A risk adjusted return measure that judges a strategy only on its downside swings, not its upside.
The Sortino ratio measures how much return a strategy earns for the downside risk it takes. It is the Sharpe ratio's smarter cousin, the same idea but counting only the swings that lose you money, since no trader has ever complained about a violent move in their favor. A higher Sortino means more reward for the pain you actually feel. Most healthy systems sit somewhere between 1 and 2 on a per trade basis.
It measures return divided by downside deviation, which is the volatility of only your losing moves. Standard deviation, the input the Sharpe ratio uses, treats a big winning week as risk. Downside deviation does not.
That one change matters for trend systems and anything with a fat right tail. A strategy that grinds small losses and occasionally explodes higher looks reckless to Sharpe and reasonable to Sortino. Sortino is simply closer to how risk feels from the seat.
Use Sortino when your returns are lopsided, which most real strategies are. Use Sharpe when you want to compare against the wider industry, since it is the common language everyone already speaks.
Sharpe punishes upside volatility, and that is its flaw. A momentum strategy can post a mediocre Sharpe and a strong Sortino at the same time, and the Sortino is the one telling the truth. Neither number means much on a handful of trades, so do not read either one off 25 closed trades and call it settled. The metrics and tools guide walks through how these fit together.
On a per trade basis, above 1 is solid and above 2 is strong. Treat anything far higher on a short record with suspicion.
Annualized figures look bigger and follow different conventions, so a Sortino of 2 per trade is not the same as a Sortino of 2 reported annually. Always check which basis you are looking at before you compare two systems, because the same strategy can wear very different numbers depending on the convention.
Quantprove computes Sortino from your closed trades and uses it inside your Edge Score as a core consistency input, capped at 1.80 so one quiet, lucky stretch cannot dominate the grade.
It pairs naturally with maximum drawdown. Sortino tells you the quality of the ride, drawdown tells you the worst of it, and together they describe risk better than either does alone.