Sortino Ratio vs Sharpe Ratio: Which One Should You Use?

When evaluating a strategy's risk-adjusted performance, the Sharpe ratio is almost always the first reflex. But it suffers from a well-known conceptual flaw: it penalizes upside and downside moves equally, even though no investor ever complains about positive volatility. The Sortino ratio was designed precisely to correct this bias.

In this article, we detail what the Sortino ratio is, how it fundamentally differs from the Sharpe ratio, how to calculate it step by step, and in which contexts it gives a more accurate picture of a strategy's real risk — with a worked numeric example.

What Is the Sortino Ratio?

The Sortino ratio measures a strategy's excess return per unit of risk, but counts only the return deviations below a minimum acceptable return threshold, known as the MAR. Everything above that threshold is not treated as risk at all.

This approach stems from a simple observation rooted in loss-aversion theory: investors do not fear variance as such, they fear losses. An exceptionally good month worries no one, even though it raises the standard deviation used by the Sharpe ratio.

Sortino vs Sharpe: The Fundamental Difference

The Sharpe ratio divides excess return by the total standard deviation of returns, which includes upside outperformance. The Sortino ratio instead replaces that total standard deviation with the 'downside deviation,' computed only from returns below the MAR.

In practice, two strategies with the same overall standard deviation can have very different Sortino ratios if one concentrates its volatility on the upside and the other on the downside. The Sharpe ratio would treat them as equivalent in risk, which is misleading for the end investor.

Why This Distinction Matters for Skewed Distributions

This difference becomes crucial for strategies with an asymmetric return distribution: options selling (steady carry with left-tail risk), trend-following (many small losses offset by rare extreme right-tail gains), or emerging-market carry trades. The Sharpe ratio penalizes these profiles in a way that poorly reflects the risk actually perceived by the investor.

How to Calculate the Sortino Ratio

The formula is: Sortino = (Rp − MAR) / Downside Deviation, where Rp is the strategy's annualized return. Downside deviation is calculated by keeping only periodic returns below the MAR, squaring each deviation, averaging them, then taking the square root — the one-sided analogue of a standard deviation.

The choice of MAR is not trivial: some practitioners use 0%, others the risk-free rate, others a strategy-specific target return. This choice has a direct, sometimes substantial, impact on the final ratio value, which is why the MAR used should always be disclosed when reporting a Sortino ratio.

Worked Example: Sharpe vs Sortino on the Same Strategy

Take a trend-following strategy with a 12% annualized return, a 2% risk-free rate, an 18% total standard deviation, and a 9% downside deviation (computed with a 0% MAR). Its Sharpe ratio is (12−2)/18 = 0.56.

Its Sortino ratio, however, is (12−0)/9 = 1.33. The gap is significant: the strategy looks average through the Sharpe lens but noticeably more attractive through the Sortino lens, precisely because a large share of its total volatility comes from occasional extreme gains rather than steady losses.

When to Favor Sortino Over Sharpe

The Sortino ratio is especially useful for comparing strategies with very different return profiles: a volatility-selling strategy against a directional one, or a macro fund against a high-frequency quantitative fund. It avoids wrongly penalizing profiles with positive skew.

It is also recommended when selecting managers or building a multi-strategy portfolio, where the goal is precisely to combine return sources whose 'felt' risk differs from the overall statistical risk captured by classic variance.

Limitations of the Sortino Ratio

The Sortino ratio is not flawless. It remains sensitive to the MAR choice, which can be used — deliberately or not — to artificially flatter a result by lowering that threshold. It also implicitly assumes the distribution of negative returns stays reasonably regular, a fragile assumption during extreme market shocks.

Finally, like the Sharpe ratio, the Sortino ratio does not correct for the number of trials or configurations tested while designing a strategy — a bias that metrics such as the Deflated Sharpe Ratio (DSR) specifically try to address, and which deserves an article of its own.

Conclusion

The Sortino ratio sharpens the reading of the risk-return trade-off by isolating what genuinely matters to investors: losses. It does not replace the Sharpe ratio but usefully complements it, especially for strategies with skewed return distributions. To compute and compare these ratios on your own return series, TrueVerdikt's free tools at /outils let you automate the calculations without a complex spreadsheet.