Tracking Error: How Closely an ETF Follows Its Index

The cheapest fund on the shelf can still trail its index by more than its fee. Fees are one cause of the gap. Cash, sampling, trading and futures rolls are the others, and a fact sheet splits the damage into figures that are easy to confuse.

Tracking difference is the gap, tracking error is the wobble

Tracking difference is the plain shortfall over a period, the fund’s return minus the index’s return, while tracking error measures how much that shortfall moves around from one month or quarter to the next, which is a different question about the same pair of numbers. A fund can have a steady, predictable difference and almost no tracking error. Another can have the same total difference and a large tracking error, because it beat the index one quarter and fell well behind the next.

That last tenth of a percent is where tracking error lives. To see the wobble, split the year into quarters.

Both funds cost you the same 0.20 points over the year. Fund A tells you in advance what it will cost. Fund B doesn’t.

What causes the gap

• Fees, taken out of the fund’s assets daily.
• Cash, which the index never holds.
• Sampling.
• Trading costs when the index changes its members or weights.
• Storage, insurance or futures rolls in commodity funds.

Sampling needs a line of its own. A fund following an index with thousands of bonds or small stocks may hold a representative subset, because buying every line would cost more than the error it avoids, and whenever the securities it skipped move differently from the ones it holds, the fund drifts away from the index for reasons that have nothing to do with fees. Cash drag works the same way on a smaller scale: dividends sit uninvested for a few days while the index assumes they were reinvested at once.

Commodity funds add their own layer. A fund that holds physical metal pays for vaults. A fund that holds futures has to sell expiring contracts and buy later ones, and the price difference between the two can cost or earn money every month. The gold ETF list separates metal funds from futures funds, and gold ETFs or gold miners covers what each one actually tracks.

Where to find it on the fund fact sheet

Fact sheets usually show fund returns next to index returns for the same periods, such as one month, a quarter, a year, three years and since inception, and subtracting one from the other gives you the tracking difference for each window. Some fact sheets print a tracking error figure directly, labeled by the period it covers and often annualized. The annual report has the audited figures. Its expense ratio and portfolio breakdown explain where the gap came from.

That separate effect is the gap between the fund’s traded price and its net asset value, covered under ETF premium and discount.

What people get wrong

The usual error is shopping on expense ratio alone. A hypothetical fund with a 0.05% fee and a sloppy sampling method can trail its index by more in a bad year than a fund charging three times as much that holds every security, and since the fee is the only cost printed in large type, the more expensive fund can look like the worse deal when it is the better one. A low expense ratio with high tracking error can cost more than it looks.

People also read a tracking error of zero as proof the fund is free. Fund A in the example has zero tracking error and still loses 0.20 points a year. And some treat a year when the fund beat its index as a sign of skill; for an index fund it’s usually lending income or luck, and it is just as likely to reverse.

Related terms

Tracking difference is the gap itself. Expense ratio is the annual fee as a share of assets. Sampling, also called optimization, is holding a subset of the index. Active risk is the name portfolio managers use for tracking error when a fund is meant to differ from its benchmark. More ETF definitions sit under ETFs.