It may sound counter-intuitive, but younger, actively managed mutual funds actually tend to outperform much older, active funds as well as their benchmarks. That is one of the startling findings from a recently published paper, “Scale and Skill in Active Management,” by Wharton professors Robert Stambaugh and Luke Taylor, as well as Lubos Pastor of the University of Chicago. The paper was recently published in the Journal of Financial Economics.
While younger funds outperform their older rivals, performance tends to diminish as they age, according to the study. That is because the active management industry is getting bigger and new competitors are coming in that make it harder for existing funds to compete, they say. After all, too many people fishing in the same pond makes it difficult for each to catch a fish.
Knowledge @Wharton sat down with Stambaugh and Taylor to talk about the implications of their findings. An edited transcript of the conversation follows.
Knowledge@Wharton: Tell us what your paper is all about.
Luke Taylor: Our paper is about scale and skill in active management. We started with a simple idea, which is that a mutual fund’s performance depends not just on skill but also potentially on scale. So in order for us to say anything about skill, we need to make sure we first understand scale. The goal of this project is to understand the nature of returns to scale and active mutual fund management.
We test two different ideas; the first is decreasing returns to scale at the fund level. There, the idea is, as a fund gets larger, that causes the fund’s performance to drop. For example, because larger funds make larger trades, the push price is more and that hurts performance. That’s the first idea. The second idea is decreasing returns to scale at the industry level. The idea here is, as the overall mutual fund industry gets larger, that makes all funds’ performance decrease. And here, the logic is as there are more mutual fund dollars out there competing and chasing after mispriced stocks, that makes prices move and it makes it harder for any mutual fund to find a good trading opportunity.
“Investors should prefer newer active funds over older active funds.” –Luke Taylor
Our main result is strong evidence of decreasing returns at the industry level. In contrast, we find mixed evidence of decreasing returns at the fund level. So more simply, we find that it’s the size of the mutual fund industry that’s very important for performance Twitter . The size of an individual fund is less important. These results that we’re finding on scale have some interesting implications for how we think about skill. We offer a new way of measuring skill that adjusts for scale. With this measure, we show, for example, that skill has been trending upwards in the mutual fund industry over the last thirty years.
Knowledge@Wharton: How do you define an active mutual fund?
Stambaugh: Well, an active mutual fund is distinguished from a passive mutual fund in the sense that a passive mutual fund would just buy a basket of stocks that represent an overall investment in the stock market or segment of the stock market of what we typically call an index fund.
An active fund deviates from that portfolio by overweighting stocks that it believes are underpriced and offer superior returns and down-weights or sells off the stocks that it believes are overpriced or offer worse returns. So an active fund will trade more than a passive fund and its return will have what we call “tracking error” relative to the passive benchmark against which it’s judged.
Knowledge@Wharton: The active-management industry has grown dramatically since 1980. What is behind the surge in popularity?
Stambaugh: Well, the surge in popularity of active mutual funds really coincides with the surge in mutual funds overall. In fact, index funds or passive funds have actually had a somewhat higher rate of growth than active mutual funds, but the overall mutual fund industry is growing so much over the decades that active funds, too, have become much bigger in the sense that they control a much bigger segment of the U.S. stock market.
We believe that the overall trend in mutual fund growth just coincides in general with the increasingly smaller fraction of the U.S. equity market that’s owned directly by individuals. The growth in retirement plans such as defined contribution 401(k) plans is one big driver of that growth and there are other institutional factors that have been identified as leading to this surge in mutual fund size.
Knowledge@Wharton: So a larger active-management industry makes it harder for funds to outperform their benchmarks. Why?
Taylor: As there are more mutual fund dollars out there competing with each other and looking for mispriced stocks, then there’s more trading on those ideas that pushes prices, and that’s going to make it harder for any mutual fund to find a mispriced stock. The analogy we like to use is that if there are more people fishing in the same pond, that’s going to make it harder for any individual person to catch a fish.
Knowledge@Wharton: According to your research, the skill of active managers has increased over time. But how do you define skill?
Stambaugh: Well, as Luke mentioned earlier, in order to understand skill, we have to understand what the role of scale or size of industry in a fund is in generating performance. We define skill as the ability of a manager to perform relative to a benchmark, abstracting from the effects of industry size and his own fund size.
One way of thinking about that is we try to ask how much outperformance would this manager have if he were the only one operating in the stock market and didn’t have competition from all the other mutual funds and in fact didn’t have to worry about how big his own fund got? In some sense, it’s his ability to generate returns on the first dollar you would invest with that fund manager.
Knowledge@Wharton: Why do you think the skill of active managers has increased over time?
Stambaugh: Well, we don’t really know for sure but we do know that certainly the education of those entering the financial services industry has increased over time. So training has gone up. Also over time, new strategies get discovered. We have, for example, quantitative approaches to investment management that exploit technology and computing capabilities to much greater degrees than we could decades ago.
So, we believe things like training and technology lead to higher degrees of skill as well as the usual sort of learning-on-the-job effects. We believe that skill has increased for those reasons and possibly others; we don’t really know why for sure.
“New funds entering the industry have more skill … possibly because of better education or a better grasp on technology.” –Luke Taylor
Taylor: We show that skill has gone up over the last 30 years, yet performance has not. How is that possible? Our explanation combines two elements. The first is the mutual fund industry has been growing steadily over the last 30 years. The second element is decreasing returns to scale at the industry level.
In other words, funds are becoming more skilled, but the industry is also growing and those two effects tend to offset each other. … Yes, fund managers are more skilled today, but they have to be more skilled just to keep up with the rest of the pack.
Knowledge@Wharton: How does a fund’s size affect its return?
Taylor: As a fund gets larger, it starts to make larger trades. Those larger trades push prices more against the fund which hurts the fund’s profits.
Stambaugh: For example, if you have a fund that thinks a given stock is underpriced and wants to buy five thousand shares of that stock, it has to go and find someone willing to sell that stock in the market. If another fund identifies a stock as underpriced and wants to buy 50,000 shares of that stock, it’s got to find more potential sellers of that stock.
To do so, it either has to offer a bigger concession — willing to pay more for that stock to get it quickly — or it has to try to parcel out its purchases over time and take longer and perhaps thereby miss out on some of the underpricing that would be corrected in the meantime.
Knowledge@Wharton: Were you surprised fund size did not play a bigger role?
Stambaugh: I think we were surprised because [a] leading theory about mutual funds these days assigns a central role to the notion that as a given fund gets bigger, it should become harder for it to perform. I think we were somewhat surprised that even though our estimates pointed in a direction, we were not able to get strong, statistical precision associated with those estimates. That was one aspect of our study we found somewhat surprising.
Taylor: The logic for why a fund size matters is compelling, but there is a different logic pointing in the other direction. For example, imagine two mutual funds that follow exactly the same strategy. They make the same trades at the same times. Well, you can see that their individual sizes may not matter. Instead, what’s going to matter is their combined size.
Knowledge@Wharton: You also found out that the age of the fund matters – younger funds actually outperformed older ones. Why?
Stambaugh: Well, again, here’s one of these findings we don’t have a definitive answer for. But one potential reason is that younger funds could also come with younger managers on average, so we get back to this issue of increases in skill and sort of technological prowess perhaps being associated with age.
“Active funds in general, on average, have underperformed passive index funds.” –Robert Stambaugh
The other thing is that younger funds could be exploiting newer strategies that the market has not sort of caught onto and exploited to a significant degree. Again, this is one of those things like rising skill. We can’t really say for sure why it’s happened but we think it’s not implausible that these factors could contribute to superior performance for younger funds.
Taylor: We find that there’s also a second way in which fund age matters and that’s a relation over time between performance and age. We find that a fund’s performance typically deteriorates as the fund ages. If you take all these results together, they’re consistent with a simple story.
The story is that new funds entering the industry have more skill than the existing funds, possibly because of better education or a better grasp on technology. Because of the superior skill, the young funds outperform initially. They outperform their benchmarks and they also outperform the older funds. But performance tends to deteriorate as a fund ages and the reason is that as a fund ages, the mutual fund industry is getting larger and larger and that hurts everyone’s performance.
Knowledge@Wharton: Should mutual fund investors prefer younger, active funds over older ones?
Taylor: Investors should prefer newer active funds over older active funds. Each month, we form a portfolio that contains all of the youngest funds — say, funds that are of age between zero and three years. We also form a portfolio of funds that are 10-years-old and older and we find that the portfolio of very young funds significantly outperforms the portfolio of older funds.
Knowledge@Wharton: Do younger, active funds outperform index funds?
Stambaugh: Well, the bad news of this finding is that even though we find that younger funds significantly outperform older funds in the active fund universe, active funds in general, be they young or old, on average have underperformed passive index funds.
So our message would be to someone who has decided to invest at least part of their money in active funds, younger funds do seem to, on average, offer superior performance. But the overall averages do favor index funds.
And potentially because of this finding that the industry size matters, so as long as the … active fund industry is as big as it is, index funds may offer a better return. Were that industry to be smaller, were it to shrink — and indeed, index funds are growing more rapidly — there could come a time when active funds are able to keep pace and perhaps even offer some superior performance. But the evidence, for now anyway, would seem to be that index funds offer superior average returns based on historical evidence.
Knowledge@Wharton: Are you doing a follow-up on your research?
Stambaugh: Well, we are following up, actually. In fact, we have a recent working paper that also explores the role of fund size in a somewhat different way. We’re looking at the turnover of funds, how much they trade and to what extent that turnover seems to be related to performance.
Do funds earn more after they trade more? Does a given fund, when it trades more heavily, produce superior returns as a result? And there, we do find that fund size does seem to play a role. In particular, small funds do seem to exhibit a stronger relation between their turnover and their performance, suggesting they can more readily exploit these time-bearing opportunities to identify mispricing.