Definition
The portfolio information ratio measures average return relative to a chosen benchmark per unit of tracking error. It addresses how consistently a portfolio adds active return. Its value depends on the benchmark, dates, costs and measurement frequency.
Information ratio formula
Let active[t] = portfolio return[t] - benchmark return[t]. For regularly spaced observations:
annualized IR = mean(active) / std(active) × sqrt(periods per year)
Tracking error is the standard deviation of active returns, usually annualized. A simple square-root annualization requires suitable assumptions about the return process. A portfolio identical to its benchmark has zero active return and zero tracking error, so this ratio is undefined rather than evidence of a perfect strategy.
Subtracting two CAGRs and dividing by portfolio volatility is not this calculation. Use the aligned periodic active-return series. State whether fees are included and ensure both series treat dividends consistently.
Information ratio versus Sharpe ratio
| Measure | Reference return | Denominator |
|---|---|---|
| Information ratio | Chosen investable benchmark | Variability of active returns |
| Conventional Sharpe ratio | Risk-free return | Variability of excess returns |
A positive information ratio depends on the chosen benchmark. A strategy could add value relative to one market and trail another. The Sharpe guide explains why measurement conventions need to be aligned before ratios are compared.
Two meanings of IR in factor research
The U.S. momentum study reports a one-month mean rank information coefficient of 0.012 and an IC information ratio of about 0.06. That IC ratio summarizes the stability of a factor's cross-sectional predictive correlation through time. It is not the active-return information ratio of the leveraged ETF portfolio.
The same study's 29.9% CAGR and the benchmark's 20.7% CAGR do not establish its portfolio IR. That would require the aligned strategy and benchmark periodic returns. Keep “IC information ratio” and “portfolio information ratio” explicit when reading a report or asking an assistant to calculate a metric.
Common pitfalls
A very small tracking error can make an estimate unstable. Changing the benchmark after seeing results creates a selection problem. A price-only benchmark against a dividend-adjusted strategy overstates active return. And a positive average active return can conceal long periods of underperformance that matter to an investor.
Use the backtesting guide to define a benchmark before testing. The factor investing overview explains the difference between a factor research statistic and an investable portfolio result.
Frequently asked questions
Is a high information ratio always preferable?
It is one measure of active-return consistency under a specific benchmark and sample. Review exposure, costs, drawdown and statistical uncertainty alongside it.
Is information ratio the same as alpha?
No. Alpha is an intercept from a specified return model. IR uses mean active return and tracking error. They can be related in a particular model, but their definitions and inputs differ.
Related terms
See beta, CAGR, turnover and the quant glossary.
For education and research only. This is not investment advice. Historical performance does not guarantee future results.
Last updated: 2026-10. Author: FinLab Quant Research.