If you're not good at something, the smart thing to do is go find a professional. If you're sick and don't know what you have, you go to a doctor. If the wiring in your house has a short circuit, you call an electrician. But does this approach work in investing?
Probably not. Doctors and electricians have to prove their skill and training to get certified. Money managers don't. For almost two decades now, the average hedge fund has delivered a lacklustre return. Actively managed mutual funds generally fail to beat the market after fees, either in a given year or over time. Although most fund managers do have skill, and could generally beat the market trading on their own, the fees they charge more than cancel out their investing edge.
There are definitely some fund managers who do beat the market, even taking fees into account. But how do you, the investor, know who they are? How do you separate the wheat from the chaff, as it were?
One way is to look at past performance. An electrician who successfully fixes most houses is probably a good electrician. There's a branch of economic theory that says picking a manager based on past performance is a rational thing to do. In 2004, Jonathan Berk and Richard Green showed how this could work. In their model, investors judge managers' skill based on their performance, so money flows to the managers who do the best. But since most investing strategies don't scale up very well, the inflows of money drive returns back down. In Mr. Berk and Mr. Green's model, inflows of money balance out manager skill exactly, so that the average fund performs as well as the market, after fees. It's an efficient, optimal world.
Unfortunately, it's probably an imaginary world. The fact that the average fund underperforms the market after fees means that customers routinely overpay, which isn't possible in a fully rational model. Also, there's evidence that the return chasing done by real investors isn't quite as canny or rational as what the agents in Berk and Green's model do.
Economists Guillermo Baquero and Marno Verbeek have a new paper, in which they examine the factors that make investors give their money to hedge funds. Using data from an advisory firm, they look at how performance predicts fund flows. What they find is that one factor is very important -- the length of a fund's unbroken winning streak, defined as the number of quarters it has beaten Treasury bills. The more consecutive quarters a fund does better than T-bills, the more money goes to the fund.
This is strange, because it's very easy to imagine a world in which this is a very bad indicator of performance. Suppose Fund A returns 0.1 per cent in 99 out of 100 years, while fund B returns -0.1 per cent half the time and 10 per cent the other half. Fund A is going to generate much longer winning streaks than Fund B. But Fund B earns a much higher return on average. So using the length of a winning streak is irrational in this case.
What about the real world? Mr. Baquero and Mr. Verbeek find that the length of winning streaks doesn't predict fund performance. Investors who put their money into hedge funds that have long unbroken runs of success are not getting a good deal.
So what does work instead? Mr. Baquero and Mr. Verbeek find that the best way for predicting returns is the good old Sharpe ratio, measured over the previous two years. Unlike the length of a winning streak, a Sharpe ratio allows for both wins and losses. It also takes into account both reward and risk, instead of just measuring how long a fund can tread water.
The authors have basically discovered that it's possible for investors to behave like rational performance-chasers. For some reason, they just don't.
Why don't investors chase returns the right way? Mr. Baquero and Mr. Verbeek suggest that people have "extrapolative expectations," meaning they expect current trends to continue. But there are other possibilities. Maybe investors have a model in their heads of how long a fund is able to keep its edge after discovering a profitable trading strategy. Or maybe investors judge risk by scanning performance records for recent losses, and just try to minimize risk.
But whatever the reason, the result is clear -- investors aren't choosing funds in a rational way. Maybe that's why returns have been lacklustre in the hedge fund industry in recent years -- investors might just be sending their money to the wrong ones.
Noah Smith is a Bloomberg View columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.