Tracking Error
How far a portfolio's returns deviate from its benchmark, measured as a standard deviation. It quantifies how different a portfolio is, not how good.
MadStockAlerts Research · Updated August 29, 2026
What to take away
- Tracking error is the standard deviation of the return difference against a benchmark.
- Low tracking error means similar behaviour, not similar quality.
- An index fund's tracking difference and tracking error are different measures.
- Active share measures how different the holdings are; tracking error measures how differently they behave.
- High tracking error requires a long record before it can be judged.
MAD Academy Training Video · 0:45
How Far a Fund Wanders
Tracking error measures how much a fund's returns differ from its benchmark, and low is only good if tracking was the point.
This lesson is part of a Stock Alerts + Tools plan.
The measure
tracking error = standard deviation of (portfolio return - benchmark return)
- measured over a series of periods, usually monthly and annualised
- a result of 2 percent means the difference typically falls within about 2 points either side of the average difference
It says nothing about direction. A portfolio consistently ahead of its benchmark by a similar margin has low tracking error, and so does one consistently behind. What it measures is how variable the difference is.
| Tracking error | Typical description |
|---|---|
| Under 0.5% | Index tracking |
| 1 to 3% | Enhanced or constrained active |
| 4 to 8% | Active, with meaningful deviation |
| Above 8% | Concentrated or unconstrained |
Tracking difference is a different thing
For an index fund, two measures are reported and are frequently confused. Tracking difference is the actual return gap over a period; tracking error is the variability of that gap.
| Tracking difference | Tracking error | |
|---|---|---|
| What it measures | How far behind or ahead, in total | How variable the gap was |
| Usual cause | Fees, transaction costs, cash drag, withholding tax | Sampling, rebalancing timing, currency effects |
| Ideal for a tracker | Small and negative, roughly the fee | As close to zero as possible |
For a fund whose job is to track, tracking difference approximately equal to the expense ratio and a very low tracking error is the expected result. A tracker with a large tracking error is doing something other than tracking.
Why deviation needs time to judge
A portfolio with high tracking error will spend substantial periods ahead of and behind its benchmark simply because it is different. Distinguishing skill from deviation requires enough periods for the average to emerge from the variability.
The higher the tracking error, the longer the record needed. This is the same sample-size problem that appears throughout this library: variability makes a short record uninformative, and the more variable the record the longer it has to be.
Information ratio
Tracking error alone says how different a portfolio is. Pairing it with the average return difference says whether the deviation was worth taking.
information ratio = average excess return / tracking error
- excess return is measured against the same benchmark used for the tracking error
- it is the Sharpe ratio's structure applied to a benchmark rather than to cash
A portfolio beating its benchmark by one percent with a one percent tracking error has an information ratio of 1.0, and one beating it by the same margin with an eight percent tracking error has a ratio of 0.125. The second took far more deviation to achieve the same result.
The measure inherits every problem of the Sharpe ratio, including sensitivity to the period and to the distribution's shape. It also requires the same long records, and for the same reason: the higher the tracking error, the longer the record has to be.