At a recent Performance Measurement Forum meeting, we touched on hedge funds and their use of benchmarks. The key here is that the managers are not managing against these benchmarks, but rather use them only for reference purposes.
Benchmarks are critically important, in all aspects of life. And while being known as the graduate who is "most likely to succeed" might be a challenge, we no doubt compare ourselves to others.
Our younger son, Douglas, has become quite an athlete and participates in various events, like the Tough Mudder (and I thought the Army's obstacle courses were tough!). He told my wife what his time was for his first such event, and she thought it was a good one (any time would have sounded good to this proud mother); I, on the other hand, wondered what it meant relative to others, because without a benchmark, such as mean, median, high, low, it's difficult to judge (oh, and his time WAS very good). But a number by itself means little.
A lot can be said about benchmarks, and more will be ...
.png)
Interesting points regarding benchmarks as a return reference vs something that managers "manage to." It's clear that a benchmark should represent a market segment reference point for the active manager. This can be helpful in evaluating not only the return the manager achieved, but also provide some information regarding how the manager achieved that return (and hopefully the excess return.) We try to strike a balance between the manager doing something different from the benchmark in terms of risk, style, market exposure and security selection while not deviating so far that the portfolio is no longer represented by the benchmark. Otherwise, the manager not only exposes the client to performance risk via tracking error, but also threatens to undo the overall asset allocation of the client (imagine a manager holding all cash because no attractive securities are available!) In this context we think of the manager as "managing to the benchmark." This works as long as the manager's performance can be replicated passively.
ReplyDeleteUnfortunately, this is not the case with benchmarking hedge funds. Why not? Because hedge funds are not an asset class; they are strategies (global macro, long-short, distressed, event driven, etc.) In reality, hedge fund managers are typically buying and selling the same publicly available securities found in "traditional" long-only portfolios. But when you invest with a hedge fund manager, you are giving wide latitude and discretion to someone whom you believe possesses great skill, in the hopes that you will generate high returns (or perhaps positive returns regardless of market conditions.) You are essentially substituting manager risk for market risk.
So, hedge funds are strategies that cannot be replicated passively, and thus cannot be benchmarked with a passive set of market exposures. The only practical way to evaluate the relative performance of hedge fund managers is against a peer group of similar hedge fund managers. Thought that peer groups made bad benchmarks? It's so much worse with hedge fund peer groups, since they have so many serious biases. When you consider that a hedge fund peer group is basically the average of those hedge funds who: a) did not blow up and... b) chose to report their performance - then you realize that this is simply an "average of the winners." The performance of the true hedge fund peer group is characterized by a lower return and higher volatility. In that regard, if a hedge fund manager can simply match the risk-adjusted return of the peer group, it's probably an above-average performer.
So, do these hedge fund managers "manage to" their benchmarks? Yes, they probably do, insofar as they are actively investing within their general mandate. That's about as close a definition as you would expect in an asset class where you expect each manager to "chase his own dreams" with the expectation that he will "find the pot of gold at the end of the rainbow."