TL;DR: Tracking error is the standard deviation of the differences between a fund returns and its benchmark returns. Illustrative example only: quarterly differences of 0.5, negative 0.3, 0.4, and negative 0.2 percent have a mean of 0.1 percent and a sample standard deviation of about 0.41 percent per quarter. Annualised by multiplying by 2, the square root of 4 quarters, tracking error is about 0.82 percent a year. It measures consistency of the gap, not the size of the gap itself.
Key Takeaways
- The worked example in this draft is fully illustrative. Its arithmetic is recorded in Sources and method below and was recomputed for this draft.
- Definitions here follow standard filing and fund reporting usage. Where providers use a different definition, the draft says so and tells you to check which one a source is using.
- This is educational information, not financial advice, and it is not a recommendation to buy, sell, or hold any security or fund.
Work four quarters of differences
Tracking error starts from differences, not returns. All figures are illustrative. Suppose an index fund and its benchmark produce quarterly return differences, fund minus benchmark, of 0.5, negative 0.3, 0.4, and negative 0.2 percent. The mean difference is 0.1 percent, since 0.5 minus 0.3 plus 0.4 minus 0.2 equals 0.4, and 0.4 divided by 4 equals 0.1. That mean gap has its own name, tracking difference, and, as a separate annual illustration, a fund returning 10.6 percent against a benchmark returning 10.0 percent over a full year shows a tracking difference of 0.6 points, since 10.6 minus 10.0 equals 0.6.
Tracking error asks a different question: how wildly did the quarterly gaps swing around their mean. Deviations from the mean are 0.4, negative 0.4, 0.3, and negative 0.3. Squared, they are 0.16, 0.16, 0.09, and 0.09, summing to 0.50. Sample variance divides by 3: 0.50 divided by 3 equals about 0.1667. The square root is about 0.41 percent per quarter. Annualising multiplies by the square root of the number of periods, 2 for quarters: 0.41 multiplied by 2 equals about 0.82 percent. That square root scaling assumes the quarterly gaps are roughly uncorrelated with a stable variance; serially correlated gaps annualise differently. Two funds can share the same yearly tracking difference while one glides and the other lurches, and tracking error is the figure that tells them apart. The fund structures being measured are in what is an index fund and what is an ETF.
Where tracking error comes from
For index funds the sources are mundane and mostly unavoidable. Fees are the steady one: the drag in expense ratio explained pushes the fund below its cost free benchmark a little every period. Cash held for redemptions earns less than the index in rising markets. Sampling matters for wide indexes, where a representative subset drifts around the full index path. Securities lending income can offset some drag, one reason tracking difference occasionally beats what fees alone predict.
For active funds, tracking error is not a defect. It is the statistical shadow of every active decision, and a high figure certifies that the fund genuinely differs from its benchmark. The same number is a quality complaint for an index fund and a neutral description for an active one. Judging either needs the companion habit in track portfolio performance vs benchmark: same periods, same benchmark, costs included.
How to use the two tracking numbers together
Read tracking difference for cost and tracking error for consistency. For an index fund, a small, stable difference close to the fee level, with low tracking error, describes a fund doing its one job tightly. A widening difference or a rising error asks what changed: fees, sampling, fund size, or index methodology. For an active fund, pair tracking error with the return gap itself: high error with no compensating return over a long span describes expensive wandering, while the risk adjusted lens of the Sharpe ratio explained asks whether the wandering was ever paid for.
Neither figure predicts the future. Both describe how a fund behaved against its stated yardstick. For the one job index investors hire a fund to do, they are the closest thing to a report card the industry publishes.
This article is educational information, not financial advice. Prices, company figures, and fund figures in the worked example are illustrative and are not figures for any real security, company, or fund.
Sources and method
- Target keyword: "tracking error explained" from .saeo/keywords.yaml (queue entry added 2026-11-01). No search volume or difficulty figure is claimed; both are recorded as unknown in the queue entry.
- All figures are labelled illustrative examples. Arithmetic verified for this draft on 2026-11-01 by recomputation in Python: mean of 0.5, negative 0.3, 0.4, negative 0.2 equals 0.1 percent; squared deviations sum to 0.50; sample variance 0.50 divided by 3 equals about 0.1667; square root is about 0.41 percent quarterly; annualised, about 0.41 multiplied by 2 equals about 0.82 percent (all rounded); tracking difference check 10.6 minus 10.0 equals 0.6 points.
- Internal links checked for this publish batch on 2026-11-01 against the live Fullport sitemap plus this batch: sharpe-ratio-explained, track-portfolio-performance-vs-benchmark, what-is-an-etf, what-is-an-index-fund.
