If you've spent any time looking at portfolio statistics, you've encountered the Sharpe ratio. You've probably also encountered the Sortino ratio. They get presented side by side, like they're two ways of looking at the same thing.

They're not. They measure different things. And which one you should pay attention to depends on what you actually care about.

This post explains both, in plain English, and helps you decide which number deserves more weight in your decisions.

The basic question both ratios try to answer

Investment returns alone don't tell you anything. A 20% return in a year when you took unlimited risk is not impressive. A 20% return in a year when you took moderate risk is impressive. Risk-adjusted returns are what matter.

Sharpe and Sortino are both attempts to measure risk-adjusted returns. They both answer some version of the question: "How much excess return did this portfolio produce per unit of risk it took?"

The difference is how each defines "risk."

How Sharpe defines risk

The Sharpe ratio uses standard deviation of returns as its risk measure. Standard deviation captures the variability of returns over time. A portfolio with a 1% standard deviation has very stable returns. A portfolio with a 20% standard deviation swings wildly.

The formula is roughly: (portfolio return minus risk-free rate) divided by the standard deviation of the portfolio's returns.

The "risk-free rate" is what you'd earn on a 3-month T-bill, which represents return with essentially zero risk. The numerator (portfolio return minus risk-free rate) is the "excess return" the portfolio earned beyond the risk-free option. The denominator (standard deviation) is the variability of those returns.

A higher Sharpe ratio is better. A portfolio with Sharpe of 1.5 is generating more excess return per unit of variability than a portfolio with Sharpe of 0.7.

The catch with Sharpe

Standard deviation treats upward volatility and downward volatility the same. A month where your portfolio jumps 15% counts as just as much "risk" as a month where it drops 15%. Mathematically that's neat. Intuitively, most investors don't actually mind upside volatility.

Imagine two portfolios that both return 15% per year. Portfolio A goes up 1.25% every single month, like clockwork. Portfolio B has months where it gains 8% and months where it loses 4%, but ends the year at the same place. Sharpe ratio penalizes Portfolio B heavily because of the volatility, even though all of that volatility is good months mixed with bad months that average out to the same return.

Most investors don't actually want to penalize the upside. They want a measure of how much downside risk they took to earn the return. Standard deviation overstates "risk" because it includes upside swings as a negative.

How Sortino fixes this

The Sortino ratio uses downside deviation instead of standard deviation. Downside deviation only counts the variability of losing months (or losing periods, depending on how it's calculated). Upside volatility doesn't count.

The formula is similar to Sharpe: (portfolio return minus risk-free rate) divided by downside deviation. The numerator is unchanged. The denominator is now smaller, because it only includes the bad volatility.

In practice, Sortino is always equal to or higher than Sharpe for the same portfolio. The difference between them tells you how much of the portfolio's volatility is actually upside (which you don't care about) versus downside (which you do).

A portfolio with similar Sharpe and Sortino ratios has volatility evenly distributed between up and down months. A portfolio with a much higher Sortino than Sharpe has mostly upside volatility, with the downside relatively contained. That's a good thing.

A practical example

Let's compare two of our portfolios:

Core 20: Sharpe ratio of 0.78, Sortino ratio of 1.17. The Sortino is meaningfully higher than the Sharpe, indicating that the portfolio's volatility skews toward the upside. The downside is more contained than the overall variability would suggest.

Tepper Tactical: Sharpe ratio of 0.63, Sortino ratio of 1.12. Similar pattern. Higher Sortino than Sharpe means the portfolio has a friendly distribution: more good volatility than bad.

In both cases, the Sharpe ratio understates the actual risk-adjusted appeal of the portfolio because Sharpe is treating beneficial upward swings as "risk" we wanted to avoid. The Sortino ratio, which ignores upside swings, gives a cleaner picture of how much downside risk was actually taken.

The same logic applies in reverse. A portfolio with similar Sharpe and Sortino ratios is producing return with symmetric volatility (which is just average). A portfolio where Sharpe is dramatically higher than Sortino is rare and would suggest the portfolio is producing more downside volatility than upside, which is unusual and worth investigating.

When to use which

For most evaluation purposes, look at both, but lean on Sortino.

Sharpe is useful when you want a quick read on overall risk-adjusted return. It's standardized, it's familiar, and it's what most professional materials cite. If you're comparing two portfolios at a high level, Sharpe is fine.

Sortino is useful when you actually care about how much downside the portfolio takes. For most individual investors, this is the question that matters. You don't lose sleep over upward volatility. You lose sleep when your account is dropping. Sortino measures the variability of the losing periods specifically, which maps more closely to the lived experience of investing.

The relationship between the two ratios is also informative. A portfolio with a much higher Sortino than Sharpe is producing asymmetric returns (more upside variance than downside). That's a feature, not a bug. A portfolio with Sortino close to Sharpe has symmetric volatility. A portfolio with Sortino lower than Sharpe is statistically unusual and probably warrants closer inspection.

What the numbers mean in absolute terms

Both ratios are dimensionless. There's no "absolute" good or bad value. They're meaningful in comparison, not in isolation. But here are some rough benchmarks:

  • Sharpe < 0.5: Below average. The portfolio isn't generating much excess return for the variability it's producing.
  • Sharpe 0.5 to 1.0: Average to good. Most diversified equity portfolios fall here over long periods.
  • Sharpe 1.0 to 2.0: Excellent. Persistent values in this range over multi-decade periods are rare.
  • Sharpe > 2.0: Suspicious. Either a short backtest, a very specific strategy, or a sign of survivorship bias / curve fitting.

Sortino values follow a similar scale, generally a bit higher than Sharpe for the same portfolio, but the same heuristics apply.

The S&P 500 has a long-term Sharpe ratio of roughly 0.4 to 0.6, depending on the time period measured. Anything meaningfully above that is producing real excess risk-adjusted return.

Common pitfalls to avoid

A few things to watch for when reading Sharpe and Sortino numbers from any source:

Different risk-free rates. The "excess return" portion of both formulas depends on the risk-free rate used. Different sources use different rates. Make sure you're comparing apples to apples.

Different lookback windows. A Sharpe of 1.5 over five years can become 0.6 over twenty. The longer the window, the more reliable the number. One-year or three-year Sharpe ratios are mostly noise.

Survivorship bias. Sharpe ratios calculated from indexes that drop failing constituents will look better than they should. The S&P 500 itself has survivorship bias built in. Real-money portfolios have it less.

Backtest curve fitting. A strategy designed and optimized using historical data can show a beautiful Sharpe ratio in backtest and collapse to nothing live. The longer and more out-of-sample the backtest, the more reliable the ratio.

Weighting methodology. Equal-weighted portfolios will have different Sharpes than market-cap weighted versions of the same holdings. Different sources sometimes don't disclose which they're showing.

The bottom line

The Sharpe ratio is a useful, widely cited measure of risk-adjusted return. The Sortino ratio is a better measure of risk-adjusted return for most individual investors because it only penalizes the downside variability, which is what investors actually experience as risk.

Use Sharpe for quick comparisons and high-level reading. Use Sortino when you want to understand how much downside risk a portfolio is actually taking. Pay attention to the relationship between them: a meaningfully higher Sortino than Sharpe is a sign of friendly asymmetry.

Either way, neither ratio alone tells the full story. Combined with maximum drawdown, alpha, beta, and a basic sense for the strategy's logic, they help you read a portfolio honestly. Without that context, both ratios can be cherry-picked or window-dressed to make any strategy look better than it is.

The numbers matter. The judgment about which numbers to trust matters more.