Paid for Performance, and Freed From the Herd (NYTimes, 10/23/05)

October 23, 2005
Strategies
Paid for Performance, and Freed From the Herd
By MARK HULBERT
EVIDENCE is mounting that mutual fund managers are more likely to think for themselves when their pay is closely linked to their fund's performance.
Just last month, the Janus Capital Group, the fund company based in Denver, established performance incentives for the managers of 13 of its mutual funds. This means that they will earn significantly more if their funds outperform the indexes they use as benchmarks. Janus thus joins several large fund families, including Fidelity Investments and the Vanguard Group, that use such incentives to compensate some of their funds' managers.
In a recent study, Massimo Massa, a finance professor at Insead, the French business school, and two Ph.D. students there, Nishant Dass and Rajdeep Patgiri, found that performance incentives have had a significant and salutary influence on the buying and selling decisions of fund managers.
The study, "Mutual Funds and Bubbles: The Surprising Role of Contractual Incentives," has circulated in academic circles since July.
(A version is at http://papers.ssrn.com/sol3/papers.cfm?abstract-id=759365.)
Funds compensate their managers in a variety of ways - and some methods, of course, offer more performance incentives than others. The managers who have the least incentive for good performance, according to the researchers, are those who are paid a declining percentage of assets under management as those assets grow. At the other end of the spectrum are managers, now including some at Janus, whose compensation is significantly greater if certain performance hurdles are met. In the middle are those whose compensation is a flat percentage of assets under management.
The researchers measured how various incentives affected a fund manager's predisposition to buy Internet stocks during the late 1990's. They found that managers with the strongest pay-for-performance incentives tended to have the smallest investments in such companies. That, in turn, tended to make their funds lag behind the average fund over the three years before the Internet bubble burst, and to outperform over the three years thereafter.
The researchers say they believe that compensation incentives affect a manager's willingness to buy and sell the same stocks that others are trading. They theorize that managers who are the least rewarded for good performance tend to be most worried about ending up at the bottom of the rankings - and are not very motivated to take a risk and try to finish at the top. This means that they will be more likely to mimic other funds' portfolios, something that the researchers call herding behavior.
During a bubble, the researchers found, these managers tend to invest in the frothiest stocks. While doing so looks risky from the investors' point of view, it is a safe strategy for managers who at all costs want to avoid falling to the bottom of the rankings. That is because they know that by jumping on the bubble's bandwagon, they will participate in the spectacular performance while the bubble is expanding, but still land in the middle of the pack when the bubble bursts.
In contrast, according to the researchers, managers who are best rewarded for good performance are less likely to run with the herd. "With high-enough contractual incentives," the researchers write, "the prospect of ranking at the top by diverging from the bubble would more than offset the incentives to have a high, but not the best, performance by riding the bubble." Such managers will tend to invest in stocks that are out of favor.
The study also helps to explain the conflicting conclusions of previous research into the effect of performance-based fees on fund returns. During a bubble like that of the late 1990's, the funds with the greatest performance-based incentives for their managers are likely to lag behind those with the least such incentives. But the opposite conclusion emerges after a bubble bursts.
The new research has major public policy implications. Securities regulators typically have taken a jaundiced view of pay-for-performance among mutual funds, based on the idea that such compensation may encourage fund managers to incur too much risk. But the researchers believe that mutual funds whose managers are poorly compensated for performance may unwittingly be contributing to the market's boom-and-bust cycle. Funds with stronger performance incentives "may provide a useful counterweight" to offset future bubbles, they said.
In light of these findings, should you be leaning toward a mutual fund that gives its managers more performance-based incentives? The answer depends on your willingness to have your fund's performance deviate significantly from that of the average fund. In the late 1990's, many investors were impatient with funds that did not load up on Internet stocks, so they dumped those funds before the bubble burst. Others were willing to tolerate their funds' low rankings in the hope of outperforming over the longer term - and, for the most part, they have had the last laugh.
Mark Hulbert is editor of The Hulbert Financial Digest, a service of MarketWatch. E-mail: strategy@nytimes.com.