- Paired observations
- 6
- Mean active return per period
- 0.0667 %
- Sample tracking error per period
- 0.1862 %
- Annualized tracking error (zero serial covariance)
- 0.645 %
Each portfolio return minus its matching benchmark return. The horizontal axis follows the order supplied.
View the chart values
| Series | Observation | Active return (%) |
|---|---|---|
| Active return | 1 | 0.2 |
| Active return | 2 | 0.1 |
| Active return | 3 | 0.1 |
| Active return | 4 | -0.2 |
| Active return | 5 | 0.3 |
| Active return | 6 | -0.1 |
On this page
Pair the same dates and return conventions
Each portfolio observation is compared with the benchmark observation in the same position. Enter percentages, so 1.2 means 1.2%, and use equally spaced dates. The demonstration inputs are fictional monthly returns. Both series need matching currency, fee and total-return conventions for the comparison to answer a meaningful question.
Understand the sample statistic
The tool subtracts benchmark return from portfolio return for each observation, calculates the mean difference, and measures deviations from that mean using n − 1 in the variance denominator. It requires at least two pairs. A constant active return has zero sample tracking error even when the portfolio consistently differs from the benchmark.
Choose periods per year and interpret the estimate
Twelve represents monthly observations; four represents quarterly observations. The tool multiplies periodic tracking error by the square root of this number. That scaling assumes equal periodic variances and zero covariance between different periods. Correlated active returns can make the annual figure differ from this estimate.
Keep the comparison within its scope
The result describes the supplied series and convention. It does not calculate a compounded annual active return, prove future tracking performance or isolate the cause of a return difference. Mismatched dates or benchmarks can create a misleading statistic even if the arithmetic is correct.