Christine Benz: Hi, I'm Christine Benz for Morningstar.com here at the Morningstar Ibbotson Conference. Today, I had the pleasure of sitting down with Martijn Cremers, he is a professor at Yale University and he is also the co-author of some recent research that attempts to identify how active certain mutual funds truly are.
Martijn Cremers: Active share is a new measure of active management. It compares the holdings of a fund to the holdings of its benchmark. So, if the holdings or the investments in the funds are very similar to the holdings of the benchmark we say the fund is not very active in terms of its holdings, and it could be closet index fund. If the holdings are very different, and so the fund invests in very different stocks than are in the benchmark, or the weights are very different, than it will have a high active share.
Benz: So one thing you found and this is very intuitively appealing is that the higher the active share, so the more active the manager is, the better their performance will tend to be. And by contrast those closet indexers with lower active share will tend to have a relatively worst performance. Let's talk about what you found there?
Cremers: Yeah, exactly. We always compared the performance of particular funds relative to the performance of its benchmark. We have basically two different results there. First, for the closet index funds, again funds with holdings that are very similar to benchmark, they underperform. The simple reason for this is that they are very similar investments, but they have much higher costs than a typical index-tracking passive fund or a passive ETF. The good news is that the funds whose holdings are most different from the benchmarks, they can do better than their benchmarks.
Benz: So, how about on a risk-adjusted basis? Do you find that those higher Active Share funds that also have better performance also have higher levels of volatility?
Cremers: They have slightly higher levels of volatility, but not strong enough to wipe out their gains. So, we look at two different measures there. On the one hand, we take the benchmark-adjusted performance, and we adjust for market exposure, we adjust for size exposure, for value growth exposure, and momentum exposure. And after adjusting for all of these things, all of these risks or characteristics measures, they still outperform.
The other thing we do is we look at the information ratio of their outperformance, risk-adjusted, divided by their volatility, and we find that the funds with the highest active share have the highest information ratios.
Benz: So another question is, where do you tend to find these closet indexers clustering? When I think of closet indexers I often think of that fund hewing closely to the S&P 500, is that where they tend to be focused in the large-cap space?
Cremers: Yes, indeed. There is a big difference across styles. Most closet index funds tend to be large-cap funds that are benchmarked to the S&P 500. Most of the small-cap funds tend to have generally high active shares. At the same time, it's still really important to look at active share because most of the assets are actually in funds that invest in large-cap funds.
Benz: So you have recently been transporting this data internationally and looking at whether you are observing some of the same trends in overseas funds, what are you seeing so far?
Cremers: So for our international sample, we have 30 different countries. We have over 10,000 different funds, and for that we look at four different styles of funds. We look at global funds. We look at regional funds. We look at main company funds. We look at other funds that invest in particular companies or particular sectors. When I say, global funds I mean funds whose benchmark implies that they invest in more than one region or more than one continent.
So for those four different styles we find in all four styles that the most active funds, funds whose holding are most different from the benchmark, that they outperform the benchmarks and that the closet index funds across all of those four different styles underperform their benchmarks. The results are strongest for global funds.
Benz: So one thing, users might say is that, well I can sort of look at active versus a more passively managed portfolio by looking at say R-squared. What do you think Active Share does differently and I'm sure in your view better than looking at those returns-based tools, like R-squared?
Cremers: So R-squared is actually very similar in spirit. And if you do not have access to holdings I would strongly recommend looking at R-squared actually. You can also look at both of them because they both capture a slightly different aspect of active management. Again if you don't have access to holdings and you cannot calculate active share, then R-squared will get you a long way there.
Benz: So, another related question is size, and as a fund grows larger you might tend to think that it might become less active and become more of a benchmark hugger. Is that something that your research bears out, as well?
Cremers: We do find a small negative relationship between active share and fund size, but in our view it's relatively minor. So for our U.S. funds the relationship between fund size and active share is basically flat for large-cap funds, until they have about $10 billion of assets. Once they reach $10 billion, we do find a relatively strong negative association.
Benz: So you think it's important to keep an eye out, but it shouldn't necessarily set off alarm bells, if a fund is growing?
Cremers: Yes. We find that the highest active share funds can outperform the benchmarks even if they're very large. At the same time for the very, very largest funds, we do find that the outperformance is a bit less than for the smaller funds.
Benz: Thank you, Martijn. Fascinating research, thanks for sharing it with us.
Cremers: Sure. Thank you very much
