"This strategy returned 40% last year" tells you almost nothing on its own. Forty percent with wild, terrifying drawdowns along the way is a completely different proposition from forty percent earned steadily. The Sharpe ratio is the standard attempt to put a number on that difference.
The idea in one sentence
Sharpe ratio measures return per unit of risk taken — specifically, average return in excess of a risk-free rate, divided by the volatility (standard deviation) of those returns. Two strategies can post the same total return with very different Sharpe ratios if one got there smoothly and the other got there via a rollercoaster.
Why this matters more than raw return
Raw return answers "how much did I make." Sharpe ratio answers "how much did I have to suffer to make it," which matters for two very practical reasons:
- Higher volatility means a higher chance of quitting at the worst time. A strategy with wild swings is one you're statistically more likely to abandon mid-drawdown, locking in the loss right before a recovery.
- Leverage can inflate raw returns without adding skill. A strategy earning 8% with low volatility can often be levered up to earn 16%+ with proportionally more volatility — the skill (Sharpe ratio) hasn't changed, only the risk taken to express it.
Rough intuition for the numbers
There's no universal "good" number — it depends on asset class and time horizon — but as a rough field guide: a Sharpe ratio below 0.5 is generally unremarkable, 1.0+ is solid, and anything reported well above 2.0 over a long, realistic sample deserves scrutiny rather than excitement, because it's uncommon and often a sign the backtest has a flaw (survivorship bias, unrealistic costs, or too short a sample) rather than a genuine edge.
Where it misleads
Sharpe ratio isn't a complete picture, for a few concrete reasons:
- It treats upside and downside volatility the same. A strategy that has occasional huge up-months looks "risky" by this measure even though nobody minds that kind of volatility. (The Sortino ratio, which only penalizes downside volatility, is a common fix.)
- It says nothing about drawdown depth or duration — a strategy can have a perfectly good Sharpe ratio while still enduring a brutal 40% drawdown along the way.
- It's sample-dependent. A Sharpe ratio computed over 20 trades is a much less reliable estimate than one computed over 500 — the number needs a large enough sample to mean anything, same as a win rate.
The practical takeaway
Never evaluate a strategy — or a backtest — on total return alone. Ask for the Sharpe ratio (and ideally max drawdown alongside it), and treat an implausibly high number as a reason to dig into the methodology, not a reason to get excited. That's exactly why every backtest run in Atradian's Strategy Lab reports Sharpe ratio, win rate, and max drawdown side by side, not just net P&L — a single flattering number is easy to produce; a full risk picture is what actually tells you whether a strategy is worth trusting.