How closely does a fund actually follow its index?

A fund promises to track an index. Two numbers measure how well it kept the promise, and they are routinely confused with each other.

Difference and error are different questions

Tracking difference is the fund's return minus the index's return over a period. It is a signed number and it is almost always negative, because the fund pays fees and trading costs and the index pays nothing. A fund charging 0.10% a year with no other leakage shows a tracking difference near −0.10%.

Tracking error is how much that difference wobbles from period to period — its standard deviation. A fund that lags its index by exactly 0.10% every single year has a large-ish tracking difference and a tracking error of zero: predictable, and what an index fund should look like. A fund that beats the index one year and lags it by 2% the next has a small average difference and a large error, which means something inside it is not tracking.

For a holder, the difference is the cost and the error is the surprise. You can live with a known cost. A surprise on an index fund means the fund is doing something the label does not say.

Where the difference comes from

In roughly the order they matter for a large equity fund: the expense ratio; the cost of trading the index's own changes; cash held between dividend receipt and reinvestment; and, working the other way, securities lending — a fund that lends its shares to short sellers collects a fee, and a well-run tracker can claw back a meaningful part of its expense ratio that way. The synthetic-and-lending lesson comes back to what that lending costs in risk.

The trap in the comparison

Here is a chart of an S&P 500 fund and of the index it tracks. Both open at their full history, which is longer than ten years; the numbers below come from the measure tool set to 2 September 2016 and 3 September 2026:

Measured from 2 September 2016 to 3 September 2026, the fund's adjusted price rose about 316% and the index rose about 255%. Read naively, the fund beat its index by a mile, which is impossible for a fund charging a fee.

The catch is what the two lines measure. The index is a price index: it counts the companies' share prices and ignores every dividend they paid. The fund's adjusted series includes those dividends, reinvested. The 60-point gap is ten years of S&P 500 dividends, not tracking. Compare the fund against the total return version of the index and the gap collapses to the expense ratio and a little friction — which is the true tracking difference, and it is small.

So the first question in any tracking comparison is: which version of the index? A price index against a total-return fund flatters the fund; a total-return index against a price series of the fund damns it. Neither is tracking.

In the data

The fund side is easy to find: the fund's adjusted price folds its distributions back in, and its data record states its own returns:

The index side is the hard one. The index series on the chart above is the price index, which leaves the dividends out, and the total-return version that a fair tracking comparison needs is published by the index provider rather than shown on most charts.

Try it now

1. Measure the same ten-year window on both charts above — 2 September 2016 to 3 September 2026, not the whole history the chart opens with — and write the two percentages down. Then divide one plus the fund's return by one plus the index's and take the tenth root: ((1 + 3.157) / (1 + 2.554)) ^ (1/10) is about 1.016, so roughly 1.6% a year of reinvested dividends. Compare it with the fund's distribution yield (0.98% on 29 September 2026); if yours is higher, the decade's average yield was higher than today's, which is what ten years of rising prices does to a yield.

2. Read the ten-year return in the fund's performance table above. Compare it with what you measured; the provider's figure is annualised, yours is cumulative, and converting one into the other is the risk-and-return course's CAGR arithmetic.

3. From memory: what is the difference between tracking difference and tracking error, and which one would worry you on an index fund?