ETFs and Tracking Error

There was an interesting article in today’s WSJ about a recent report issued by Morgan Stanley on how the tracking error between exchange-traded funds and their prospective indexes increased last year:

Last year, 54 ETFs showed tracking errors of more than three percentage points, up from just four funds the prior year. And a handful of the 54 missed by more than 10 percentage points.

Nearly all exchange-traded funds, which are baskets of securities that trade all day like stocks, are designed to track indexes. So investors expect returns to closely mimic those of market gauges like Standard & Poor’s 500-stock index or the Barclays Capital (formerly Lehman) U.S. Aggregate Bond Index.

The article also mentions that the average tracking error for ETFs was 1.25% in 2009—over twice the .52% average in 2008.

For those of you who don’t know, tracking error is the difference in returns (both positive and negative) between an investment vehicle and the index it tracks. Sometimes the investment outperforms the index but it’s much more common for the investment to underperform due to fund expenses. The difference in performance can also come from how the ETF is set up. According to the article, if the ETF buys every stock in the index, then its trading costs can be higher. If it buys fewer stocks in order to just represent the index, then it can become more vulnerable to tracking error.

The article doesn’t mention how many ETFs were studied and I can’t find a copy of the report. If anyone finds a link to the report, please send it my way. I’d like to check it out.

4 thoughts on “ETFs and Tracking Error”

  1. You basically hit the key reason ETFs can track their indices. Many of the ETFs are relatively new so they have low asset bases. This means the ETF provider will use replication methods in hopes of tracking the underlying index since they don’t have enough assets to buy all the postions. As the asset base increases, then the ETF provider will add more positions.

    The best example of this is the Vanguard and iShares holdings. Many of their ETFs track the same Morgan Stanley indexes. In most of these ETFs, the Vanguard ETFs hold substantially more positions – sometimes double or triple the number.

    Vanguard created ETFs, but the Vanguard ETFs are actually a different share class of Vanguard’s established index mutual funds. The Vanguard funds have accumulated substantial assets over the years whereas iShares have only been in the business for a decade or so.

    So investors need to examine the tracking error as they may not be getting what they expect.

  2. ETF’s which track commodities are notorious for this. Other than silver and gold (which are usually pruchased and stored), other commodities are usually traded via futures and swaps. Where the error comes into play is from the effects of contango and backwardation. You can read the definition on wikipedia

    Sometimes the error can huge, as in the case of UNG (the natural gas etf) in which the spot price of natural gas increased many times that of the ETF itself. It is important to understand how the etf tracks the underlying financial instrument. Otherwise you could be in for a big surprise!

  3. JLP,

    Thanks for the article. Those are some large variances. A few aren’t too surprising as they are new ETFs and haven’t built the asset base as I discussed in my previous comment. However, the Vanguard Telecom was shocking as Vanguard usually prides itself on closely tracking the index. Of course, investors are probably happy they deviated since it topped it by 17%. The Powershares Emerging was another surprise as that is a key product for them. Their other ETFs are newer and a smaller slice of the market.

    Overall, I tend to use Vanguard ETFs because their tracking error is the tightest due to the large holdings. Also, the Vanguard expense ratios are the best of breed (lowest cost).

Comments are closed.