Tracking Error

Why tracking error matters

Tracking error converts the abstract idea of active management into a measurable risk budget. It helps investors distinguish index-like portfolios from genuinely differentiated mandates, compare intended and realized activity, and size complementary managers. The information ratio uses it as the denominator. A portfolio can have low absolute volatility yet high tracking error if it differs materially from its benchmark, or high volatility and low tracking error if both move together.

How it is applied

Calculate portfolio return minus benchmark return for each period, then take the sample standard deviation of that series and annualize consistently. Ex ante tracking error comes from current holdings, factor exposures, and a covariance model; ex post tracking error uses realized returns. Managers monitor contributions by security and factor, forecast changes before trades, and compare realized risk with mandate ranges. Benchmark, currency, pricing time, and derivative exposure must align.

Formula

Portfolio example

A global equity fund has annualized active return of 1.5% and tracking error of 3%, producing an information ratio of 0.50. A second fund has the same return and 8% tracking error. The second took more benchmark-relative risk. If both hold illiquid securities priced at different times from the index, observed tracking error may reflect valuation timing rather than deliberate active decisions.

How to interpret it

Higher tracking error means active returns varied more widely, not that the manager performed worse. The appropriate level depends on mandate, fee, concentration, and investor portfolio. Realized tracking error above target may signal changing markets, unintended exposures, or weak controls. A very low value for a high-fee active fund can suggest closet indexing, but active share and holdings analysis are needed because the two measures capture different dimensions.

Limitations and common misconceptions

Historical tracking error is backward-looking and sample-dependent. Forecasts rely on unstable correlations and factor models. The measure treats favorable and unfavorable deviations equally and misses tail shape, liquidity, and absolute loss. Benchmark mismatch makes it meaningless, while stale pricing can suppress or distort it. Investors should examine active return, attribution, active share, stress tests, and economic exposures alongside the headline number.