Information ratio (IR) measures the consistency of excess returns above a benchmark, adjusted for the risk taken to achieve those returns. It’s the key metric for evaluating active management—showing whether deviating from an index is actually worth it. High information ratio means you’re generating alpha efficiently; low IR suggests you’re taking active risk without adequate reward.
- IR = Active Return / Tracking Error (volatility of difference from benchmark)
- Above 0.5 is good; above 1.0 is excellent
- Shows whether your active bets are paying off
How Information Ratio Works
Information ratio compares the return you add above the benchmark to the volatility of that difference (tracking error).
Information Ratio = (Rp - Rb) / Tracking Error
Where:
- Rp = Portfolio return
- Rb = Benchmark return
- Tracking Error = Standard deviation of (Rp - Rb) over time
Quick Reference
| Information Ratio | Interpretation | Skill Level |
|---|---|---|
| Below 0 | Underperforming benchmark | Negative skill |
| 0 to 0.3 | Marginal outperformance | Slight edge |
| 0.3 to 0.5 | Moderate outperformance | Decent active management |
| 0.5 to 0.75 | Good outperformance | Skilled manager |
| 0.75 to 1.0 | Very good | Top quartile |
| Above 1.0 | Excellent | Elite performance |
Example Calculation
Your Trading vs. S&P 500:
Monthly returns over 12 months:
| Month | Your Return | S&P 500 | Difference |
|---|---|---|---|
| Jan | 4.2% | 3.1% | +1.1% |
| Feb | -1.5% | -2.3% | +0.8% |
| Mar | 2.8% | 2.5% | +0.3% |
| … | … | … | … |
| Avg | 1.5%/mo | 1.0%/mo | +0.5%/mo |
Step 1: Calculate Active Return Active Return = 1.5% - 1.0% = 0.5% monthly = 6% annualized
Step 2: Calculate Tracking Error Tracking Error (Std Dev of differences) = 1.8% monthly = 6.2% annualized
Step 3: Calculate Information Ratio
IR = 6% / 6.2% = 0.97
Your information ratio is 0.97—very good. You generate almost 1% of alpha for every 1% of active risk.
Information ratio measures alpha divided by tracking error—the volatility of your difference from the benchmark. An IR above 0.5 is good, above 1.0 is excellent. It shows whether your active trading decisions add value relative to the risk of deviating from the index.
Information Ratio vs Sharpe Ratio
| Metric | Numerator | Denominator | Measures |
|---|---|---|---|
| Sharpe | Return - Risk-Free Rate | Total Volatility | Absolute risk-adjusted return |
| Information | Return - Benchmark | Tracking Error | Active management skill |
When to Use Each:
- Sharpe: Comparing any strategy to a risk-free alternative
- IR: Evaluating whether active trading beats passive investing
Improving Information Ratio
Increase Alpha (Numerator):
- Better stock selection
- Superior timing
- Exploit consistent edge
Decrease Tracking Error (Denominator):
- Focus on high-conviction bets only
- Avoid unnecessary portfolio churn
- Don’t take positions without clear edge
The goal is concentrated, high-quality active bets—not lots of random deviations from benchmark.
Information Ratio by Strategy Type
| Strategy | Typical IR | Tracking Error |
|---|---|---|
| Index-Enhanced | 0.2 - 0.4 | 1-3% |
| Active Core | 0.3 - 0.6 | 3-6% |
| Concentrated Stock Picking | 0.4 - 0.8 | 6-12% |
| Sector Rotation | 0.3 - 0.7 | 5-10% |
| Long/Short | 0.5 - 1.0 | 8-15% |
Higher tracking error isn’t bad if it comes with proportionally higher alpha.
Common Mistakes
-
Ignoring tracking error – 10% alpha sounds great, but if tracking error is 25%, your IR is only 0.4.
-
Wrong benchmark choice – Small-cap strategy compared to S&P 500 will have distorted IR. Use appropriate benchmark.
-
Too short a period – IR from 6 months is unreliable. Need at least 2-3 years for meaningful calculation.
-
Conflating IR with Sharpe – They measure different things. High Sharpe doesn’t guarantee high IR and vice versa.
How JournalPlus Tracks Information Ratio
JournalPlus calculates your information ratio against common benchmarks, showing both your alpha and tracking error separately. You can see whether your active trading decisions are adding value or just adding volatility without commensurate returns.
