AgPa #16: Concentrated Stock Markets (7/7)

Mutual Fund Performance at Long Horizons (2022)
Hendrik Bessembinder, Michael J. Cooper, Feng Zhang
SMU Cox School of Business Research Paper No. 22-11 via SSRN, URL

This is the seventh and final AGNOSTIC Paper on the extreme concentration in stock markets. While the last six weeks focused exclusively on stocks, this one is an out-of-sample test with a somewhat different focus. As a reminder, the previous six papers document substantial concentration (positive skewness) in the distribution of long-term stock returns. This means that although the overall performance of the stock market was quite good, the majority of stocks underperformed. In fact, the positive net performance was entirely driven by a few very successful outliers (“big winners”).

In this paper, the authors find a similar pattern (although not as extreme) for equity mutual funds. The results have some important implications for the ongoing “Active vs. Passive” debate and are very relevant for investors. So let’s dive into it.

• Setup and Idea
• Data and Methodology
• Important Results and Takeaways
• Conclusions and Further Ideas

By now, all posts of this series are online and you can use the list to navigate. If you also read the previous posts (I hope so!), you know that I recommend to read the series chronologically. But as always, feel free to do whatever serves your needs.

• Week 1: Concentration in the US Stock Market between 1926 and 2019
• Week 2: Concentration in Global Stock Markets between 1990 and 2019
• Week 3: Dominance of the Tech-Industry?
• Week 4: Characteristics of Big Winners?
• Week 5: Identifying Big Winners Upfront?
• Week 6: Even Big Winners had Bad Drawdowns
• Week 7: The Same Pattern for US Mutual Funds

Everything that follows is only my summary of the original paper. So unless indicated otherwise, all tables and charts belong to the authors of the paper and I am just quoting them. The authors deserve full credit for creating this material, so please always cite the original source.

Setup and Idea

A lot of empirical studies on mutual funds examine a very simple question: do actively managed funds add value when compared to a passive market-cap-weighted benchmark? Or more practically: can professional fund managers “beat the market” and achieve better performance than a simple ETF on an index like the S&P 500?¹The appropriate benchmark depends of course on the asset class, region, and/or investment style. Since active funds charge substantially higher fees than their passive counterparts, they should perform better to compensate for that. Nobody wants to pay more for less, right?

In practice, however, this is unfortunately not the case. A large amount of empirical research unambiguously shows that the majority of active managers fail to outperform their benchmarks.²This general result is fairly robust and holds for different regions, time horizons, etc. Of course, this doesn’t mean that all active managers are bad. But statistically and in aggregate, investors were much better off with the passive benchmark. In fact, the literature also shows that the active managers who do outperform, usually fail to do it persistently and are very hard to identify upfront.³For example, Fama and French (2010).

The authors add to this evidence and link it to their previous analyses of positive skewness in the distribution of long-term stock returns. Unsurprisingly, the argument against active management is even more compelling in light of extremely concentrated long-term returns.

Data and Methodology

The authors construct a sample of 7,883 domestic mutual funds for the period between 1991 and 2020 in the US.⁴“Domestic” means that the funds only invest in US equities. The data comes from CRSP, a state of the art service for research on the US equity market.⁵CRSP stands for Center for Research in Security Prices and is an affiliate of the University of Chicago. More information here. Most importantly, this data is free from survivorship-bias. This ensures that funds which have been closed are not dropped from the database. As benchmarks for the active mutual funds, the authors use risk-free treasuries,⁶Yes, some equity funds even managed to underperform risk-free treasuries. the market-cap-weighted US stock market (not investable; “market”), and the popular “SPY” ETF on the S&P 500 index (investable).

Apart from the empirical part, the authors also introduce a simulation procedure to replicate the empirical distribution of mutual fund returns. While this is also an important contribution, I will skip it at this point and only focus on the empirical results. Like in the first two papers on stocks, many of the key findings are just interesting statistics and don’t require sophisticated econometric models.

Finally, the authors also apply their Shareholder Wealth Creation (SWC) methodology to the aggregate assets of their mutual fund sample. As a reminder, SWC measures the total wealth created by a particular security over its lifetime relative to some benchmark. This is particularly interesting for mutual funds because aggregate fund-flows are not properly reflected by long-term returns. A small fund that generates 10% per year but doesn’t attract new assets is a good investment for the existing investors. However, it doesn’t produce much wealth from the top-down perspective. That’s the difference between total returns and SWC.

Important Results and Takeaways

Longer investment-horizons lead to extremer return distributions – also for mutual funds

Like in the first two papers on stocks, the authors start with simple frequency distributions of mutual fund returns over different time horizons. The following charts show the results for annual-, decade-, and lifetime-returns. You can find the same charts for US- and international stocks in the first two posts of this series and I recommend to compare them in split view.

The shape of the annual distribution is fairly similar to that of stocks, though the dimensions are much smaller. The most frequent annual fund return is slightly larger than zero and most observations are in the range between -100% and +50%. However, even annual returns already exhibit a pronounced “right tail” with positive returns of up to 300%.

The “right tail” triples to 900% for the distribution of decade returns. However, the most frequent decade return for mutual funds remains close to 0%. This is an important difference to the distribution of decade stock-returns where the most frequent observation is close to -100%. Even beyond that, the distribution looks much more “normal” than those for stocks. Nevertheless, the impact of positive skewness is now observable: the majority of mutual funds generates returns of less than 100% but there are quite few that perform much better.⁷Just to put it into perspective: 100% return over 10 years equals about 7.2% compounded per year. This is pretty much in line with common estimates for the long-run performance of stock markets and shouldn’t require too much manager skill.

For lifetime returns, the distribution becomes again extremer. The “right tail” now ranges till 1,500% (which is an imposed upper limit by the authors) and there are quite a few very successful funds that reach this point. The most frequent return remains once again positive but close to 0%. Admittedly, those numbers are somewhat hard to interpret because the lifetime of funds is obviously different. However, the overall pattern is very similar as for stocks: the longer the investment-horizon, the extremer the return distributions and the larger the impact of a few very successful winners.

But why are the magnitudes so much smaller for mutual funds? Most likely because of volatility. I mentioned in the first post of this series that extreme return distributions are a result of high volatility and compounding over time. By combining multiple stocks into a mutual fund, volatility decreases and the resulting returns become less extreme. This is of course the powerful idea of diversification and in my opinion, this is best illustrated by the most frequent outcomes. For single stocks, the most frequent decade-return was close to -100%. For mutual funds, it was close to 0% and modestly positive. No matter if active or passive, I believe these are strong arguments for some degree of diversification.

Most active managers underperform passive benchmarks – especially over the long-term

As mentioned above, the literature on mutual funds is anything but a sales-pitch for active management. This paper is no different and the authors additionally show that the fraction of outperforming funds decreases substantially over longer investment-horizons. The following chart summarizes the results.

For the annual horizon, 39.3% of funds outperform the market and 41.1% outperform the S&P 500 ETF, respectively. Those numbers decrease to 34.1% and 38.3% for the decade-, and to 24.1% and 30.3% for the lifetime horizon. All of those are worse than a 50-50 guess. So any investor who picked a random mutual fund most likely ended up underperforming the passive benchmarks. Depending on the horizon, 20-30% of funds didn’t even outperform risk-free treasuries.⁸This is really not that difficult with a reasonably diversified portfolio of stocks.

So this is already disappointing, but we are by far not finished. For individual stocks, the majority also performed poorly but the few big winners were sufficient to turn the distribution into “net positive”. Unfortunately, this is not the case for mutual funds. On average, they performed worse than both passive benchmarks over all three investment horizons. Although the differences in annualized returns seem minor, they are statistically significant and compound over time.

None of these results are groundbreaking new but it’s interesting to see how the argument against active managers becomes stronger for longer investment-horizons. In my opinion, this is quite intuitive. It is very hard to get a competitive advantage in the stock market and it’s even harder to maintain it for a long time.

Those are of course only the high-level results. The authors present many more statistics and tests on fund characteristics and their impact on performance. For example, older funds performed better but are still significantly below the 50-50 threshold for both benchmarks. None of these analyses change the overall message: most mutual funds underperformed simple passive benchmarks and investors were worse off on average.

Despite their problems, we must be fair to active managers. Just because most of them underperform doesn’t mean that there are no skilled fund managers. Beneath those averages, there are some really good funds that outperformed the benchmarks by a great margin. But as mentioned above, other studies⁹For example, Fama and French (2010). concluded that those are very difficult to identify before they outperform (quite similar to the few big winners in the stock market). So the pure existence of skilled managers doesn’t change the fact that mutual funds are statistically a bad deal for most investors.¹⁰We all know that we could win in the casino, but this doesn’t make gambling smart.

Compared to the S&P 500, mutual fund investors lost about $1.3T between 1991 and 2020

So far, the authors haven’t been very kind to the mutual fund industry. But after the disappointing results above, the question naturally arises why most of those funds underperformed their benchmarks? It doesn’t get much better for fund managers…

There are two possible explanations (or a combination of both). First, the (well-paid) fund mangers simply have no skill. They invest in the wrong stocks, trade too much, have bad timing or whatever else. Second, they have skill and outperform but charge too much for it. For example, consider a fund that beats the S&P 500 by 0.75%-points but charges a management fee of 1% per year. Despite outperformance, the net return after-fees is still worse than the benchmark.

The authors examine those explanations and add back fees to look at the same statistics for returns before-fees. The numbers obviously improve but they are still not convincing. Over their lifetime and before-fees, 37.6% of funds outperformed the market and 45.2% outperformed the S&P 500 ETF (vs. 24.1% and 30.3% for returns after-fees). Both numbers are still significantly lower than the 50-50 random guess. In addition to that, this is a purely theoretical exercise. Investors ultimately returns after-fees. So it doesn’t really help them if their fund manager outperformed before overcharging for his/her services.

Overall, the results suggest three types of fund managers with respect to skill and fees. In any case, fund investors should pay close attention. Fees matter a lot and are perfectly observable at the time of investing (manager skill is not).

• Fund managers without skill who underperform even before-fees (55% of all).¹¹The fractions are based on mutual funds that beat the S&P 500 ETF over their lifetime.
• Fund managers with some skill but too high fees who therefore still underperform after-fees (15%).
• Fund managers with enough skill (or cheap enough fees) to also outperform after-fees (30%).

The overall conclusion is very simple: by and large, the mutual fund industry charged far too much for their services. The majority of fund managers lacked the skill to outperform a passive benchmarks and are thus (in theory) not worth any fees at all.¹²This statement only refers to performance. Fund managers may deliver other services to their clients that justify fees. The following table summarizes the tremendous scale of this issue.

The table shows SWC for all 7,883 US mutual funds from 1991 to 2020 relative to risk-free treasuries, the market benchmark, and the S&P 500 ETF. Compared to risk-free treasuries, US mutual funds generated wealth enhancements of $9,500.9B before- and $8,664.9B after-fees. This number sounds impressive, but it’s actually not that special. A reasonably diversified portfolio of stocks should outperform the risk-free rate over time. That’s just the well-known equity risk premium.

The comparison to the investable S&P 500 ETF is therefore more interesting.¹³The total market portfolio is not directly investable and more of a theoretical benchmark. Before fees, mutual funds added $936B in value compared to this ETF. Outperforming the S&P 500 is very difficult, so in aggregate, this is a strong accomplishment.¹⁴This result is different from the average underperformance of mutual funds in terms of total returns. SWC is an absolute measure and apparently, the largest funds in terms of assets performed quite well before-fees. However, the added value decreases to $-1,308.4B after-fees. So mutual fund managers added $936B in value but charged $2,244.4B¹⁵1,308.4 + 936 = 2,244.4 for this service. A great deal for fund managers but not for investors.

Conclusions and Further Ideas

Overall, the authors find the same pattern of concentration and positive skewness for the return-distribution of mutual funds. The magnitude is less extreme but the overall message remains: the longer the investment-horizon, the more extreme return distributions. In case of mutual funds, this has important consequences for performance evaluation and the old debate about “Active vs. Passive”.

The fraction of outperforming funds decreases substantially when measured by compounded returns over long horizons. This is a problem because practitioners often measure fund performance by (short-term) arithmetic returns, Sharpe-ratios, or alphas from regressions. The authors show that such measures can be misleading. For example, there are funds with positive regression alphas but long-term underperformance. Investors should therefore always look at the full compounded returns and don’t get blinded by other performance statistics.

However, I believe the most important results remains that (at least statistically) active funds were a bad deal for investors. Most fund managers didn’t add value in terms of performance and/or charged too high fees for that “service”. Except for those who have a legitimate reason to believe they can identify the few outperforming managers upfront (which is difficult), I would argue that most people should just go for a cheap passive index fund or ETF.¹⁶The debate about “Active vs. Passive” has of course more dimensions than just performance. For example, some people wouldn’t invest at all without the marketing of mutual funds. It’s sad that those people pay too much fees and underperform. But in the end, a bad mutual fund might be still better than just holding cash.

And this hopefully closes the circle to the first post of this series. The tremendous wealth-creation of the stock market was driven by very few successful companies and finding them ahead of others was very difficult. Therefore it was a perfectly sound strategy to just buy an index fund and enjoy the fact that you automatically owned the few big winners. By doing that, you also outperformed about 70% of the “professionals”. Past performance never guarantees future results, but I am quite confident that those patterns will persist in the future. Extreme success is rare and why should it suddenly become easy to identify it before others?

---

• AgPa #76: ESG Myth Debunking2024-04-15
• AgPa #75: Optimal Investment Committees2024-04-01
• AgPa #74: Peer-Reviewed Research is Not Helpful to Predict Returns – Really?2024-01-01
• AgPa #73: Country and Industry Momentum2023-12-17

---

This content is for educational and informational purposes only and no substitute for professional or financial advice. The use of any information on this website is solely on your own risk and I do not take responsibility or liability for any damages that may occur. The views expressed on this website are solely my own and do not necessarily reflect the views of any organisation I am associated with. Income- or benefit-generating links are marked with a star (*). All content that is not my intellectual property is marked as such. If you own the intellectual property displayed on this website and do not agree with my use of it, please send me an e-mail and I will remedy the situation immediately. Please also read the Disclaimer.