In Monday’s post, I reviewed the major factors that contribute to an index fund’s tracking error. Here are some other things to consider when you’re comparing your fund’s performance to that of its benchmark. These can cause tracking errors to seem unusually large or small, but they need to be understood in context.
Changes to the index. A number of ETFs changed their benchmark index during 2012, including some core equity funds from BMO and Vanguard. When there is an index change in the middle of the year, measuring tracking error becomes difficult and the numbers can be misleading. Until late September, the BMO S&P 500 Hedged to CAD (ZUE) held just 100 large-cap stocks selected using a different methodology. ZUE ended up lagging the S&P 500 by less than its management fee, which is normally an excellent result, but in this case it was a fluke.
A small number of ETFs in Canada are not tied to any third-party benchmark. The BMO Canadian Dividend (ZDV), for example, includes 30 stocks selected using an in-house methodology. In cases like this, there’s not much you can do other than compare the fund’s performance to other Canadian dividend ETFs, or perhaps to a broad market index such as the S&P/TSX Composite. Even then, this will only be meaningful over multi-year periods.
Differences between market price and NAV. Last month I explained how an ETF’s market price can differ from its net asset value (NAV). In most cases, tracking error is measured by comparing the index return to the fund’s NAV. But things get complicated when a Canadian ETF holds an underlying US-listed ETF, a structure that’s common at iShares and Vanguard. Now you’ve got two different ETFs with two different NAVs and two different market prices.
Most of the large tracking error in the Vanguard MSCI U.S. Broad Market (VUS) was likely the result of currency hedging, but its annual report also cites “differences between the market price and net asset value of the underlying US domiciled Vanguard funds in which the ETF invests.” Don’t be concerned about this sort of tracking error: it’s a short-term anomaly that would disappear if you used different start and end dates.
Fair value pricing. One of the reasons an international equity ETF’s market price and NAV can diverge is time zone differences, as I explained in a recent post. This same phenomenon can also cause the ETF’s tracking error to appear unusually large, something I first wrote about back in 2010. Vanguard uses a technique called fair value pricing to correct for some of these distortions, and this idea is described in the most recent annual report for its emerging markets ETF (VEE):
The ETF’s management fees subtracted 0.54 percentage points from the fund’s performance. Other factors, such as security selection and fair-value pricing, added back 0.46 percentage points. Fair-value pricing is a policy intended to address pricing discrepancies that may arise because of time-zone differences among global stock markets. This policy ensures that the ETF’s net asset value doesn’t include “stale” prices from markets that close before the U.S. stock market.
The key point here is that if your ETF happened to show a large tracking error recently, that’s not necessarily a reason to abandon it. It’s critical to check the fund’s Management Report of Fund Performance to determine the explanation. As with so many investing decisions, tracking error should be considered with a focus on the long-term.
For the latter two reasons for ignoring tracking error (market price/NAV and fair value pricing), it appears that they should average out over time. So, there might be tracking error in any one year, but if they are in the same direction every year, there is a problem.