated funds, fund managers can hedge their investments; in other words, they can take steps toward moderating possible losses. Selling short is an example of this kind of strategy. By selling short, you borrow a stock or commodity from a broker for a specified period of time in order to sell it, hopefully at a high price. Then, you find somewhere to buy it at a lower price and return the lower-priced entity to the lender. In this way, you make money by selling something you did not own, buying its replacement for less, and returning the replacement to the lender.
Short selling is a common way to make money off an asset whose price has gone down. However, you would take a loss if the price goes up after you have made your sale. Appropriate combinations of long and short positions can lead to a portfolio whose returns are unrelated to returns on aggregate market indices. However, most categories of hedge funds are not fully hedged, exposing investors to market risk. Those risks are harder to quantify than risks with conventional investments, due to biases built into the different ways they are traded.
Biases can smooth over the risks
Korajczyk points out that measuring the true risk of a portfolio of illiquid assets poses considerable difficulties. Biases tend to underestimate risk and overestimate results, even results that are adjusted for risk. Two notable biases that investors need to look out for include nonsynchronous pricing and price smoothing, Korajczyk says.
Figure 1: The gray area is the overlap in time between the measured return on asset I and asset J. Their common movement will be underestimated using these measured returns because it does not take into account the lead-lag relation between the two.
In nonsynchronous pricing, the closing price for a measurement period (say one month) on different assets occurs at different times (Figure 1). The closing price is the last transaction price of the month, but the last transaction of one asset might have come much earlier in the period than the last transaction of another. The later closing transaction encompasses more information on the trends of the entire period, with the asset being heavily traded. The earlier closing transaction contains less information because it missed some significant portion of the period’s activities and was not heavily traded—it was not a full participant. So a lightly traded stock with an early closing price can look like it has less volatility and market exposure than it really has.
Price smoothing can result when assets are not priced for long periods of time. Prices of illiquid assets actually might be estimates of what a price would have been if a transaction had occurred, even though no transaction took place to establish a factual price. Relying on these estimates or “appraisals” could mean that the stated value of an asset is stale.
These biases and more will apply across the broad range of alternative assets, including hedge funds, private equity, and, to a lesser extent, real estate and assets such as timberland. Taken as a whole, Korajczyk concludes, the adjusted level of risk for these asset classes is significantly higher than risk levels derived from standard methods that do not account for stale pricing and asymmetric risk exposures.
So in considering alternative assets classes, let the buyer beware, let the borrower beware, and let the seller beware—especially if they are all the same investor.