Downside Deviation: Risk Below a Target Return

Key Takeaways

• Downside deviation counts only returns below a minimum acceptable return, treating upside as zero risk.
• It squares shortfalls below the target, divides by the total number of observations, and takes the square root.
• It is the denominator of the Sortino ratio, replacing standard deviation.
• A common error is dividing by only the count of bad periods rather than all periods.

What It Is

Standard deviation treats every move away from the average as risk, including big gains. Downside deviation fixes that by measuring dispersion only below a threshold called the minimum acceptable return (MAR), sometimes the risk-free rate, sometimes zero, sometimes a goal return.

Any period that returns above the MAR contributes nothing to the risk figure. Only shortfalls below the MAR matter. This matches how most investors actually think: a year of strong gains is not a problem, but a year of losses is.

Downside deviation is closely related to semideviation, the difference being that semideviation uses the mean as its reference point while downside deviation lets you set any target.

The Intuition

Imagine two funds with identical standard deviations. One got there with wild gains and modest losses. The other got there with modest gains and frequent steep losses. Standard deviation calls them equally risky, which feels wrong.

Downside deviation separates them. By zeroing out everything above the target, it focuses purely on the shortfalls that hurt. The fund with frequent steep losses shows a high downside deviation, while the fund with mostly upside volatility shows a low one. The measure rewards the path investors actually prefer.

How Downside Deviation Works

The formula squares each period's shortfall below the MAR, averages over all periods, and takes the square root:

Downside Deviation = sqrt( (1/n) x sum of [min(Rt - MAR, 0)]^2 )

Where Rt is the return in period t, MAR is the minimum acceptable return, n is the total number of observations, and the min function keeps only the negative shortfalls.

The key subtlety is the denominator. The standard convention divides by n, the total number of periods, not by the count of below-target periods. Dividing by only the bad periods inflates the figure and is a frequent source of mismatched numbers across reports. Whatever convention is used must be stated.

Worked Example

A portfolio has 6 monthly returns: plus 4 percent, minus 2 percent, plus 1 percent, minus 5 percent, plus 3 percent, minus 1 percent. The MAR is 0 percent.

Keep only the shortfalls below zero, then square them:

(-2%)^2 = 0.0004
(-5%)^2 = 0.0025
(-1%)^2 = 0.0001

The three positive months contribute zero. Sum the squared shortfalls and divide by all 6 periods:

(0.0004 + 0.0025 + 0.0001) / 6 = 0.0030 / 6 = 0.0005

Take the square root:

Downside Deviation = sqrt(0.0005) = 0.0224, or about 2.24%

If you had wrongly divided by only the 3 down months, you would get sqrt(0.0010) = 3.16 percent, overstating the risk by nearly half. The denominator choice clearly matters.

Common Mistakes

1. Dividing by the wrong count. The standard formula divides by all periods, not just the losing ones. Using only the down periods inflates downside deviation and breaks comparability.
2. Forgetting to set the MAR. The result changes with the target. A MAR of zero, the risk-free rate, or a goal return give different numbers, so the threshold must be stated.
3. Confusing it with semideviation. Semideviation measures below the mean, downside deviation below a chosen target. They coincide only when the MAR equals the mean.
4. Annualizing carelessly. Like standard deviation, monthly downside deviation must be scaled correctly to annualize. Mixing frequencies corrupts any ratio built on it.
5. Assuming it captures tail risk fully. Downside deviation measures average shortfall magnitude, not the worst case. Pair it with drawdown or value-at-risk for the extremes.

Frequently Asked Questions

What is downside deviation in simple terms? Downside deviation measures how much returns swing below a target you set, ignoring all the gains above it. A higher number means deeper or more frequent shortfalls below your acceptable return.

How does downside deviation affect investment decisions? It gives a risk figure that only counts the outcomes investors actually dislike, the shortfalls. As the worked example shows, focusing on below-target moves can reveal that two funds with the same standard deviation have very different real downside risk.

What is a real-world example of downside deviation? A portfolio with monthly returns including minus 2, minus 5, and minus 1 percent against a zero target has a downside deviation of about 2.24 percent when you divide by all six months.

How can investors use downside deviation effectively? State the minimum acceptable return clearly, always divide by the total number of periods, and keep the frequency consistent before annualizing. Then use it as the risk input for the Sortino ratio.

How is downside deviation different from standard deviation? Standard deviation counts every deviation from the mean, including gains, as risk. Downside deviation counts only the shortfalls below a chosen target, so upside volatility adds nothing to the figure.

Sources

1. Financial Edge. "Semi-Deviation Explained." https://www.fe.training/free-resources/asset-management/semi-deviation/
2. Breaking Down Finance. "Downside Deviation." https://breakingdownfinance.com/finance-topics/performance-measurement/downside-deviation/
3. CFA Institute. "The Sortino Ratio." https://rpc.cfainstitute.org/sites/default/files/-/media/documents/code/gips/the-sortino-ratio.pdf
4. PerformanceAnalytics. "DownsideDeviation." https://timelyportfolio.github.io/PerformanceAnalytics/reference/DownsideDeviation.html

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.