Hosting an industry conference? Ask us about including it in this ticker?
What do you think of our site upgrade?

A 401k Must Read: Mutual Fund Expense Ratio Myth Busted

October 09
02:42 2012

One of the most pervasive myths in the investing universe – and one that impacts the world of 401k plan sponsors and investors the most – states that “lower expense ratio mutual funds perform better than higher expense ratio mutual funds.” Like all absolute statements, this one is grounded in a partial, but by no means universal, truth. It gained national media attention when an August 2010 Morningstar study “How Expense Ratios and Star Ratings Predict Success” concluded lower expense ratio funds did better at predicting mutual fund success than Morningstar’s star rating system. The study went as far as to say, “In every single time period and data point tested, low-cost funds beat high-cost funds.”

That’s a pretty strong statement, but is it true? It’s not clear Morningstar still seems to think so. As recently as last year, at its June 2011 Investors Conference, Laura Lutton, editorial director at Morningstar Inc., said, “cheaper funds are far more likely to outperform.” (“Morningstar exec Laura Lutton: Stick with cheaper funds,” Investment News, June 8, 2012). There’s a reason why scientific researchers like to use hedging language such as “far more likely” rather than absolutist assertions like “in every single time period” – it’s too easy to knock the absolutist off its pedestal.

We took a look at the August 31, 2012 Morningstar data disk and ran a screen for all index funds with a best fit benchmark of the S&P 500, that were true no-load funds, that weren’t feeder funds, that had no short positions and that had either a “growth” or “growth and income” objective. This produced 50 funds (well, actually 51, but one was a “bear” fund meant to perform opposite of the S&P 500 and we eliminated it). We ran several different scatter graphs like these two and they all looked the same:

In both cases, the scatter graph appears as we would expect. The implied regression line (the dotted red line) is slopping downward, indicating a negative correlation between the expense ratio (on the x-axis) and the total return (on the y-axis). Indeed, the correlation in both sample sets is -0.92. In a situation where lower fees would always mean higher returns, the correlation would be -1.00. This is pretty close, but, as you can see, there are several data points showing cases where higher expense ratios correspond to higher returns. Still, while not supporting the August 2010 Morningstar study’s “every single case” language, these results are consistent with Lutton’s “far more likely” comment. This makes perfect sense. An index fund, particularly S&P 500 index funds like we’re looking at here, all have the same portfolio producing the same baseline performance. The only real difference in actual returns of different funds should be due to different expense ratios.

Unfortunately for this theory, however, the universe of active funds proves not a happy place. Using the same base Morningstar data disk, we ran the same screen, changing only the requirement that the fund be an index fund to the requirement that the fund not be an index fund. This produced 110 candidates (again, after removing a “bear market” fund). As you can see, the resulting scatter graphs look nothing like the index fund graphs above.

In the case of these actively managed funds, the scatter graphs show plenty of examples where higher expense ratio (x-axis) funds produced better performance (y-axis) – the exact opposite conclusion of the August 2010 Morningstar Study. Indeed, the correlation of the Annual 2011 Returns was much less (-0.18) while the correlation was actually positive for the 10 year data (+0.16). This positive correlation may be an artifact of survivor bias, but that doesn’t remove the fact that some high expense ratio funds earned higher performance numbers than their less expensive counterparts.

What are we to conclude? While expense ratios may be important when conducting due diligence on index funds, they may be less important – and even misleading – on actively managed funds. As we said in last week’s article (“What’s a Fair Fee to Pay a Fiduciary,”, October 2, 2012), “it would not be unusual to find a high expense ratio actively managed fund with better investment returns than a low expense ratio actively managed fund.”

And regulators echo this sentiment. The SEC doesn’t say to stick to low expense ratio funds, it merely (correctly) points out, “As you might expect, fees and expenses vary from fund to fund. A fund with high costs must perform better than a low-cost fund to generate the same returns for you.” Along the same lines, the DOL also stipulates fees should only represent part of an overall analysis of the appropriateness of an investment.

Alas, limiting ourselves to the written word, we cannot bust this particular myth with the fanfare of brilliant (and loud) pyrotechnics, but the results are nonetheless the same for every alert fiduciary: as your eyes can plainly see from the actual data presented in the scatter graphs, this myth is busted.

Interested in learning more about this and other important topics confronting 401k fiduciaries? Explore Mr. Carosa’s new book 401(k) Fiduciary Solutions and discover how to solve those hidden traps that often pop up in 401k plans.

About Author

Christopher Carosa, CTFA

Christopher Carosa, CTFA


  1. John Gingas
    John Gingas October 11, 10:35


    Nicely done! I’ve been making this argument for 3 decades and your data and comparatives kept the issue in CONTEXT. Thanks for your work.

    John Gingas

  2. Tom Rouse
    Tom Rouse October 11, 11:21


    Based on what I have seen, actively managed mutual funds in a majority of 401(k) plans not only have higher expense ratios, no-loads included, but also have higher levels of transactions/turnover costs when compard to a passively managed portfolio of asset class mutual funds. An index fund must mirror the index and threrefore will incure a higher number of transactions compared to a passively managed portfolio of low cost asset class funds.

    I wonder what your graphs would show if the transactions are included.


  3. Mason Flinn
    Mason Flinn October 11, 11:42

    Great article. I heard Jack Bogle speak last night. It is always good to hear a different opinion. Well done, Mason

  4. James Osborne
    James Osborne January 04, 13:58

    Is the data that you selected here across all asset classes? Each asset should be studied separately to get a good regression – otherwise returns will be all over the map (much like your scatterplot). It is hard to tell from your post just which funds you are examining.

  5. Christopher Carosa, CTFA
    Christopher Carosa, CTFA Author January 04, 15:30

    James, good question. The article actually did one better than your suggestion. Rather than just looking at asset classes (some which do not use the S&P 500 as a benchmark), we only looked at funds using the S&P 500 as a benchmark. That’s more relevant than asset class.

  6. Eric
    Eric May 21, 14:58

    Given that only about 50% of active mutual funds survive a 10 year time frame I imagine the regression stats are highly biased.

    Have you thought of doing an analysis using a survivorship adjusted database like CRSP? It would be a great follow up to this article.

  7. Christopher Carosa, CTFA
    Christopher Carosa, CTFA Author May 21, 17:05

    Eric: Good idea. Do you have access to that database? Is it as thorough as Lipper’s (who will also provide data without survivorship bias)?

  8. Eric
    Eric May 22, 16:20

    @Christopher I do not have access so I am no help in knowing which database would be better. I imagine even with the adjustments it would take significant time, effort, and thought to recreate the regression bias free. A worthy undertaking for someone willing to pursue it! Appreciate your efforts to further the conversation this subject (I may be an asset allocator but I think the day an advisor is not open to other investment strategies is the day they cease to be an effective advisor)

  9. David
    David June 06, 22:36

    Point is that the average (even most) investor can’t identify the ones that are going to perform better. If they picked 10 (diversified), on average they would be close to the line. Even picking into the 25% left of center would hit the target better.

  10. Christopher Carosa, CTFA
    Christopher Carosa, CTFA Author June 09, 13:38

    Actually, since the average actively managed fund performed well above the index, the average person – even if they randomly selected their fund – would have a better chance to outperform the index by going with active. This, of course, assumes they avoid over-diversification by picking more than one fund. The real point is, no one really knows whether active or passive will outperform in any random period. It’s less obvious from the data here, but other surveys have proved if either of the two philosophies has a statistical edge, its active. You’re not hearing that now because we’re in the cycle where passive is outperforming, and this is skewing all the snapshot-in-time performance measurements you’re currently seeing.

Only registered users can comment. Login is sponsored by…

Vote in our Poll


The materials at this web site are maintained for the sole purpose of providing general information about fiduciary law, tax accounting and investments and do not under any circumstances constitute legal, accounting or investment advice. You should not act or refrain from acting based on these materials without first obtaining the advice of an appropriate professional. Please carefully read the terms and conditions for using this site. This website contains links to third-party websites. We are not responsible for, and make no representations or endorsements with respect to, third-party websites, or with respect to any information, products or services that may be provided by or through such websites.