Are Young Managers All That?

Friday's The Wall Street Journal carried an article by Mark Hulbert entitled "Why New Mutual Funds Are Better." The answer in Hulbert's article is because new funds have younger portfolio managers, who are better at the task than are veterans. In that, Hulbert cites an unpublished paper called Scale and Skill in Active Management by Chicago's Lubos Pastor and Wharton's Robert Stambaugh and Lucian Taylor.

To echo Sam Gamgee, this subject needs a week's answer, or none at all.

Let's start at the beginning. In 2004, when Jonathan Berk and Richard Green, business school professors at Berkeley and Carnegie Mellon, respectively, argued that mutual fund investors are rational in buying mutual funds based on their past performance.

The claim will likely surprise you. In the popular press, the academic community generally is portrayed as fully agreeing with index-fund proponents that the past performance of active fund managers is immaterial. Because the investment markets are a random walk, per Nobel Laureate Eugene Fama and Princeton's Burton Malkiel, manager skill does not exist. There is only chance, which masquerades as skill.

Many professors, however, do not believe this thesis to be entirely correct. They do not believe that market returns are fully random, and that no opportunity exists for managers to exhibit skill. They view this very simple model of fully efficient markets/no manager ability as being a useful first-stage approximation of the truth, but far from the final word. The reality is rather messier.

Berk and Green offer one of the strongest explanations of this messy reality. Their argument is:

1) Mutual fund managers do have skill. In fact, they have a massive amount of skill--6.5 percentage points per year on average before expenses, 5 percentage points after expenses.

The estimate of 6.5 percentage points strikes me as absurdly high, particularly as the authors claim that figure applies to "all mutual funds." It's difficult enough to believe that for stock funds, but 6.5 points per year including municipal-bond funds and investment-grade taxable funds? But I'll table this discussion for a later column, as Scale and Skill also offers estimates of manager skill.

2) Unfortunately for investors, this skill is quickly arbitraged away. Funds that perform well are rewarded with new assets. These assets hamper the fund manager because it is easier to execute trades, find enough good ideas, and so on when running a smaller fund. At first, the inflow of assets only moderately harms a fund that is run by a skilled manager, so that the fund continues to post good performance, albeit not as strong as previously. This continued good performance encourages investors to place more money into the fund, which slows it more. Eventually, equilibrium is reached, when the damage caused by the asset inflows equals the level of the manager's ability, meaning that excess performance has been fully removed.

Once again, I'll let Scale and Skill address this thesis. It's worth pointing out, though, that to the extent that the argument is correct, it surely must apply very differently across investment categories. The effect of cash inflows on a micro-cap stock fund, or one that trades in an obscure corner of the convertible-bond market, is quite different than the effect on a blue-chip U.S. stock fund.

3) On average, younger funds will outperform because they are smaller. It takes time for a fund to reach its equilibrium point. During that period, the fund has an advantage because it is not carrying the full amount of assets that its manager's skill will eventually attract.

Surprisingly, Berk and Green do not directly make this argument. But it follows, unless there is reason to believe that the managers of younger funds are significantly less skilled than those who run older funds--and Berk and Green make no such suggestion and offer no such proof.

4) There's no chance for excess returns with older funds. My wording, not Berk and Green's, but that is the upshot. The exception would be if an established fund changed portfolio managers, and the new manager were significantly more talented than the previous manager. That manager should fare well until his or her higher level of talent becomes recognized by the marketplace and showered with the appropriate amount of new assets.

5) Because investors cannot perfectly measure manager skill, there is noise in the process. Sometimes investors will underestimate the ability of managers. Those funds are thus smaller than they should be. The fund thus will perform relatively well, which will bring it more assets, until it can no longer outperform. Conversely, investors sometimes overestimate a manager's ability. That fund will be too large and perform poorly, leading to redemptions until equilibrium is reached.

Think of a horse race, where each horse is handicapped by a different weight. The horse that wins the race is not necessarily the horse that is the fastest. In the Berk and Green framework, it is the horse that is most underestimated. It won because it carried too little weight for its abilities. Conversely, the horse that finishes last is the horse that was most overrated, and thus it was saddled with an unmanageably high burden.

6)Fund investors are rational. They are correct in believing that fund managers have skill, and collectively they allocate their money rationally, giving the greatest rewards to those managers who have the most skill, and proportionately less rewards to those who are less skilled. Unfortunately, fund investors don't benefit from their decisions, as the asset inflows that come from their decisions fully consume the extra gains that fund managers can deliver.

In which case, I suppose, index-fund proponents might argue that fund investors are not rational after all, since there is no payoff for all their work (aside from those who got into the right young funds). However, that's a different definition of rationality than is used by economics professors. The point is, with the Berk/Green theory, past performance is meaningful. It just ends up meaning something a bit different than investors realize.

Tomorrow's column will look further at the issue of fund size. Scale and Skill argues that fund size is less important than the size of the overall fund industry--a belief that affects both its estimates of manager skill, and its evaluation of the abilities of young portfolio managers.

Note: Thanks to Morningstar's Lee Davidson for his help, both in walking through the papers' math and in pointing out Scale and Skill's debt to the Berk/Green paper.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.