Active Advantage 2026: What the Evidence Really Says About Active Management

Martijn Cremers
So thank you, and good morning. I'm so privileged to be here, honored to give this talk. Thank you to the organizers for inviting me. Again, thank you, all of you, for being here. I have been doing research on active management for about 25 years, and I'm very excited to share with you what I think I have learned.

So I'm Dutch. That's my strange accent and also the strange spelling of my name. I'd like to introduce myself as the guy with a J too many. Ij, I think, is only appearing in Dutch. Wonderful to be in my first trip here in Australia. I'm here way too short, but I will be back. Had a great walk through Melbourne yesterday. Beautiful city and also especially loved loved the river and everything around it. The subtitle of my talk is I read hundreds and hundreds of academic papers so that you don't have to. At the same time I'm going to show you, I'll give you some references to some of my survey articles where you can go and you can find all the references that I think you would be interested in, if you want to learn more.

My talk proceeds in three steps. As a starting point, I'm going to talk about what I call the conventional wisdom. What generally people think about active versus passive. And a simple question I'm going to ask is, well, the conventional wisdom by now is about 25 years old. What has academic evidence really shown in these last 25 years after the conventional wisdom first was manifested? I'm going to argue, show you a bunch of evidence, that the academic evidence suggests that the conventional wisdom is way too negative about the contributions of active management.

Then my second part is going to talk about what I call the three pillars framework. What are the three requirements for successful active management in the long term? Skill, conviction, and opportunity.

And the third and final part is where I apply this three pillar framework to active management and try to answer this question, where and when is active management most valuable?

I'm going to focus primarily on US equities and most of my evidence, because that's where most of the academic research has been done, is focused on US equities. But I'm also glad to talk about other parts of the investment world as well, especially if you have any questions on those. So let's start with the conventional wisdom.

The key study here is one of the most important papers among academic papers ever written by Mark Carhart published in 1997. He argued, showed evidence, that there is no support for active management. He argued that the average active manager underperforms. This is one of the best cited papers in finance. It's generally considered a top five academic paper. And I think it's one of those foundational papers that has led to general consensus about what I call the conventional wisdom.

One of the consequences of the conventional wisdom, of course, has this massive outflow of actively managed funds and a massive inflow into passively managed products. Now, this graph is very incomplete. It only looks at US equities, doesn't consider many other changes within equities. For example, portfolios have become a lot more international over this period. It also abstracts away from private markets, alternatives, real estate, and very importantly, bonds. This just looks at equities.

So just for U.S. equities, only looking at mutual funds and ETFs, in 1996, around the time that Carhart wrote his paper, 36% of assets in the U.S. equity mutual fund and ETF industry were passively managed. The most recent update was 59%. So that means for the last 25 years or so, last 30 years, massive headwinds for active management. It was difficult to run an active fund in 96’. I think it has become harder over the last 30 years because of all of those headwinds.

So let's dive a bit deeper. There's three components of the conventional wisdom. The first one, as I mentioned, The key conclusion by Carhart, the average fund underperforms.

Now, I hesitate to go here because I'm arguing here with the Nobel Prize winner, Bill Sharpe. Bill Sharpe had a powerful theoretical insight. He argued that there's a zero-sum game in investing, that in the aggregate, investors own the market, and so active managers who deviate from the market, in the aggregate, they will perform like the market, and they will therefore underperform by the amount of fees that they charge.

And that gave rise to this, what Carrick says, a logical conclusion, the average fund underperforms. The second component is that, yes, some funds can outperform, but that outperformance doesn't persist.

Then the third part is if there are skilled managers, then that skill is eaten up by fees.

So what I will try to do next is show you how the academic evidence in this 25 years subsequent to Mark Carhart writing this has, I think, largely overturned these three conclusions. If you want to check my references, this is the paper where we have hundreds and hundreds of references that you can check out. This is joint work with two other guys. And again, published in the Financial Analyst Journal. And a key reason to actually look at this paper is I think we did a careful job in just looking at this massive literature where I think our reference list, I think it may be the longest reference list ever published in the Financial Analyst Journal.

All right, so let's take a look. The average fund underperforms after fees. Let's first think about Sharpe's argument of the market as a zero-sum game

Pederson, in a paper published in 2018, he has an alternative theoretical model that I think, and you decide, but I think much more closely describes the actual world we live in from what Sharpe describes. Because the world is not consisting of only passive products and active managers. There's many other players in this market as well. There's a lot of people who trade on noise, a lot of speculators who are not fundamentally interested in the long-term value of companies. You can debate how much of active management is doing that, but the main point that Pederson makes is that there's so many noise traders that the Sharpe model doesn't really apply to our world.

One minor point in that is that the market, in some sense, is not a static thing. The market changes over time, what is in the S&P 500 is changing over time. It's actually quite surprising if you look at passive products themselves and you look at the turnover in passive products. That turnover in passive products may be much higher than you would expect ex ante. And I document that in detail in some of my papers.

Because the market is a lot more complicated than what Sharpe described, Pederson concludes, even theoretically, there's a strong point to make that active managers can be worth fees in aggregate because there's a very important asset allocation function that they provide.

Then let's look at the data. Let's look at Mark Carhart's methodology that led him to conclude that the average fund underperforms

Berk and van Binsbergen, I should be able to say his name because he happens to be Dutch, in a paper in 2015, they make a very simple move, but it's a very powerful one. I hesitate to criticize something that is fairly widespread in academic research, because I love academic research. But a lot of academic research is not always directly applicable to the real world. Here's one simple example. When you are doing performance evaluation, when you are comparing the performance of an actively managed fund, what do you compare it to? Are you comparing apples to apples? Are you comparing an investable product, an active fund after fees, to another investable product? Let's say a passive product, its returns after fees. That is not what most academic papers have done

And so what Burke and Van Binsberger do is to make that simple change. Let's compare apples to apples. Let's compare the returns of actively managed funds, not to these theoretical benchmarks, these factor models, that Mark Carhart uses, the Fama-French Carhart model, the standard workhorse in academic papers. But let's compare it to the after-fee returns of the most closely related Vanguard benchmark. And once they do that, the Carhart results disappears. There's no longer evidence for underperformance

So here in red, you see Mark Carhart's results. Then in blue, you see the results relative to the Vanguard benchmark. On the left, you see if you put all actively managed mutual funds and U.S. equities, you treat them equally. So you equally weigh them. On the right, you weigh them by assets. There's not strong evidence here of outperformance. Certainly, there's a little bit of underperformance evaluated. But the strong evidence that Carhart showed of underperformance is completely gone. And that's a sobering result. I'm not surprised by it.

In 2012, I published a paper with two co-authors, titled, ‘Should Benchmarks Have Alpha?’ Now, that's a rhetorical question, obviously, if you have a benchmark that is like the S&P 500 or the Russell 2000, those are effectively the market. So, of course, they should not have alpha. So what we did here is something extremely basic. We used the standard academic model, the Fama-French Carhart 4-factor model, and applied it to the S&P 500, and we applied it to the Russell 2000. Extremely basic and simple. What we found was that the standard 4-factor Fama-French model allowed very large abnormal returns that were statistically very strong over long periods of time. The S&P 500 outperformed, the Russell 2000 underperformed. So we need to be very cautious by reading academic papers for the purpose of performance evaluation.

In my research, I always try to focus on comparing apples to apples, so trying to use actual achievable returns on the right-hand side when I do performance evaluations. So compare active funds to, as close as possible, the returns on ETFs.

What about the SPIVA scorecard? I think there was a question in the previous session about this. The most recent SPIVA scorecard for U.S. equity mutual funds make this claim. 90%, according to the SPIVA scorecard for U.S. equity, actively managed funds, 90% of U.S. equity funds have underperformed over the last 10 years. 85% over the last five, 87% over the last three.

And currently, this is ongoing work, everything else I'm talking about today has all been published and vetted. So this is the one exception. This has not been vetted yet. So this is ongoing work, we're currently writing it up.

What we do in the new paper we're working on, we're taking a close look at the actual assumptions, the actual methodology of the SPIVA. Here are some choices that the SPIVA report makes that we think strongly overemphasize the underperformance. A key assumption that SPIVA makes, that S&P makes, is when they look at a 10-year period, they fixed all of the funds that are sample at the beginning of that 10-year period. And over time, many funds drop out for various reasons. A typical reason is a fund merges with another fund. Irrespective of how the fund actually performs over that period, if they drop out of the sample, S&P classifies that fund as underperforming. Even though, as we found, many of those funds did not actually underperform. Many did. The majority actually did. But a substantial minority of funds that disappear from the sample did not actually underperform their benchmark. So one of the things that we do is simply, well, let's not assume that. Let's actually look at the data and let the data decide if they underperform or did not underperform.

SAP makes many other assumptions. And so we're currently in the process of going through those systematically. They look at benchmarks. I think we can quibble a lot about how they assign benchmarks. They do not incorporate the size of the assets. They also assume that investing in the benchmark is costless. There's no fees. And they assume that there's only one beta that matters relative to the benchmark.

So these are tentative results where we've changed some of these assumptions. So let's change three. Let's look at the actual performance for the funds that exit. Rather than assuming that they underperform. Use the actual index performance after fees, and incorporate the assets, let's say, to keep things simple, at the beginning of the time period.

Here are the results. From 90% on performance, that drops down to 63%. Over a 10-year period, over a five-year, it dropped from 85% to 57%. And over a three-year period, it dropped from 87% to 52%. So the conclusion is, at very least, I believe, that the SPIVA reports vastly, vastly overstate the underperformance of actively managed funds.

What about performance persistence? Bollen and Busse in 2005 concluded that Mark Cahart just didn't really look very hard. If you're willing to go a little bit more to the extremes, they found significant performance persistence, especially for positive performance persistence, that funds that have outperformed in the past, that they continue to outperform in the future.

Kosowski et al., in a paper in 2006, they bring advanced statistics to the table. And advanced statistics, that means better statistical tests that will allow you to better determine whether or not there is actually positive performance persistence. Think about if there is performance persistence, would your statistical test allow you to document it? That means that your statistical tests have good power. You also want statistical tests to have good size, so that if there's really nothing there, it's all this noise, your statistical test doesn't tell you that there is something there. And so what Kosowski et al., do is they come up with much more powerful statistical tests than Carhart uses, and they find much stronger evidence for positive performance persistence.

So the bars here are a reflection of statistical uncertainty associated with the tests. In red is the Carhart results, and then in blue, the Kosowski. The much lower bars means that their statistical tests have much stronger power. They're able to much more accurately decide whether or not there is actually positive performance persistence.

So the Active Share paper, where we introduced the measure of Active Share, we published it in 2009. A couple of years ago, with two new co-authors, we decided to do an out-of-sample test. So our data ended in the original paper in 2003 or so. And now this new paper we started in 2007 using the eVestments database, but also all open-ended US equity mutual funds in US equities. We found that most of the results of our original paper still held robust in the subsequent period, especially for the investments products. For mutual funds, I'm going to tell you a little bit more about that in the next slide. But for mutual funds, the key results that we found, was very persistent over time, is that the funds that are most actively managed, with the highest Active Shares, that they show positive performance persistence.

So on the left here, you see the future performance of funds that had weak past performance, and they continue to underperform. And on the right, you see the future performance of funds that had a good past performance, depending on the Active Share. And you see for high Active Share funds, you see that they are significant positive performance in the future, and for low active funds, there isn't.

So just very quickly, what is Active Share? Active Share is the most basic measure of active management. It compares the holdings of a fund to the holdings of a benchmark. If the holdings are overlapping, that lowers Active Share. So think about all the securities that are both in the fund and the benchmark. All of those securities, if you long only, will create an overlapping holding. 5% in the benchmark, 3% in the fund, that's a 3% overlapping holding. 10% in the benchmark, let's say it's 10% in the fund, all those overlapping weights you sum up, deduct that from 100%, and you have Active Share. So Active Share is very literal, the percentage of weights in the fund that is different, that is not overlapping one-for-one with the benchmark.

One of the key results of our original paper in 2009 was showing that the conventional wisdom that funds tend to underperform was really driven by the low Active Share funds. So here's our original main result. It shows the abnormal performance adjusted for the benchmark for the future, predicting, depending on Active Share.

So you see that the lowest 60% of funds in terms of Active Share, they show significant underperformance. But for the top 40% in terms of Active Share, there was no evidence of underperformance. We did find strong outperformance, but that turns out to be depending on the time period. In some time periods you find it, for others you do not.

When we just replicated our study after 2003, we find that for US mutual funds, we did not find the outperformance. But again, no underperformance either. So the key point is that you need to distinguish funds that you have to distinguish between those who are truly active with really high Active Shares versus funds that are much less active.

After looking at the conventional wisdom, let's move to the three pillars framework. What are the characteristics of successful managers? So how do I think about it? This paper I published in 2018 is basically introducing that framework and kind of summarizes my academic research on active management at that point in time.

The first thing that successful active managers need is the skill to identify long-term investment opportunities. And we heard from different portfolio managers this morning of how they think about that. Skill is very hard to identify. Skill is hard to measure. And Active Share certainly is not a measure of skill. But if a manager has skill to identify investment opportunities, the manager also needs certain courage of one's convictions to actually apply that skill to create a portfolio that is meaningfully different. What the manager is doing needs to be difficult to do, otherwise, others will come in and do the same thing.

If the manager is successful, what's going to happen? A lot of money will flow in, not just in that management portfolio, but also in other portfolios that are doing similar things. You need to have certain barriers of entry to what the portfolio manager is doing. That's where the conviction comes in.

There's different ways to think about conviction. My most straightforward example is a manager who is very active, let's say high Active Share, and patient. That combination, I believe, is very, very difficult and really requires conviction of one's opportunities. It's difficult to be patient in an impatient world. Now, most active managers are not very patient. And most patient capital is not very active. Most patient capital is index funds, quasi-index funds, closet index funds.

So if most active managers are not very patient and most patient managers are not very active, what about the combination? Well, that's very rare. We heard about active quality investors. Those tend to be managers with high Active Shares and long holding periods, and that's a small group indeed. But because it's a small group, and assuming there are opportunities out there, there's not that many managers that pursue those opportunities. And that means that managers that do can be more successful. That is exactly what we found.

And third, you need opportunity. You need to be in a market, but also in an environment where you can be active over a sufficiently long period of time.

So I talked about Active Share as not a measure of skill. It just measures how different you are from the benchmark. And that doesn't necessarily require skill. However, having a high Active Share does require the other two, it does require conviction and opportunity

So let's do a quick thought experiment. Let's take a manager, let's call him Martin. Martin is a very overconfident, blunt Dutchman. Maybe I'm overly specific here. Martin has very strong convictions and lots of opportunity. So naturally, Martin, as a portfolio manager, is running a high Active Share portfolio. Alas, Martin doesn't have skill. How long would Martin survive?

Just a thought experiment. I think it will depend very much on how competitive this market is and how bucked up distributional channels are, et cetera. But in today's world, I don't think Martin would survive for very long, because running a high Active Share portfolio is incredibly, incredibly difficult.

This psychedelic chart shows you where the assets are in actively managed, large cap US mutual funds, US equity mutual funds, only large caps. In orange and blue is the assets in low Active Share funds, the funds with Active Shares of 60% or below. You see that around 2014, closet indexing, somewhat arbitrary, below 60% Active Share, had become somewhat uncommon. And since then, it has become dominant in this market, large-cap U.S. equity. The simple explanation is a very strong correlation across time is the concentration of the benchmark, right, the S&P 500 or the Russell 1000. There's a very strong correlation, 71% correlation between the percentage of closet indexing and the Herfindahl index in the market here.

So what this all basically means is it's become much, much harder to run a high Active Share fund in the United States for large caps. For mid and small caps, things look very differently. Active Shares generally are much higher. There's actually no problem with low Active Share as long as your fees are low. In many markets, it's difficult to have a high Active Share. Again, there's no problem with that as long as fees are proportionally low. Because the lower the Active Share, the higher the hurdle rate to earn back your fees.

To take, again, a very simple example, let's take a fund with expenses of 100 basis points above the index. Then what's the hurdle rate for the total holdings of the fund? Well, the hurdle rate here defined as earning back your fees. Well, the hurdle rate would be 100 basis points. Now I'm also going to tell you that the Active Share of this fund is 50% to keep things simple. What then is the hurdle rate for the active holdings? With the Active Share of 50%, half the holdings are identical to the benchmark. If both the benchmark and the fund have the same weight in the stock, that's not going to help you to earn back your higher fees. So what's the hurdle rate if the Active Share of 50%? How much does the 50% of the portfolio that is different, how much does that have to outperform for the investors to earn back their fees? Anyone? Exactly! 200 basis points.

And so are you paying the right amount for your funds? Well, my answer is that very much depends on your Active Share. So we introduce what we call Active Fee. It's the expense ratio divided by the Active Share. And if you do that, then many products all of a sudden don't look so cheap anymore, including many factor investing funds or enhanced index funds. They may not actually be as cheap as you think because their Active Share may be quite low.

We talked about the evidence both for the investments, the separate managed accounts, but also mutual funds, the strong performance persistence for active managers. Here's the result for active and patient managers. So here we look at the abnormal performance only of high Active Share managers. But then across high Active Share managers, we look at patience. What we found is that across the board, the high Active Share managers have performed well, but it's one standout group. That among the high Active Share managers, it was the patient high Active Share managers that did incredibly well. And our explanation is that they're doing something that requires the strongest conviction, with the fewest active managers trying to do this because there's relatively few, those who are able to do this, they have outperformed.

In my final minutes I have, I'm going to give some suggestions of how to apply this three pillar framework.

One important consideration is competition. How many funds are trying to do the same thing? And if there's too many funds trying to do the same thing, then there's less evidence of outperformance.

“The benefits to active management,” I'm quoting here, “are highest where deviation from a fundamental value is likely to be the largest and where competition is likely to be the lowest.” According to this paper, emerging markets would be one example where this is most favorable, then Europe, then non-US, and then finally US think of your large caps.

The paper I wrote with three others published in 2016, we looked at Active Share and explicit indexing in over 30 different mutual fund markets around the world. We have two key results. Where do active managers perform the best? On the one hand, we found that they've performed the best when there's more explicit indexing, when there's less active management, because there's so much indexing. That's certainly where we are today. So I believe today is actually a great time for active managers because there's many fewer that are trying to do this.

However, there's a counter argument here. We also found that, as a group, active managers performed much weaker if there is a lot of closet indexing. And right now, in the US large-gap equity, there's so many closet indexers. So that means that, as a group, we would expect that, in the aggregate, active managers would be less effective.

There's many other things that I don't have time to talk about, including that managers have different sets of skills. There's security selection, there's also risk management, and there's also trying to time the market, especially downside protection. So important studies have shown that active managers often have skills across the board, and that they particularly have this skill to adjust relative to where the market is.

A final thought, and this I think does describe the markets where we are today, when you think about opportunity, an important aspect of opportunity is how much cross-sectional dispersion there is, related to also Professor Bessembinder's insight. In some markets, there's a lot of cross-sectional dispersion, and other markets or at other times, stocks tend to all trade together according to some broad themes. The key point of this graph is that active managers tend to thrive during times when there is more cross-sectional dispersion, when stocks are not all trading the same way and there's more opportunity for security selection. So let me conclude, and then I look forward to engaging with your questions and comments.

A very thorough read of the academic literature in the last 25 years convinced us that the actual evidence for active management is much more positive than the conventional wisdom has told us.

Active Share is important to distinguish between managers that have portfolios that are very different versus portfolios that have large overlapping holdings with the benchmarks. I talked about skill and conviction, using patience as an example for conviction. And finally, we looked at opportunity, the when and the where of investing in active funds. Thank you, and let's go to questions and comments.

Gizelle Roux
Well, unsurprisingly, we've got a lot of questions. And so I'm going to try and summarise some of the themes that have come through that have repeated themselves over some of the questions. And these are very helpful questions because they're all going to assist in looking at funds themselves rather than necessarily just the pure academics of it.

So there's a lot of questions about concepts of Active Share. And I'm just going to give you three little points which you might want to raise. There's the question about Active Share versus tracking error, so which one and how to treat those differently or the same. The second one is just exactly how high an Active Share do you need or does that depend on the market? And the other one, that I think is quite prominent at the moment, we've actually seen a lot more cross-sectional dispersion in the market, and most recent data is showing that there's a lot more outperformance from active managers. If we go into a period where the US exceptionalism becomes less pronounced, as you point out, there is more alpha to be generated outside the US. Should that mean that we should get higher performance even with the same level of Active Share if we disperse ourselves across those markets?

Martijn Cremers
Thank you. Thank you. These are great set of questions. Let me start with the first one. I try to be brief, but that is not my strength. As a professor, I, you know, I often like to take a long time for my answer, so bear with me. I'll try to be brief.

So tracking error is the volatility of the difference in return between the fund and the benchmark. So how different the returns have been. Active Share ignores returns and only looks at holdings. So the two measures should be used together. They're very complementary.

In our 2009 paper, we go through a great detailed discussion about how to think about the two. One simple way to think about it is that Active Share ignores the correlation structure of the securities in the portfolio versus the benchmark. Tracking error volatility is very much about the correlation structure. And so key question is, think about the active positions, the positions in the funds that are different from the benchmark. Are those active positions well diversified or not? And so if those active positions, the ways in which the fund is different from the benchmark, if those active positions are very well diversified, you can very much have a high Active Share but not a high tracking error. If those active decisions are not very well diversified, you're going to end up with a high Active Share and a higher tracking error. So that's the key distinction between the two.

It's not in the question, but it's one of my favorite talking points. I don't think that tracking error is a good measure of risk. I think a key question to ask for investing in active funds is how does the overall risk profile of your total portfolio change if you add or do not add this particular fund? And so it could be the case, depending on the circumstances, that a fund with a high isocratic volatility or a high tracking error may actually make the overall portfolio better diversified.

Can you remind me of the second?

Gizelle Roux
Let's go to the question about market regimes, when active or Active Share does better and as we may diversify more across more regions.

Martijn Cremers
Yes, so the good news is that academic research has looked in detail about different skills of managers. And let's keep things simple and look at two different types of skills. One is the security selection skill. And for that skill to work, you need markets with more cross-sectional dispersion. And then a different type of skill is more about risk management and downside protection. Are you able to protect the assets? And so is the beta, for example, lower when markets go down, and is your beta going up as markets go up?

The academic research concluded that it's really the same set of managers that have skill in both. So it's not the case that some managers are good at one and the other. Actually, the managers that perform well, they really have skill in both.

The other thing I would say about the question about opportunity, I think that if you have a very skilled manager, why constrain that manager with a very narrow mandate? Especially if the opportunities will shift over time across different parts of the equities market. The most extreme would be like a global equity manager. But if the mandate is too narrow, then it's going to be very hard for that manager to actually apply the skill in a consistent way.

Gizelle Roux
Again, the question that's replicated a few times is, any comments on the current both participants in the market? In other words, has the rise in passive meant that we get lower returns from Active Share or not? In other words, because of the long-term thing.

And secondly, we've had now quite a number of markets, the US and frankly here (Australia) is even more pronounced, where you have some sectors that totally dominate markets and make it either, you know, people, if they don't participate in those, they implicitly have a high Active Share, even though they're not necessarily making that implicit decision.

Martijn Cremers
Yes. So for the first part of the question, we don't have that much evidence yet. So this is an area where we just need to do more research. The paper that we published in 2016, I think, is still the most comprehensive study that looks at this, but it's a bit dated. So I really need to update our own study.

What we found there was a two-fold result, that active managers do significantly better in countries with more explicit indexing. So when there's fewer active managers around. But they also, on the other side, is if Active Shares are generally very low in the country, then the performance is significantly lower. And we attribute that largely to fees. I think about fees as relative to the Active Share.

What's the second part of your question?

Gizelle Roux
That was against passive, and the other was the shape of the market. So we get high IT in the US or mega sevens, or here (Australia) we get banks and resource stocks dominating the index. And the manager looks like it's got a high Active Share because it doesn't participate in those sectors. So some ask, should we actually exclude those sectors when we're looking at?

Martijn Cremers
Yeah, no, that's a great question. I'm not sure I have a good answer to that. It really is, if you step away a little bit, is what is the right benchmark? And if a manager decides to not participate in an important sector of the economy, that, yes, will increase the Active Share. But it also means the risk profile of the funds may be very different than the market. So in my mind, it also raises the question, what is the right benchmark for that manager?

I also very much understand that for a lot of investors, their clients, they want the manager to have a sector distribution that is similar to the market. And when you are then in a market where a few names dominate the key sectors, you naturally will end up with a lower Active Share. So I understand why for US large caps, you see much lower Active Shares today. I think that's very understandable.

The flip side of that, I think then, earning back your fees becomes a lot more challenging. And I think that's one of the reasons where I think the lower Active Share funds generally have underperformed, unless they've also significantly lowered their fees.

Gizelle Roux
Again, following almost – all these questions follow a great pattern. So there's a lot of questions about what data, for example, should we use rolling period timeframes for assessing returns? Should we look at only seven- or 10-year even returns? It's pointed out that market regimes are now something like 10 years. What time frame should we use? And how should we look at the time frame if we're actually going to look at performance of managers?

Martijn Cremers
Great questions. So as an academic, I have certain advantages. I can just look at 25, 30 years of data. I don't have to worry about making a decision today. Because often when you make a decision, you have much more limited data. So it's very challenging.

I would say that in the end, when you make a decision to invest with an active manager, yes, you want to look at past performance, but you also have to look at the team that you're actually investing with. And so for me, Active Share and all these past performance, that's a starting point. But then does the team that you would invest with, is that the right team? Do they have the right investment philosophy that suits you? Do they have the right investment process? Do they maybe challenge your own thinking?

And so there's a lot of aspects here that as an academic I can't answer or that is too specific to, that I do not incorporate in my studies. So I think of active sharing and those other things I talked about as only a starting point, before you can even get to, I think, the key question of, is this the right manager or the right team?

Gizelle Roux
And just a side note to that, because that question came up afterwards, is should we look at these rolling returns or rather say we need to look at short-term returns to understand what they're doing and what their process is, but focus on the long-term outcome, if you like, rather than the returns? Because we may get short periods of under or out performance, I think that was pointed out. You can go through periods of underperformance, which may not be misleading.

Martijn Cremers
Yeah, so let's look at this group of funds that, in my research, has performed the best, the high Active Share, patient managers. Those managers have outperformed over long periods of time, but they also go through periods of one, two, three years of underperformance. So a key question, I think, is how does the manager react and what does the manager say? What does the manager learn on the performance over one, two, three years? And it's related to conviction, but also opportunity. Does the manager actually have conviction? And at the same time, how does the manager manage the risk of underperformance over one, two, three years of time, which is, for many clients, that's a substantial period of time. But I also think that it's really hard to draw conclusions on one, two, or three years. And so to get any type of confidence about results, certainly in academic studies, you need 10, 15, 20 years. And of course, that's not a luxury that you often have when making investment decisions.

Gizelle Roux
And again, the following question, which is relatively related to it, is size of the fund, stability of the fund, and implicitly also therefore, and I'll relate the question to it, is funds that set caps or trim positions or have loss of mandate, not necessarily because of bad performance, just because of structural issues. Were you able to assess how those kind of factors affect their performance?

Martijn Cremers
Yes, that very much, so like size of the fund is very dependent on the style or the mandate that the manager has. For large cap, it's not that much evidence that size matters a lot. Once you control for Active Share. So larger funds, some of them at least, especially in the extreme, they tend to have lower Active Shares. Once you get to small caps or less liquid markets, size becomes more of a challenge for performance.

Gizelle Roux
There's actually an article in the AFR today which we shared for a brief read with Martijn this morning.

Martijn Cremers
Oh, yes.

Gizelle Roux
And it sort of looks at the question of what causes, or do you get, you generate alpha from roughly, let's say, 30% of the market. I think you're suggesting it could be 50%, depending on how it's measured and how you arrange it. How do we create a portfolio that actually can capture that and say, and I think there's a question from the floor saying, it's just gonna be easier to be passive and I don't need to make excuses, you know, just why not capitulate? Your opinion.

Martijn Cremers
Well, the study you referred to is a fascinating study where the authors use artificial intelligence to describe trades that active mutual funds make. And they were able to explain using their machine learning model 71% of the trades that mutual fund managers make. At the same time, I think one of the other key results was that the trades that the model doesn't explain, those were actually trades that made money or were much more likely to make money than the trades that were explained.

So I would say your question depends on your view on the world. Is it worth pursuing? I think in today's world, the number of managers who actually is trying to analyze markets and try to understand where the investment opportunities are has shrunk. And I have a strong conviction based on this academic research that means there's a lot better opportunity now. I guess that's contrarian to where the flows have gone. But for me, it also just makes intuitive sense. There's many fewer managers still doing what the top Active Share managers have been doing. And so that means there's, on the one hand, more opportunity, and on the other hand, fewer people doing it. Because so much of the investment decisions are made not based on views of fundamental economic value. And a lot of investment decisions are based on something else. I think of it as a herd moving in and out, noise traders, speculators, short-term perspectives. But it does require a lot of discipline and conviction and analysis and, well also frankly, meetings like this to get to know the people who would run your money or your clients' money.

But yes, I'm very much a believer that it is possible, and I also believe that today's markets are actually more attractive for active management than they have been.

Gizelle Roux
The question was the rise of benchmarks, equity benchmarks. Everybody knows, reads about, hears about every single day. Very few people bother to look at the credit benchmarks every single day and when it comes to other asset classes, it's whatever benchmark the manager cares to nominate. Is there any way around us becoming so obsessed with equity benchmarks and should the others have similar ones or is the equity benchmark flawed or is it actually the best approach?

Martijn Cremers
Benchmarks are very important. Comparing, doing careful evaluation of a manager's past performance is very important and getting that benchmark right is an important piece of that. I like to use benchmarks that are very relatively broad. So if you're a large cap manager, using the Russell 1000. And then if you have very specific value of growth, you use components of that. But keep things somewhat simple. That doesn't work for all asset classes, but I think it works for the typical US equity managers in my research studies.

Things are very different often if you go to more particular markets or certainly outside equities. In recent years, I've been also looking at actively managed bond funds. It's a completely different world where the benchmarks in fixed income have way too many securities, and most of those securities are very illiquid. And so it turns out that the actual passive products in bonds don't actually follow the benchmarks very closely at all. So passive investing in bonds is actually an oxymoron. There is no such thing, because the passive products have Active Shares that are typically 50%, 60%. And they tend to underperform because they trade too much.

The difference between US equities and bonds is very vast, of course. And so there's many other equity markets, but getting the benchmark right is very important. And one simple way to think about that is if you can take the return of the fund and you subtract the benchmark, is there any systematic factor exposure left? And it's a potential red flag if there is. So to what extent does the benchmark actually capture the overall investment strategy of the fund? And if you just take the difference and you regress it with a bunch of systematic factors, you would like to see very little that still loads on that. That's a simple test if you have the right benchmark.

Gizelle Roux
We have a flashing light, but I'm going to lop in one extra question. In your assessment of funds, what is the one metric that you think funds should do? Higher Active Share, longer holding periods, is there anything else? Is there one thing you would like to see funds that we can measure do that would help us?

Martijn Cremers
To have an actual differentiated view on the market. Not necessarily contrarian, but what is different about your process? What is unique, different? How do you think differently than others? Because at the end of the day, we already know what the market thinks. What you're looking for in an active manager is someone who thinks differently. An Active Share is just one aspect of that. And that is a question that I think is the key for active managers is to come with a differentiated thesis that is a different perspective than what anyone else has given you.

Gizelle Roux
Thank you, Martin. Thank you for the questions. Thank you.

The views expressed here are those of the speakers. These views do not necessarily reflect the views of MFS or others in the MFS organisation. No forecasts can be guaranteed. These views should not be relied upon as investment advice, as securities recommendations, or as an indication of trading intent on behalf of any the Advisor.

Neither MFS nor any of its subsidiaries is affiliated with Martijn Cremers, Baillie Gifford and Orbis.

The information provided in this communication is of a general nature and should not be considered investment advice or a recommendation to invest in any security or to adopt any investment strategy. It does not take into account your objectives, financial situation or needs. Before acting on any of the information you should consider its appropriateness, having regard to your own objectives, financial situation and needs. It is not intended as a complete analysis of every material fact regarding any market, industry investment or strategy. In Australia, this communication is issued by MFS International Australia Pty Ltd (ABN 68 607 579 537, AFSL 485343), Orbis Investment Advisory Pty Limited (ABN 15 101 387 964, AFSL 237862) and Baillie Gifford Overseas Limited (ARBN 118 567 178), registered as a foreign company under the Corporations Act 2001 (Cth) and holds Foreign Australian Financial Services Licence No 528911 in respect of these financial services provided to Australian wholesale clients. Baillie Gifford Overseas Limited is authorised and regulated by the Financial Conduct Authority under UK laws which differ from those applicable in Australia. Orbis Investment Advisory Pty Limited and MFS International Australia Pty Ltd are regulated by the Australian Securities and Investments Commission.

 

67818.1