William J. Bernstein
Itís the Execution, Stupid
Taking oneís lumps comes with the authorial territory and, as these things go, Iíve been lucky. Ninety-five percent of my feedback is positive; in fact, this psychic pay far exceeds my royalties and advances. Further, most of the discontents are pretty obviously the owners of various gored oxen, even if the precise breed is never identified.
But one particular criticism ranklesóthat my books are "out of date." Why? Because I "ignore" ETFs. In the first place, I donít: The Intelligent Asset Allocator was published in 2000, and although I briefly discussed the blossoming field, at the time they were just too new to recommend. In The Four Pillars of Investing, which came out a scant eighteen months later, I actually did recommend their purchase to those who were so inclined, and listed them along with the appropriate open-end vehicles. I did state that ETFs were not appropriate for value averaging/dollar cost averaging, for obvious reasons.
Apparently, that was not enough for those who think that open-end funds are old fashioned, if not downright foolhardy, especially in an environment rife with late trading and market timing.
How do I really feel about ETFs? I donít buy them. Not for myself, my family and, in particular, not for the clients of our advisory firm. The reason? Because, in most cases, you can do better. To show you why, Iíve put together a table from the Morningstar database (note that ER is expense ratio and TE, tracking error):
ETF TE (Gross of Fees)
Vang. TE (Gross of Fees)
The first column lists the asset class being tracked, and the next three columns are the returns of the appropriate index, ETF, and Vanguard fund, respectively. All of the ETFs are iShares, except for the SPDR S&P 500 offering, which provides a much longer baseline. The last two columns are the tracking errors versus the index for the ETFs and Vanguard funds respectively, gross of fees (i.e. taking expenses into consideration). Performance net of fees can be determined simply by looking at the third and fourth columns. I cut the analysis off at the end of April 2003 for the S&P 500 value and growth, the S&P 600 value and growth, and the Midcap (S&P 400) category because of Vanguardís switch away from the S&P benchmark system after that.
In all seven cases where a direct head-to-head comparison can be made, the Vanguard funds outperform the iShares. The results are highly statistically significant, with a t-stat of 2.78 for net returns and 3.44 for gross returns (p values less than 0.016 and 0.007, respectively). Pretty impressive for just seven data points. In some cases, it isnít even close; the Vanguard Small-Cap Value Index Fund beats the relevant ETF by about 80 basis points (bp) both before and after expenses.
The conclusion here should be obvious even to the most rabid early-adopter (if it hasnít already hit everyone else over the head in light of recent events in the fund industry): corporate culture counts. Itís not that thereís anything wrong with Barclays; their tracking errors are pretty respectable. Itís just that theyíre not Gus Sauter.
Another company, like Vanguard, that executes superbly is Dimensional Fund Advisors. Their flagship U.S. Microcap Fund has outperformed its index, the obscure CRSP 9-10 decile, by almost 1% per year since inception in 1992, despite a 0.56% expense ratio. (All of that margin came in the first half of the periodóin the past decade, theyíve managed to keep up with the index on a net basis, no small accomplishment in view of the fact that the fund now owns upwards of 10% of the market cap of each of its names.)
The "problem" with DFAís other funds is they donít really take their benchmarks seriously, providing multiple versions for some, none for others, and thumbing their noses at the benchmarks in any case, on the theory that itís best to aim for low turnover and negative transactional expenses, rather than slavishly follow an index, which can be expensive. I agree strongly with them and, in fact, wish that Vanguard would have the moxie to do the same. If DFA has the credibility to put performance above index tracking, certainly so does Vanguard.
But I digress. The arguments in favor of ETFs (that they are theoretically more tax-efficient, that they are less susceptible to timing and late-trading shenanigans, that in a few cases their expense ratios are a tad lower than the corresponding open-end funds) pale in comparison to the performance difference. What good is it that the SPDR expense ratio is 7 bp less than Vanguardís 500 Index Fund when the latter has a 16 bp execution advantage? Just how likely is it that anyone is timing or late trading Vanguard funds? I would not be surprised if somewhere someone had figured out how to get a bit of late or rapid trading past Vanguard, but Iíd be shocked if there were any monkey business internal to the company. Regardless, Vanguard still manages to outperform the corresponding ETF in every case.
And thatís before we get to the cows in the ETF living room: commissions, spreads, and the truly awful performance of the single-country offerings. Call me old fashioned: Iíll go with outperformance born of transactional skill every time, even if my fund choices are as out of date as my khakis, eyeglasses, and minivan.
Copyright © 2004, William J. Bernstein. All rights reserved.
The right to download, store and/or output any material
on this Web site is granted for viewing use only. Material may not
be reproduced in any form without the express written permission of William
J. Bernstein. Reproduction or editing by any means, mechanical or electronic,
in whole or in part, without the express written permission of William J.
Bernstein is strictly prohibited. Please read the disclaimer.
The right to download, store and/or output any material on this Web site is granted for viewing use only. Material may not be reproduced in any form without the express written permission of William J. Bernstein. Reproduction or editing by any means, mechanical or electronic, in whole or in part, without the express written permission of William J. Bernstein is strictly prohibited. Please read the disclaimer.