International Tracking Error: Part 1

In a previous post, I pointed out that many international equity ETFs showed large tracking errors in 2009. For example, the iShares MSCI EAFE Index Fund (XIN), which tracks European and Japanese stocks, trailed its benchmark index by 5.34%. The comparable Vanguard Europe Pacific ETF (VEA) also lagged by 3.44%.

I contacted iShares, Vanguard and Claymore to ask how their international ETFs got so far off track last year. The answers turned out to be varied and complicated, so I’ll take a couple of posts to explain them.

The first point to understand is that the net asset value (NAV) of international funds can be temporarily distorted because of the time differences between North America, Asia and Europe. The Tokyo markets close at 1 a.m. Eastern Standard Time, while the London Stock Exchange closes at 11:30 a.m. EST. By the time the US and Canadian markets stop trading at 4 p.m., the Asian markets have been already been closed for 15 hours, and the European markets for four-and-a-half hours.

International indexes are calculated using the end-of-day prices of stocks in their native country, but a Canadian or American ETF tracking that index will continue to trade after those overseas markets are closed. Here’s why that can make a big difference: imagine that Nestlé’s share price is $20 when the Swiss Stock Exchange closes at 5:30 p.m. in Zürich. The index value for the day will be calculated using that $20 price. Two hours later, at 1:30 p.m. EST, there’s a news report that a dozen Americans have died from drinking tainted Nestea. The value of Nestlé shares plummets to $15—call it a Nestea plunge—and North American ETFs that hold the stock decline accordingly. The European index, meanwhile, won’t change until the Swiss exchange reopens the next day. So if you compare the value of the ETF to that of the index when the New York and Toronto markets close, there will be a significant tracking error.

This example is tongue-in-cheek and a bit extreme, but price movements like this occur regularly. To mitigate the effect, Vanguard uses a technique called “fair value pricing” to adjust the NAV of their international index funds and ETFs. (I highly recommend reading this article and watching this presentation about fair value pricing on Vanguard’s website.) This may sound like financial sleight-of-hand, but it actually protects the funds’ shareholders by preventing a kind of market timing called time zone arbitrage. iShares accomplishes the same goal by not allowing the creation or redemption of new units in their international ETFs when the underlying markets are closed. “This prevents anyone gaming the stale NAV prices,” they explained to me. While these measures are beneficial for long-term investors, they can cause short-term differences between the funds’ performance and the benchmark index.

In both late 2008 and late 2009, the markets happened to be unusually volatile, and ETFs that track international indexes were thrown for a loop by big one-day price changes in many stocks. The funds’ tracking errors looked large when measured from December 31, 2008, to December 31, 2009. But this was a random and temporary anomaly. Had you measured the funds’ 12-month tracking error from different start and end dates, the results would have been quite different, and could have been positive or negative. Over the long term, these anomalies should even out. Indeed, if you look at longer periods—say, three to five years—both Vanguard’s and iShares’ international ETFs have an excellent record of tracking their indexes closely.

In my next post, I’ll look at another reason why international index funds and ETFs can have large tracking errors: representative sampling.

Filed under: ETFs