Why the SEC shouldn’t push index funds

By Felix Salmon
August 19, 2011
David Swensen's op-ed in last Sunday's NYT.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Peter Rudegeair, who has the memory of an elephant, thought of me when he saw David Swensen’s op-ed in last Sunday’s NYT. Back in March 2009, he recalled, I posted a short blog entry entitled “The Dangers of Listening to David Swensen”. In light of the new piece, Peter asked, what do I think now?

The answer is that Swensen still doesn’t look so good, despite the fact that he’s mainly advocating something perfectly sensible — investing in low-cost index funds. Still, there are two problems with the piece. His argument doesn’t really hold up, and it also looks suspiciously self-serving.

Swensen is right, of course, that the mutual fund industry has a conflict between making money for itself and making money for investors. But then he starts getting specific about how the industry takes money from investors and turns it into profit for itself:

Mutual fund companies, retail brokers and financial advisers aggressively market funds awarded four stars and five stars by Morningstar, the Chicago-based arbiter of investment performance…

In 2010, investors redeemed $152 billion from one-star, two-star and three-star funds and placed $304 billion in four-star and five-star funds. In the crisis-scarred year of 2008, even as investors withdrew $174 billion from one-star, two-star and three-star funds, they added $47 billion to four-star and five-star funds. Year in and year out, flows to four-star and five-star funds prove remarkably resilient and overshadow flows to the three bottom categories.

This churning of investor portfolios hurts investor returns. First, brokers and advisers use the pointless buying and selling to increase and to justify their all-too-rich compensation. Second, the mutual fund industry uses the star-rating system to encourage performance-chasing (selling funds that performed poorly and buying funds that performed well). In other words, investors sell low and buy high.

Ill-advised buying and selling of funds costs the investing public a substantial sum. In 2010, Morningstar found that if mutual fund investors in 2000, as a whole, had simply bought and held their funds for 10 years, their investment outcomes would have improved by an average of 1.6 percentage points per year. That 1.6 percent may not sound like much, but it adds up to tens of billions of dollars per year.

Swensen is right that people are more likely to sell underperforming funds, and buy outperforming funds. And he’s also right that there are fund managers out there who recommend precisely that course of action: I wrote about one of them a few weeks ago. But he’s wrong that the tendency to rotate from underperforming funds into outperforming rival funds is responsible for that 160bp of negative alpha. If that was the case, all you’d need to do is trade the opposite way — sell the five-star funds, buy the one-star funds — and you’d make a fortune.

The fact is, as Swensen well knows, that within any given sector, the difference between the best-performing funds and the more mediocre performers is almost always modest. If you look very carefully, over five years or longer, then you might be able to discern a performance differential between, say, five-star and four-star funds. But as a rule, it’s nothing to write home about. So unless you pay massive up-front fees every time you buy and sell funds, being rotated from one value fund to another, or from one emerging-market fund to another, is unlikely to do massive harm to your net worth.

What’s more, investors are actually pretty smart win it comes to picking mutual funds. They’re cost-conscious: the lower the fees, the more popular the fund is likely to be. And if you look back at the funds which investors piled into before Morningstar existed, they tend to be precisely the ones which would have got 4- and 5-star ratings, and which over the long term turned out to be the best bets.

So the 160bp of underperformance — the difference between investor returns and market returns — doesn’t seem to come from the area that Swensen identifies: it’s a lot more complex than that.

A look at the report in question is probably useful at this point; Swensen doesn’t provide a link, but I’m nicer. What we’re looking at here is a gap between investor returns, on the one hand, and the average return of a mutual fund in that sector, on the other. Morningstar’s John Rekenthaler confirmed for me that we’re looking at the unweighted average of all the funds in a sector, regardless of size — so in the smaller sectors, if there were one or two small funds which did extremely well, that could easily throw off the average considerably. What’s more, there’s significant survivorship bias over the course of a decade: the underperforming funds tend to close and not get counted in the figures, while the best performers all live on.

All of which means that Swensen isn’t right when he characterizes the average fund return as the return that investors would have received had they “simply bought and held their funds for 10 years”.

And there’s certainly no evidence to suggest that buying actively-managed funds rather than index funds will cost you anything close to 160bp a year. As I say, mutual fund investors in aggregate are pretty smart: they tend to buy funds which might not be quite as cheap as the cheapest index funds, but which aren’t far off. Even once you take into account the idea that most fund managers underperform the index, you’re still not going to get to that 160bp level.

Instead, Rekenthaler has his own theory for the underperformance, which rings much more true to me. We’re not seeing investors rotate within sectors, we’re seeing them rotate between sectors, something he calls “category chasing”. When tech stocks or real estate or gold are hot, investors pile in; when those sectors crash, the same investors pile out. “Within a category, people are making rational decisions,” says Rekenthaler. “The problem is that they’re chasing categories.”

I hasten to add here that Rekenthaler doesn’t have the empirical data to prove this thesis in an iron-clad manner. But it’s more credible than the Swensen thesis, since it seems as though it’s better at explaining the big differences between stated mutual-fund returns, on the one hand, and the actual returns that investors see, on the other.

And if Rekenthaler is right, then Swensen’s prescription — everybody should just pile into index funds — does no good at all, and in fact might even do harm. As index funds steadily get supplanted by ETFs, the ease of trading in and out of them, from one hot sector to the next, becomes all but irresistible: everybody becomes a market timer. I have a very expensive broker, who performs no useful service for me at all, beyond simply being expensive — and that’s a good thing, because if I know that trading is expensive, then I won’t do it, and if I don’t trade, I’ll make more money over the long term. Fees are bad — but sometimes ease of trading can be bad, too, as Emanuel Derman notes today.

On the other hand, there’s one group which would benefit greatly if Swensen’s advice were taken and the SEC put a lot of effort into pushing all individual investors into index funds. And that’s the kind of hedge-fund managers who Swensen invests in. Because there’s a paradox of index funds: while they make perfect sense for any given individual, it would be disastrous if everybody invested in them. The whole point of having a market is that you need millions of people all competing against each other to allocate capital most efficiently. Investing in an index fund is a way of free-riding on the collective wisdom of active investors, and so you need to have more active investors than index-fund managers in order for the market to continue to work.

The more people that invest in index funds, the easier it becomes, in theory, to beat the index — which is Swensen’s job. And while I’ll continue to recommend that any given individual should invest in index funds, I wouldn’t suggest, as Swensen does, that there’s a public-policy benefit in pushing everybody to do so.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/

To me as a public investor, the most insidiously pro-Wall Street SEC bias may be the ubiquitous “past performance does not guarantee future results,” as required by Rule 482. It is a negative pregnant that suggests that past performance, while no guarantee, is actually relevant. If you believe in any of the the efficient markets/random walk theories, the only accurate statement is that past performance is completely irrelevant. Failure to state this clearly has allowed the mutual fund industry to pitch its essentially worthless active management and associated fees as something worth paying for. This has imposed billions of dollars of fees/costs on small investors over many years, with no corresponding benefit.

Posted by maynardGkeynes | Report as abusive

Alexi Savov from my department says flows in and out of index funds also play a role:
http://pages.stern.nyu.edu/~asavov/index _files/FreeForAFee.pdf

Posted by guanix | Report as abusive

Felix, you’re one of my favorite bloggers, but this was shoddy work.

1) “But he’s wrong that the tendency to rotate from underperforming funds into outperforming rival funds is responsible for that 160bp of negative alpha. If that was the case, all you’d need to do is trade the opposite way — sell the five-star funds, buy the one-star funds — and you’d make a fortune.”

Even if mean reversion existed, buying one-star funds and selling five-star funds could still be a bad idea if one-star funds charge high enough fees–that is, have persistently negative alpha. However, mean reversion is real. Underperforming asset classes and stocks have a tendency to bounce back up over the long run. Thaler and De Bondt conducted the most famous study on this effect. They found that buying multi-year loser stocks outperformed over the next few years. Hot stocks and asset classes tend to outperform for about a year but underperform beyond that.

2) “The fact is, as Swensen well knows, that within any given sector, the difference between the best-performing funds and the more mediocre performers is almost always modest.”

This is false. I just pulled up a list of all mutual funds in Morningstar’s Large Value category. The best 1-year performance is about 23%. The Vanguard Value Index’s performance, a decent proxy for the average, is about 3.53%. I found similar spreads with other categories I checked.

3) “And there’s certainly no evidence to suggest that buying actively-managed funds rather than index funds will cost you anything close to 160bp a year.”

Untrue. By my own calculations, the asset-weighted expense ratio of the entire mutual fund and ETF industry was 0.88% in 2010 excluding index funds (yes, I have access to a comprehensive mutual fund database, and, yes, this is the type of analysis I do for a living). This figure doesn’t include inflated commissions, bid-ask spreads, and market impact costs incurred by active funds. Index funds’ asset-weighted ER was .26%.

4) “And if Rekenthaler is right, then Swensen’s prescription — everybody should just pile into index funds — does no good at all, and in fact might even do harm. As index funds steadily get supplanted by ETFs, the ease of trading in and out of them, from one hot sector to the next, becomes all but irresistible: everybody becomes a market timer.”

This speculation is not only implausible, but not supported by a shred of evidence. 401(k) participants are a lazy bunch, as study after study has demonstrated.

5) “On the other hand, there’s one group which would benefit greatly if Swensen’s advice were taken and the SEC put a lot of effort into pushing all individual investors into index funds. And that’s the kind of hedge-fund managers who Swensen invests in. Because there’s a paradox of index funds: while they make perfect sense for any given individual, it would be disastrous if everybody invested in them.”

This is absurd. First of all, if most people invested in index funds, they’d still be better off as a whole simply because they’re paying less. Second, just think about what you’re saying. Swensen’s master plan is to 1) write a NYT op-ed 2) ????? 3) get everyone to index to the point where the market becomes inefficient (nevermind this could never happen even if the government mandated indexing) and 4) make it easier for him make even more money for…Yale. And finally, Swensen indexes much of his conventional asset class exposure, using hedge funds for less liquid asset classes. The Yale endowment is massively overweight in alternative asset classes beyond the reach of individual investors.

Just a bad post, Felix.

Posted by oculus | Report as abusive

Kenneth French (the Cost of Active Management) estimated the difference at ~60 bps, back in 2009.

http://onlinelibrary.wiley.com/doi/10.11 11/j.1540-6261.2008.01368.x/full

Posted by Th.M | Report as abusive

I agree with oculus that you make an error on the 160 bps difference. Your analysis assumes that you could short the actively-managed funds and buy the market and make that 160 bps difference. However, that difference comes mainly from fees for both transaction costs and the portfolio manager. If you short the actively-managed fund (is that even possible for non-ETF’s?), you do not get the fees. Unlike a pure stock, if you borrow a stake in an actively-managed fund, someone still has to pay the fees to the mutual fund company on that share.

The logistical issues with short-selling get at why I believe so strongly against the strong forms of the EMH. The EMH assumes that all prices reflect all publicly available information because, if the securities are overpriced, then short-sellers would make a killing. However, this analysis assumes short-sellers have infinite money and infinite time, but nearly all investors do not have the money to make margin calls if the stock turns against their short. Nearly all investors also do not have the sheer patience to wait years for the market to find the last buyers and eventually correct itself. Only the most liquid hedge funds, with the most patient and understanding investors, can make such bets.

For all those reasons, bubbles are an inherent feature of any security market. For most real-life investors (not the paper investors in economic papers), they earn much more money for themselves chasing relative returns instead of absolute returns. Swensen and Buffett may have been wise enough not to invest their money in the Tech bubble of the late-90′s, but the regular retail broker would have lost tons of clients if they invested on a completely fundamental basis. Swensen and Buffett can make outsize returns in part because they have very understanding, long-term oriented investors.

The main issue with index funds then is that it buys in all market conditions. According to the index fund dogma, the passive investor should invest in an index fund whether it’s the peak in 2000 or the trough in 2009. The index fund dogma, however, relies on the EMH dogma that no bubbles can exist and therefore there’s no reason to not invest in the market.

For passive investors, a better approach is to realize that markets can and do form mispricings for real-world, behavioral-finance reasons. IMO, the best approach for passive investors then is still Benjamin Graham’s method of allocating a portfolio between bonds and stocks based on relative yields. This approach would have put less money in the market in 2000 and more money in the market in 2009. I do not know of a systematic analysis of how it would do compared to just dumping all savings into an index fund, but with the reality of bubbles in the stock market, it makes much more sense to me.

Posted by mwwaters | Report as abusive

“The more people that invest in index funds, the easier it becomes, in theory, to beat the index”… Really? Maybe the market may get inefficent, but are there any emperical evidence or theoretical argument that it’s easier to beat the index in a less efficient market? To beat the index, you must be at least better than average. Is it easier to be above average amoung 100 people than among a million people? I don’t think so.

Posted by ychu | Report as abusive


According to the strong forms of the Efficient Market Hypothesis, you are correct. If the EMH holds, then if 100% of investors go into index funds, somebody with infinite money would enter the market and correct all the mispricings. If some stocks were too low, the mythical investor would buy all the stocks up to when their correct price. Through shorting, the mythical investor would also short the stock until the stock went down to its correct price.

To a certain extent, this is true. When there was blood on the streets in late-2008 and early-2009, many wise investors followed Buffett’s Oct. 2008 column and bought stocks. Stocks actually continued going down after his column, but in the long run his advice would have made a lot of money.

But this argument gets back to the fundamental issue with the EMH: It assumes the smart investors have both infinite money and infinite time. In the real world, if 100% of investors went into index funds, then stocks would experience horrific mispricings. By definition, the mythical investor would not have any money to correct the mispricings.

So, what happens then if 90% of investors invest in index funds? The EMH would say the remaining 10% of investors have infinite money and time and would price the market correctly. However, the 10% investors in real life do not have infinite money and the mispricings are not corrected automatically. That leaves active smart investors who do have a ton of money, like Swensen with the Yale endowment, to make a lot of money off the mispricings.

The answer to index funds is not actively managed funds. Indeed they have conclusively been shown to have negative alpha after fees. But the answer is also not to always plow money into index funds year after year, regardless of the price of the index or the time horizon. The answer is a more heterogeneous approach to investing, which realizes can and will be mispriced due to the lack of infinite money for smart active investors. Plowing all money into index funds merely zeros the market’s historical risk premium and makes sure the historical returns don’t happen.

Posted by mwwaters | Report as abusive

Thanks, mwwaters, but my point has nothing to do with the market efficency. What I wanted to say is, as an active investor, you are not really competiting with an index (which by definition will earn an average return regardless of market efficency), but rather you are competeting with other active investors. Only if other active investors lose (against index), you have a chance to win (against index). So, my question is how you can be sure you are smarter than other active investors, and if the number of active investors gets smaller and the market gets less efficecient, would you become more confident that you are smarter than others? If almost everyone goes index, the remaining active investors may be people like Swensen and Buffett. Would you want to compete with them?

By the way, here is a quote from Buffett on active mutual funds.

“Dividends continue to run about 2%. Even if stocks were to average the 5.3% annual appreciation of the 1900s, the equity portion of plan assets – allowing for expenses of .5% – would produce no more than 7% or so. And .5% may well understate costs, given the presence of layers of consultants and highpriced managers (“helpers”).
Naturally, everyone expects to be above average. And those helpers – bless their hearts – will certainly encourage their clients in this belief. But, as a class, the helper-aided group must be below average. The reason is simple: 1) Investors, overall, will necessarily earn an average return, minus costs they incur; 2) Passive and index investors, through their very inactivity, will earn that average minus costs that are very low; 3) With that group earning average returns, so must the remaining group – the active investors. But this group will incur high transaction, management, and advisory costs. Therefore, the active investors will have their returns diminished by a far greater percentage than will their inactive brethren. That means that the passive group – the “know-nothings” – must win.”

Posted by ychu | Report as abusive

“So you need to have more active investors than index-fund managers in order for the market to continue to work.”

I am not an expert on EMH but 50:50 is no special point in the ratio of active to passive investors.

So long as you have a FEW active investors such that the liquidity is sufficient to iron out most inefficiencies then the market would be considered to “work”.
Of course 1 active investor is probably too few but there is no dividing line at 50% of volumes.

Posted by TinyTim1 | Report as abusive

Many of Swenson’s points are valid, but there are several different potential paths to reach his objective for individual investors.

1. Understand your overall time horizons and acceptable risk for various pots of money. I have been teaching my kids to take NO risk with short-term cash savings (FDIC accounts) while dollar-cost averaging into heavily equity weighted Target-Date funds early in their careers for their lifetime tax sheltered savings. This decade as they move out into the work force is likely to be one of the best investment opportunities of a lifetime when they look back 50 years from now. Dollar-cost averaging through regular payroll deductions through the massive troughs should yield very good long-term results.

2. Focus on expenses and a long-term strategy focus for the funds that they are in. For various reasons, different members of our family have money in Vanguard Target date funds, T Rowe Price Target Dates, Wells Fargo Dow Jones Index Target Dates etc. All of these have relatively low costs (less than 80 basis points per year) and a long-term, diversified strategy. No, I don’t know which of these will do the best in the long-term although I will guess that the Vanguard ones will do marginally better because they do have a fews 10s of basis point cost advantage.

2. Pay attention to the big picture. One of my kids only has MassMutual target date funds in their 401k – they are a bit more expensive but a very generous employer match more than makes up for the difference. In the end, if (when) he leaves that job, it can get rolled over into a lower cost alternative.

3. Even index fund strategies often have an active component. The Vanguard LifeStrategy series typically have a 25% holding of their Asset Allocation fund which has an active strategy of allocating between cash, S&P 500, and Total Bond index funds. You get the cost advantage, but you are still subject to manager error. In the short-term, Asset Allocation got hammered over the past few weeks because they were almost 100% in equities. The jury is still out on how Asset Allocation will do over a lifetime, but even massive blunders on their part would only impact 25% of the LifeStrategy fund.

4. Save enough to be able to weather the long-term storms and needs. 15% annual savings is pretty much essential. 20% to 30% means you should be able to do ok in just about any long-term market condition and long-term needs. One of the big adjustments coming in the Big Reset decade ahead is much more of the population shifting to this saving structure as the consequences of increased job and social safety net uncertainty hits home.

In the end, control costs, understand your acceptable risk profile, save money, and find diversified fund/ETF investments that match your long-term goals.

Posted by ErnieD | Report as abusive

Here’s a humorous look at what happens if everyone invested in mutual funds: “3 Reasons to Outlaw Index Investing Right Now (and One Selfish Reason Not To) in 5 Acts” http://bit.ly/q6kGuq

Warning: If you don’t know who Groucho Marx is, you won’t get it.

Posted by FiduciaryNews | Report as abusive

If you hunt game birds such as quail or pheasant, you use a trained dog to flush them, to cause them to fly up in front of you and your shotgun. Wall Street works the same way. Very low fee index funds that do not generate brokerage commissions are the answer. Costs are the largest destroyer of performance for ordinary investors. Avoid all fees, commissions, transfer charges, or any other charge and stay in a very broad, internationally diversified low cost mutual fund. Or feed the buzzards.

Posted by txgadfly | Report as abusive

ychu, I guess my point is that it’s too simplistic to say active investors just have to beat other active investors. It’s too easy to look at investing as purely a zero-sum game.

Yes, it IS a zero-sum game if you just average out all alphas. By definition, the average of all alphas (before fees) has to equal zero. If an investor does nothing but look at alpha, then you are correct that active investors cannot gain from less active investors.

But alpha is actually a pretty meaningless statistic if you think about it. If you average out all equity investors in either 1999 or 2009, their alpha will average out to zero. But the investors in 2009 will do far better than 1999 because stocks were fundamentally underpriced in 2009 and overpriced in 1999.

Fundamental mispricing has absolutely nothing to do with alpha. Rather, it has to do with how much active liquidity exists in the market to correct mispricings. In early-2009, so many investors had bailed that the smart money pouring into stocks could not keep up with the money going out. And if there is less smart money because more money goes to index funds, then investors Swensen like Swensen have an easier time picking up fundamentally mispriced stocks.

Posted by mwwaters | Report as abusive

To elaborate that last paragraph, it’s important to understand why fundamental mispricings are so important to Swensen vs. Alpha.

If the EMH holds, then ALL investors, including Swensen, are merely playing a game of coin-flipping. Over 10 years, 1 in 1024 investors will have “beat the market” before fees even if they were just using darts to pick stocks.

To make real, sustainable gains, there has to be some fundamental reason why an investor can beat the market. Some hedge fund managers seem good at riding speculative bubbles, having the knack to realize when the bubble has run its course and selling at the top. But typically, for an investor to make real gains, it needs fundamental mispricings. As I argue above, if there’s less active money in a market, that should generate more fundamental mispricings.

Posted by mwwaters | Report as abusive


Felix’s later post on S&P 500 correlations between stocks indicates that we are in a market that continues to be dominated by program traders reacting to macro news, especially government intervention activities such as ZIRP and QE. So currently, the indexes are floating along driven by fear and greed as expressed by the quants and their high-speed toys.

At some point in time, we will have a less macro-driven market which will probably occur when the market is in the late stages of entering a deep hole. At that time, the Graham and Dodd investors will make out like bandits in the ensuing decade.

The equity side of my investments is a mix of broadly diversified index funds and value funds with long, proven track records with either the original management teams or proven management transitions. These are typically funds that tend to lag during manic bull market blow-offs but do well through downturns and early years of a bull market.

Posted by ErnieD | Report as abusive

I love all these hopeful articles about how Index funds are so great. Problem is, Index Funds also charge fees. That means for their returns to be the same as the index, they have to take more risk than the index as a whole represents; for their risk to be the same as the index, their returns will always be beaten by the index. Unless they charge no fees, they will never do what it says on the tin.

Then there are all these seemingly perfect arguments why Index Funds are better than Active funds – the average return argument. Of course, that works if you compare the Index Fund to the whole index, but most advisers I know are smart enough to look at the consistency of above average returns. If you do this, you will see that the same fund managers inhabit the below average funds in an index, and another group of companies hang around pretty much most of the time in the above average return area. Comparing the average Index Fund to the average “above average” fund will most always result in the active fund pick beating the index.

Then there’s the timing issue: Morningstar averages (and their star system) measure date to date averages, not average consistency. While they can be very useful indicators anyway, it is always important to remember that the whole performance of a fund can be significantly enhanced by one lucky year, or even one lucky week, which could fall at the beginning or end of the date range assessed. Looking at the average of the individual calendar year returns over a date range (perhaps with the YTD figure) gives a helpful alternative that shows the effect of rogue results. You can take this view further with other analyses, but relying only on the term averages (the basis of the Morningstar Star calculations) can be misleading.

As for category chasing, in some cases this can work as it allows for macro-economic effects – even a bad fund in a good sector will beat a good fund in a poor sector. It’s like the “location effect” for housing. But this might need advisors to work with multiple companies since different fund manager companies are consistently better in some sectors than in others.

Posted by FifthDecade | Report as abusive