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.