Sortino takes your return above the risk-free rate and divides it by the downside deviation — the volatility of your losing periods only. Sharpe punishes a strategy for moving up violently, Sortino doesn't. That matters because real traders don't lose sleep over a 5% green day; they lose sleep over a 5% red day. By ignoring upside volatility, Sortino gives you a cleaner picture of how much bad swing you're actually paying for your returns.
A strategy earns 14% per year, the risk-free rate is 2%, and the downside deviation of returns (only the negative ones) is 6%.
How traders read the number: - Below 1.0 — the downside is eating most of your return. Edge is fragile. - 1.0 to 2.0 — workable. Most consistent retail strategies live here. - 2.0 to 3.0 — strong. Your losing periods are tame relative to the gains. - Above 3.0 — exceptional, but check the sample size before celebrating. Few losing trades in the window can flatter the number.
Rule of thumb: Sortino should always come out higher than Sharpe for the same strategy — if it doesn't, you've got a bug somewhere.
Sortino is the metric you check when Sharpe alone feels unfair. A momentum strategy that explodes upward in trend phases gets penalised by Sharpe for that very explosion — Sortino doesn't care, it only weighs the pullbacks against you. So when you're running anything trend-following, breakout, or asymmetric-payoff, Sortino is the more honest read.
In TradeOnyx your Sortino sits in the Overview tab under Advanced metrics, directly next to Sharpe. The two are placed side-by-side on purpose: one glance tells you whether the strategy's volatility is the kind that hurts (downside) or the kind you want more of (upside).
When Sortino comes out much higher than Sharpe, your edge is mostly upside-volatile — that's a strategy worth scaling. When the two are close, your gains and losses swing equally hard, and Sharpe is doing its job. Use the Overview period filter to compare both ratios across week, month, year, all-time and you'll see exactly when your strategy's character shifted.