The Sortino Ratio — A Metric That Looks Only at Downside Risk

The Sortino Ratio is a 'risk-adjusted return' metric that measures return relative to risk. It's a variant of the widely used Sharpe Ratio, and the core difference is just one thing.

The Sharpe ratio treats 'all volatility' as risk equally, whether the price spikes up or down. But think about it—what an investor is truly afraid of is 'losses,' not 'making a large gain.' A sudden +15% in one month is a welcome thing, yet the Sharpe ratio penalizes it as being just as 'unstable' as -15%.

The Sortino ratio considers this unfair. So it counts only volatility that fell 'below' a target return as risk. It does not penalize upside volatility.

The Sortino ratio's formula looks like this.

Sortino ratio = (average return − target return) ÷ downside deviation

Breaking it down one by one: First, the 'average return' is the average return you actually achieved. Second, the 'target return (MAR, Minimum Acceptable Return)' is the minimum threshold you want. Commonly used values are 0%, the risk-free rate (e.g., a Treasury rate of 4–5%), or a personal target such as '9% a year.' Third, the 'downside deviation' is the standard deviation calculated using only the movements in windows that fell below the target return.

Downside deviation is obtained by taking, from each period's return, the difference below the target return, squaring 'only the negative ones,' averaging them, and then taking the square root. The key is that months exceeding the target are treated as 0 and excluded from the calculation.

Here's a simple example. Suppose we set the target return at 2%.

Portfolio A: (average 10% − 2%) ÷ downside deviation 4% = 2.0 Portfolio B: (average 14% − 2%) ÷ downside deviation 7% = about 1.71

Looking only at average return, B is higher at 14%. But the Sortino ratio is higher for A (2.0) than for B (1.71). Why? Because A's 'downside turbulence (downside deviation of 4%)' was far smaller.

In other words, the Sortino ratio lets you compare two investments based on 'how much less it hurt when it fell.' For the same return, the side that wobbled less during declines receives a better score.

The higher the Sortino ratio, the better the 'return relative to downside risk.' Generally, 1 or above is considered good, 2 or above very good, and 3 or above excellent.

However, these cutoffs are not absolute standards. They differ slightly by source, and the value changes greatly depending on which period, target return, and data frequency (daily/monthly) you use. So simply comparing Sortino ratios computed under different conditions leads to misunderstandings.

One more thing: if downside cases are very few, the downside deviation can become abnormally small, exaggerating the Sortino ratio. With any metric, don't judge by one number alone—you must look together at 'how long and how deeply it actually hurt,' such as the maximum drawdown (MDD) and the loss duration.

The Sortino ratio was devised by Dr. Frank A. Sortino in the early 1980s and became widely known through a 1994 paper co-written with Lee N. Price, 'Performance Measurement in a Downside Risk Framework.'

The comparison, the Sharpe ratio, is the industry's basic yardstick, created in 1966 by Nobel laureate economist William Sharpe. Sortino's concern was clear: 'The Sharpe ratio penalizes big gains and big losses equally as risk—but what an investor really wants to avoid is only the losses.'

So in practice, the two metrics are viewed together. The Sharpe ratio shows efficiency relative to total volatility, and the Sortino ratio shows efficiency relative to downside risk, complementing each other.