Target Downside Deviation: Formula and Worked Example

Key Takeaways

• Target downside deviation (TDD) measures the dispersion of returns that fall below a minimum acceptable return (MAR); returns at or above the MAR contribute zero.
• The formula squares only the shortfalls, averages them over all periods, and takes the square root, a one-sided "semi-deviation" relative to a target.
• The MAR is a deliberate choice (0%, the risk-free rate, or a required hurdle); raising the MAR raises the TDD because more periods count as shortfalls.
• TDD is the denominator of the Sortino ratio and a better risk gauge than standard deviation for asymmetric or goal-based portfolios.

What It Is

Target downside deviation is the square root of the average squared shortfall below a target return. Where standard deviation measures dispersion around the mean in both directions, TDD measures dispersion in one direction only, below the MAR. Formally:

TDD = √( (1/n) · Σ [min(0, R_i − MAR)]² )

Every period at or above the MAR drops out of the sum (its shortfall is zero); every period below it enters as a squared miss.

The Intuition

Investors do not fear volatility as such; they fear losses, or falling short of a goal. A pension that must earn 6% does not care how far above 6% it lands, only how far below. Target downside deviation encodes that asymmetry directly: it ignores every period that met the target and penalizes only the shortfalls, scaled by how large they were.

How It Works

The calculation is five steps:

1. Choose the MAR, 0%, the risk-free rate, or a required return.
2. Compute each shortfall as min(0, R_i − MAR): zero if the period met the target, negative if it missed.
3. Square each shortfall so larger misses are penalized disproportionately.
4. Average over all n periods, not just the down periods. Dividing by the full n is the target-semideviation convention, and it is deliberate: a strategy that rarely falls short is rewarded for it.
5. Take the square root to return to return units.

Worked Example

Twelve monthly returns (%): 3, −1, 2, −4, 1, 0, 5, −2, 2, −3, 4, 1, with a MAR of 0.5%.

Only months below 0.5% produce a shortfall. The shortfalls (R − 0.5) are −1.5, −4.5, −0.5, −2.5, −3.5; their squares are 2.25, 20.25, 0.25, 6.25, 12.25, summing to 41.25. Divide by n = 12 → 3.44, and take the square root → TDD ≈ 1.85% per month (about 6.4% annualized, ×√12).

Now change the MAR to 0%. Only the outright-negative months count (−1, −4, −2, −3): squares 1 + 16 + 4 + 9 = 30, ÷12 = 2.5, √ = 1.58%. Raising the MAR from 0% to 0.5% lifted TDD from 1.58% to 1.85%, a higher bar reclassifies near-flat months as shortfalls. The MAR is not a footnote; it drives the answer.

Common Mistakes

1. Dividing by the number of downside periods instead of all periods. That inflates TDD and breaks comparability with the standard Sortino convention.
2. Leaving the MAR unstated. TDD without its target is uninterpretable; always report the MAR next to it.
3. Mixing frequencies. Compute at one frequency (say monthly) and annualize by √12, never blend monthly shortfalls with an annual target.
4. Treating TDD as a probability. It is a magnitude of shortfall, not the odds of a loss; pair it with drawdown or VaR for a fuller risk picture.

Frequently Asked Questions

Q: What is target downside deviation in simple terms? It is the one-sided cousin of standard deviation: it measures how much returns vary below a target you set (the MAR), ignoring everything at or above that target.

Q: How is target downside deviation different from standard deviation? Standard deviation counts dispersion in both directions around the mean; target downside deviation counts only shortfalls below the MAR, so upside volatility never inflates it.

Q: What MAR should I use for target downside deviation? Common choices are 0% (any loss), the risk-free rate (opportunity cost), or a required return or hurdle (goal-based). State it explicitly, because the same returns give different results at different MARs.

Q: How does target downside deviation relate to the Sortino ratio? It is the denominator of the Sortino ratio: Sortino = (return − MAR) ÷ target downside deviation. A lower TDD, all else equal, produces a higher Sortino.

Q: What is a real-world use of target downside deviation? Liability-driven investors, pensions, insurers, endowments with a spending rate, use it to measure risk relative to the return they must hit, rather than penalizing the upside they welcome.

Sources

1. Investopedia. "Downside Deviation." https://www.investopedia.com/terms/d/downside-deviation.asp
2. Investopedia. "Sortino Ratio." https://www.investopedia.com/terms/s/sortinoratio.asp
3. Investopedia. "Semideviation." https://www.investopedia.com/terms/s/semideviation.asp

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.