Tracking error

Tracking error measures how much an ETF's or index fund's returns deviate from its benchmark's returns over time, expressed as the standard deviation of the difference between the fund's return and the benchmark's return across a series of periods. A low tracking error means the fund's returns move almost in lockstep with the benchmark period after period, while a high tracking error means the gap between the two bounces around more from one period to the next.

Tracking error is different from simply looking at whether a fund beat or lagged its benchmark in total over a year, that comparison is closer to what tracking difference measures. Tracking error instead captures consistency, a fund could have a very small average shortfall against its benchmark over a year while still showing a fairly high tracking error, if that shortfall came from a bumpy mix of periods well ahead of the benchmark and periods well behind it, rather than a smooth, steady gap.

For a fund explicitly designed to replicate an index as closely as possible, low tracking error is generally seen as a sign of good fund management and an efficient replication process, while unusually high tracking error can point to problems such as poor sampling of the index's holdings, high trading costs from frequent rebalancing, or difficulty accessing certain securities in the benchmark, for example in a fund tracking an index of hard to trade international or emerging market stocks.