Hedge funds are a trillion-dollar market that isn't nearly as transparent as other corners of the financial industry. A professor at Duke University's Fuqua School of Business has made a career out of finding the missing pieces of that puzzle.
"It is like detective work," says Professor David Hsieh. "You have to sniff out and find things that are not easily available, and even when you find it you have to figure out how to reverse engineer what the fund managers do."
Hedge funds pull together large investors and seek to increase returns with speculative strategies, including borrowing and short-selling, not commonly used by mutual funds, according to the U.S. Securities and Exchange Commission. When Hsieh began studying the industry in 1994, it was worth an estimated $100 billion. Today, it's close to $3 trillion, according to the multiple research firms that track the market.
Hsieh expands on his work in this Fuqua Q&A.
Q: Why isn't hedge fund performance as easy to measure as in other sectors of the financial market?
Hedge funds are private investments, so their managers are not required to release much information about them. The academic research that has accumulated is based mostly on self-reported databases of fund performance. But not all fund managers contribute to those databases, and there's no way to know if those who do supply data are including all of the funds they manage. Since the data we have is self-reported, how good are those numbers? Are they telling the whole truth? Or are they only giving you the numbers they want you to see? No one has access to all of the possible data. So when we draw conclusions from the data we do have, are they valid for the universe of all hedge funds, or only the ones we can see? There are all these unknowns.
Q: So how do you go about filling in those blanks?
Most of my work is on understanding how the hedge funds returns differ from traditional investments like mutual funds. When I started this research, I only had access to the self-reported data. More recently, I have found a lot of additional information in a private database. This allows me to look for performance information from funds that remove themselves from databases, as well as those that never report to them in the first place. Then I look at how that missing information changes the measurement of hedge fund performance as a whole. The additional information does not change the main conclusions based on the self-reported databases. We do find certain strategies used by hedge funds produce returns that look very different from mutual funds. Most of the work until now has been on what individual funds look like. But we know that most of the assets in the industry - probably a third to a half of all the money - are managed by 10 percent of the firms. These are humongous, very big firms, running many large funds. I try to identify which firms own which funds, so I can look for similarities in behavior. The idea is to try to understand how the strategies work. You cannot see what the funds are actually doing. You only have some description of what the strategy is. You have to reverse engineer how they generate the returns that you see. Then you wait to see if your hypothesized trades have the same returns as these funds over time - which should be the case if your hypothesis is correct.
Q: You also model how hedge funds can be expected to perform in different economic environments. Why is that important?
Since the databases have existed only since the 1990s, there's no data to tell you how hedge funds might have been affected by an event like the Great Depression, for example. However, if you understand the underlying strategy and its risks, you can extrapolate from the observed data. An example is the funds that are trade credit spreads before the 2008 recession. People thought this was a great strategy because the interest rate environment in the 1990s and early 2000s was very good to these funds. But it's like selling catastrophe insurance. You collect the premiums and you're calling that profit, and it looks like you're making money - until disaster strikes and you have to make a huge payout. All you're doing in the data is observing the good times, and you haven't seen a bad time yet because the bad times happen very rarely. And then it hits and the strategy loses a lot of money which is what happened in the financial crises in 2008 and 2009. Another interesting strategy is known as managed futures, in which the manager follows the trend of futures prices in currencies, interest rates, and commodities. This strategy tends to do well in chaotic markets, such as the fall of 2008.
Q: How can investors benefit from your research?
Investing in hedge funds is like investing in individual stocks. An investor should diversify across strategies, and across managers within each strategy. My research can identify the risks of different strategies. An investor can use this information to decide if a strategy makes sense as a part of the portfolio. If it does, then the investor should diversify across managers in that strategy. It is a good bit of work.