Information Ratio Calculator

The information ratio divides active return by tracking error. Active return is the portfolio return minus the benchmark return, and tracking error is the standard deviation of that difference over the same window. With a portfolio at 13%, a benchmark at 11% and tracking error of 2 percentage points, active return is 2 points and the ratio is 1.00: one point of excess return for each point of active risk taken.

The result card groups the number into one of four illustrative bands: 1.00 and above, 0.50 to under 1.00, 0.00 to under 0.50, or below 0.00. Those cut-points are this calculator's own grouping rather than a formally defined scale. A ratio below zero means the portfolio trailed its benchmark across the window measured. Where a figure sits in that range says nothing on its own about whether the same result repeats.

A tracking index fund does not land at zero. Fees, sampling and cash drag leave a small negative active return, and the tracking error underneath it is also small, so the ratio is a quotient of two small numbers. Its sign is reliably negative, but its size is governed by the ratio of drag to tracking error and is not bounded near zero: 0.05 points of drag against 0.02 points of tracking error produces −2.50. A tightly run full-replication tracker can therefore post a more negative figure than a loose one, because the smaller denominator magnifies the same drag. Where tracking error is genuinely zero the ratio has no value at all, because the division has no denominator, and the calculator returns a message in place of a number.

What moves the number most

The two return inputs are linear levers and mirror each other. At a tracking error of 2 points, a 1-point move in either return shifts the ratio by 0.50, upward for the portfolio and downward for the benchmark, and because both are linear, that step is the same size in both directions at any starting value. Tracking error is the one that behaves differently. Sitting in the denominator makes its effect hyperbolic and asymmetric: cutting it from 2 points to 1 lifts the ratio from 1.00 to 2.00, a step of a full point, while raising it from 2 to 3 drops the ratio to 0.67, a step of a third. Measured as discrete one-point steps, the tracking error lever is the odd one out at every starting value, defaults included. The three levers coincide only under the instantaneous-slope reading, and then only where the ratio is exactly 1.00.

The formula behind this

Information ratio = active return / tracking error. The denominator is tracking error rather than total volatility, and that is what separates this measure from the Sharpe ratio. Total volatility includes the market movement the benchmark already carries, while tracking error strips that out and leaves only the variability the manager introduced by deviating. The ratio therefore answers a narrower question than Sharpe does: how much return each unit of deviation from the benchmark produced.

What the ratio cannot separate

A high figure can come from manager skill or from a benchmark that was a poor match for the mandate, and the arithmetic cannot tell those apart. A portfolio measured against an index it was never built to track shows large active return and large tracking error, so the ratio that comes out of that pairing describes the mismatch as much as the management. An easy benchmark distorts the figure by a different route: it lifts active return, while a mandate that stays close to that benchmark keeps tracking error low, so both terms of the quotient push the ratio in the same direction.