Tracking Error

The divergence between the price behavior of an investment portfolio and the price behavior of its benchmark index.

Tracking error is a statistical measure that quantifies how closely an investment portfolio follows the performance of its benchmark index. While tracking difference simply shows the net return gap over a specific period, tracking error measures the volatility of those return differences over time. It is calculated as the annualized standard deviation of the daily or monthly excess returns of the portfolio relative to its benchmark. For passive investors, tracking error serves as a primary scorecard for fund manager efficiency. A low tracking error indicates that an index fund or exchange-traded fund (ETF) is successfully replicating its target index. Conversely, a high tracking error reveals that the fund is deviating significantly from its stated target, which can expose investors to unexpected risks or performance shortfalls. In practice, tracking error is driven by several real-world factors. These include management fees, transaction costs from buying and selling assets, cash drag (holding cash to meet redemptions), and index reconstitution delays. Additionally, funds that use index sampling rather than full replication often experience higher tracking error because they hold only a representative subset of the index's components.

An index fund and its benchmark index have the following monthly excess returns (Fund Return minus Benchmark Return) over four months: Month 1 is +0.1%, Month 2 is -0.2%, Month 3 is +0.1%, and Month 4 is 0.0%. The average monthly excess return is 0.0%. The differences from this average are 0.1%, -0.2%, 0.1%, and 0.0%. Squaring these differences gives 0.01%, 0.04%, 0.01%, and 0.00%. The sum of these squared differences is 0.06%. Dividing by 3 (sample size minus 1) gives a variance of 0.02%. Taking the square root of the variance yields a monthly tracking error of approximately 0.141%. To annualize this, we multiply by the square root of 12, resulting in an annualized tracking error of 0.141% 3.464 = 0.49%.