What Is Downside Deviation? Volatility That Only Counts Losses

Downside deviation is a one-sided cousin of standard deviation. Instead of measuring how far all your results scatter from the average, it measures only the spread of results that fall below a target — usually zero. Big winners are ignored; only the painful side counts.

Why one-sided volatility is useful

Plain standard deviation treats a +5R winner as just as much "risk" as a −5R loss, which feels wrong — nobody loses sleep over an unexpectedly large profit. Downside deviation fixes that by only penalising the swings that actually hurt, giving a truer picture of how rough a system feels to trade.

Where you'll see it

Downside deviation is the denominator of the Sortino ratio: return divided by downside deviation. A system with explosive winners and small, controlled losers has a low downside deviation and therefore a high Sortino — even if its ordinary standard deviation, and so its Sharpe ratio, looks unremarkable.

How to use it

• Lower is better — it means your losing trades are small and consistent.
• Compare it to total standard deviation — if downside deviation is much smaller, your volatility lives mostly on the upside, which is exactly where you want it.
• A rising downside deviation is a warning — your losses are getting larger or less predictable, often a position sizing or stop-discipline issue.