Comments on: ETFs start to underperform A slice of lime in the soda Sun, 26 Oct 2014 19:05:02 +0000 hourly 1 By: ReutersRat Mon, 22 Feb 2010 22:33:16 +0000 Felix, why isn’t the problem true with normal index funds?

By: DanHess Mon, 22 Feb 2010 18:51:50 +0000 WCW, Ginsbu, good points, thank you (scales falling). Index sampling is not insidious because while there may be deviations, it is not biased against the investor. The investor might just as easily beat the index as fall short.

Many mutual funds underperform because of front-running, a problem exacerbated by the fact that their managers like to trade all day long. ETFs, with low turnover, are the solution to the front-running problem! Few trades, little front-running!

Long Live ETFs!!

By: wcw Mon, 22 Feb 2010 07:31:26 +0000 Dan, read the first comment. EEM samples. Felix has a point (SPY has no business failing to track perfectly, full stop), but this column wasn’t his best work.

Expatsb, what? You can front-run anything, but especially smaller, less-liquid names.

Felix, please revise and resubmit. ETFs are a funny space, and it’s worth asking why SPY didn’t track and why EEM doesn’t replicate. But this column isn’t the one you wanted to write.

By: DanHess Sun, 21 Feb 2010 22:40:46 +0000 Felix, consider this:

EEM has an annual turnover of just 5%.

If there is no turnover, there is nothing to front-run. How could poor pricing on %5 of the portfolio lead to underperformance of 6.7%. That would suggest that the fund is overpaying and underselling by more than 50% on buys and sells respectively. Likely?

There must be some other cause for this difference, and it must balance out for other years, no?

By: mjturner Sun, 21 Feb 2010 19:48:33 +0000 Your piece doesn’t really say avoid commodity ETFs, it just says that the price of commodities might be hit if commodity ETF investors decide to sell. Which is true, but rather different than saying the ETF itself is defective (they might or might not be, hard to make the case in gold where they are backed by a physical asset but perhaps more possible in others)

By: expatsp Sun, 21 Feb 2010 19:05:15 +0000 I would presume these errors don’t apply to indices that attempt to cover the /entire/ stock or bond market, as opposed to just large caps like the SPY. If everything’s in the index already, you can’t front-run it.

Given that small investors of modest means probably shouldn’t be making sector or country bets, these total ones are or should be the most important ETFs.

If I’m not mistaken, Vanguard for several years has been steering investors away from the ETF/index funds that track the S&P 500 and toward its total market index for just this reason.

By: ginsbu Sun, 21 Feb 2010 17:42:25 +0000 I don’t think it’s so clear that the evidence points to size as being a substantial contributor to ETF benchmark underperformance. At the very least, the front-running claim should apply to index mutual funds as well. Why should ETFs be different in this respect? Furthermore, if front-running index changes is a cause of underperformance, why should it matter whether you’re invested in, e.g., a large S&P500 fund or a small one? The problem would be attempting to track a popular index.

Citing EEM as evidence seems particularly problematic. As has been discussed widely elsewhere, EEM’s index sampling opens it to tracking error. VWO more fully replicates the same index and performed much better. According to your front-running hypothesis, shouldn’t VWO have suffered similarly?

Finally, regarding securities lending, ETF providers do not all remit the same amount of lending revenue to their funds. Vanguard remits substantially all, SSGA (SPDRs) somewhat less, and Barclay’s (iShares) only half. Where tracking errors of tenths or hundredths of a percentage points are at issue, these differences can matter.