What are the key limitations of the Sharpe ratio as a performance measure?

Question

AcadiFi · Accepted Answer

The **Sharpe ratio** — defined as (R_p - R_f) / sigma_p — is the most widely used risk-adjusted performance metric, but it has well-documented shortcomings that every CFA candidate should understand.

**Limitation 1: Assumes Normal Returns**

The Sharpe ratio uses standard deviation as the sole risk measure, which only fully describes risk when returns are normally distributed. For strategies with **negative skewness** (e.g., short volatility, merger arbitrage) or **excess kurtosis** (fat tails), standard deviation underestimates true risk.

**Example:** Thornfield Macro Fund and Eastwood Volatility Fund both have Sharpe ratios of 1.2. But Thornfield has positively skewed returns (occasional large gains) while Eastwood has negatively skewed returns (occasional catastrophic losses). The Sharpe ratio treats them identically.

**Limitation 2: Symmetric Treatment of Upside and Downside**

Standard deviation penalizes upside volatility equally to downside volatility. An investor who earns +15% one month and +3% the next has "high volatility" even though both months were profitable.

**Limitation 3: Susceptible to Manipulation**

| Technique | Effect on Sharpe | True Risk Change |
|---|---|---|
| Selling deep OTM puts | Increases returns with rare large losses | Much riskier |
| Smoothing returns (illiquid assets) | Artificially low volatility | No actual risk reduction |
| Using longer return intervals | Lower annualized volatility | Same underlying risk |
| Compounding before dividing | Different ratio than arithmetic | Numerical artifact |

```mermaid
flowchart TD
    A["Sharpe Ratio Limitations"] --> B["Non-normal returns: skew & kurtosis ignored"]
    A --> C["Symmetric: penalizes upside vol"]
    A --> D["Manipulable: return smoothing, option selling"]
    A --> E["Negative Sharpe: ranking breaks down"]
    A --> F["Benchmark-agnostic: ignores tracking error"]
```

**Limitation 4: Meaningless When Negative**

When Sharpe ratios are negative (excess return < 0), ranking funds by Sharpe can give perverse results. A fund with return -2% and sigma 10% has Sharpe = -0.20, while a fund with return -2% and sigma 20% has Sharpe = -0.10. The "better" Sharpe belongs to the fund with **higher** volatility — clearly wrong.

**Limitation 5: No Benchmark Sensitivity**

The Sharpe ratio compares performance to the risk-free rate, not to a relevant benchmark. A US equity fund with Sharpe 0.8 might just be riding a bull market — the information ratio (vs. S&P 500) might be negative.

**Better Alternatives:**

| Metric | Advantage Over Sharpe |
|---|---|
| **Sortino ratio** | Uses downside deviation only |
| **Information ratio** | Measures active return per unit of tracking error |
| **Calmar ratio** | Uses maximum drawdown instead of volatility |
| **Omega ratio** | Captures the entire return distribution |

**Exam Tip:** CFA questions often present two funds with similar Sharpe ratios but different return distributions and ask which metric would better differentiate them.

For more on performance evaluation, check our CFA question bank.

What are the key limitations of the Sharpe ratio as a performance measure?

Master Level III with our CFA Course

Related Questions

Practice Questions

Technique	Effect on Sharpe	True Risk Change
Selling deep OTM puts	Increases returns with rare large losses	Much riskier
Smoothing returns (illiquid assets)	Artificially low volatility	No actual risk reduction
Using longer return intervals	Lower annualized volatility	Same underlying risk
Compounding before dividing	Different ratio than arithmetic	Numerical artifact

Metric	Advantage Over Sharpe
Sortino ratio	Uses downside deviation only
Information ratio	Measures active return per unit of tracking error
Calmar ratio	Uses maximum drawdown instead of volatility
Omega ratio	Captures the entire return distribution