The information ratio measures expected return above a benchmark per unit of volatility in the portfolio’s return difference from that benchmark. Unlike the Sharpe ratio, it evaluates performance relative to a chosen comparison portfolio rather than the risk-free rate. section 7.3 section 7.2.1
Compute and interpret the ratio by:
- Choosing a benchmark, such as the S&P 500, and calculating portfolio returns relative to it. section 7.2.3
- Dividing the average active return by the standard deviation of active returns. section 7.2.3
- Annualizing a monthly ratio by multiplying it by the square root of 12. section 7.2.3
Evaluating an ESG fund
For the iShares MSCI USA ESG ETF (SUSA), calculate monthly returns, use the S&P 500 as the benchmark, and compute active returns as y - X[0]. The implementation IR = (y-X[0]).mean()/(y-X[0]).std()*sqrt(12) produces an annualized information ratio of -0.176, indicating negative benchmark-relative performance per unit of active volatility over the measured period. section 7.2.3
