Everyone wants to have a top-performing manager. That’s the whole point of hiring a manager, right? No one intentionally hires a manager that lags. The difficulty comes about when good managers have temporary periods of underperformance. How do you know it’s temporary? Should you fire the manager and switch, or should you stick with them through the period of underperformance?
The traditional method of dealing with this issue has been to terminate managers who underperform over some intermediate time period, say three years, and to then replace them with a manager that outperformed over that time period. As has been shown amply in the research literature, the traditional method doesn’t work. In the inimitable words of the Psy-Fi blog,
What the plan sponsors are doing, like many behaviorally compromised individuals, is chasing returns in the same way a retriever brings back a lit stick of dynamite.
Some resources are provided in this article on the Psy-Fi blog, which deals with behavioral issues in finance. For instance, there is a link to a paper on pension funds hiring and firing practices that shows how poorly this strategy-which is in fact used by most funds-works in actual practice. In fact, doing the opposite tends to work better!
The Brandes Institute has also addressed manager selection and performance chasing. Psy-Fi points out:
…as Death, Taxes and Short-term Underperformance shows, the probability is that even the best managers will go through significant periods of poor returns. Studied over a decade even the top performing funds managed, on average, to underperform the S&P500 by over 8% during at least one three year rolling period. One (unnamed) fund managed to trail the index by over 40% in one year. All of the top five performing funds managed to underperform at some point during the decade. 53 out of 59 funds appeared in the bottom 10% of performers during at least one quarter and 10 of them managed this for a rolling three year period.
This complicates the picture quite a bit. It becomes clear that even very good managers, for whatever reason, have significant periods of underperformance, although that doesn’t stop them from outperforming over the longer term. Psy-Fi equates short-term performance with noise:
Basically firing a fund manager because of a couple of years of poor performance is simply shooting them because you don’t like noise in the system.
We all know that short-term performance may not be indicative of the longer run, but how do we decide? How do we distinguish between signal and noise?
From a practical standpoint, it seems to me that the investor has to rely on some kind of tested return factor. If a return factor has worked in the past, especially for a very long time, it seems more likely that it will continue to work in the future-even if there is enough noise in the system to create short-term underperformance from time to time.
There aren’t a lot of candidates for return factors that have worked over long periods of time. Many studies, both practitioner and academic, suggest that value and relative strength are the main return factors that have outperformed over many decades. Both factors are robust and work in numerous formulations, but both factors also move in and out of favor-essentially creating the noise that Psy-Fi talks about.
Happily, it turns out that value and relative strength as strategies are not very correlated. AQR Capital Management makes a strong case for blending relative strength and value, as the two strategies are largely complementary. From a client standpoint, then, the best recipe for success might be to find a disiplined relative strength manager (us, we hope!) and a disciplined value manager. Apportion the account appropriately and let bake for Warren Buffett’s favorite holding period-forever.
