The headline was dramatic: “Hedge Funds Do Half as Well as You Think.”
In August, Bloomberg reported on a study of the hedge fund industry that argued hedge fund databases did not sufficiently control for biases when presenting average annualized returns. The study’s authors found, when they scrubbed the raw data, that the average return for the industry between 1996 and 2014 dropped from 12.6 percent to 6.3 percent.
As a practitioner who reviews hedge funds regularly on behalf of clients, I was a bit surprised at the magnitude of the reported drop. But the bigger surprise was that many investors believed hedge funds were achieving such high returns on average in the first place.
While the difference between 6.3 and 12.6 makes for a catchy headline, it is also important to consider the information in the wider context of what hedge funds are and what they do. Hedge funds are generally less risky than stocks and riskier than bonds, so it makes sense for the average return to be somewhere in between the two. Based on historical data going back to 1926, stocks have provided annualized returns in the low double digits, and bonds have provided returns in the low single digits.
The term “hedge fund” is an umbrella categorization, and as the study’s authors point out, not all hedge funds are alike. While they may follow many different strategies, what hedge fund managers have in common is the goal of making money regardless of the market environment. Hedge fund strategies often involve market timing, heavy borrowing to boost returns and investing in highly illiquid securities. These strategies can carry significant risks, and can lead to a wide dispersion of outcomes across different hedge funds. Some hedge funds implode spectacularly; others chug along and generate acceptable returns for their investors.
Since hedge funds can voluntarily opt in or out of indexes, it makes more sense for funds to report when their performance is good. This can lead to “survivorship bias,” where winning funds generally stay in databases and losers drop out as those funds stop reporting. “Backfill bias” occurs when hedge fund managers fill in historical performance data after a track record of success has been established, instead of simply beginning to report performance at the fund’s inception. While there are other biases in the data, these are the two biases that were adjusted for in the study Bloomberg described, which led to the 50 percent drop in returns.
In short, because hedge funds report their numbers to indexes on a voluntary basis, those figures present a highly distorted image of the industry. To be clear, no one is accusing specific funds of misreporting returns to their own clients or other misbehavior. But because hedge funds are not required to disclose their performance to the general public, the databases are often the best one can do when trying to get an estimate of the industry’s overall performance.
This study also serves as a good reminder of a truism for any investors performing due diligence: Garbage in, garbage out. Or, put a more formal way, statistical data can appear to support nearly any conclusion. Investment consultant Meredith Jones, in a blog post, pointed out that the input data for hedge funds is much messier than average, writing: “Hedge fund data is dirty. Actually, maybe even make that filthy. It’s ‘make my momma want to slap me’ dirty. Which is why it is critical to understand exactly what it is you may be looking at before jumping to any portfolio-altering conclusions.”
If you don’t think about the assumptions you are using, you can wind up with conclusions that are far off base. The study itself refers to how portfolio optimization software may use hedge fund risk-return data and spit out a recommendation that you invest 200 percent of your portfolio in hedge funds. (If you are wondering how that is possible, the idea is to invest all of your money into hedge funds and borrow an equal amount – and then put that into hedge funds too.)
I would obviously never recommend a 200 percent hedge fund allocation to a client. I wouldn’t recommend a 20 percent allocation to any of my individual clients either, because of the illiquid nature of hedge funds. But this illustrates the importance of understanding data before making allocation decisions. At Palisades Hudson Asset Management, we run portfolio optimization calculations for all of our clients, but we make sure to adjust the “optimal portfolio” to reflect practical considerations, such as lack of a meaningful long-term track record for certain asset classes.
This study also shows the importance of being aware of biases inherent in industry data, especially with hedge funds, and how those biases can shape our perceptions and expectations. An adviser touting an “exciting new hedge fund” is likely to generate a lot more interest than an adviser recommending an investment with a long-term expected return of 6 percent that charges a 2 percent management fee plus 20 percent of profits, and that cannot be liquidated for a year or more.
Not all hedge funds are bad. But this new study reminds investors that, on average, they should expect hedge fund returns and volatility to fall somewhere between stocks and bonds. Many good hedge funds can deliver on that promise; many bad ones cannot. What neither kind should be expected to do is outperform stocks in the long run.