What a Sharpe ratio is, and why a high return with a low Sharpe ratio is a red flag
A trading strategy that returned 50% sounds better than one that returned 20%. It isn't necessarily — not without knowing how bumpy the ride was to get there. That's what the Sharpe ratio measures, and it's the number that separates "this strategy has an edge" from "this strategy got lucky."
The definition, in plain terms
Sharpe ratio = (return above a risk-free baseline) ÷ (volatility of those returns). It's a return-per-unit-of-risk number, not a return number. A strategy that grinds out steady, consistent gains scores high. A strategy that lurches between big wins and big drawdowns — even if it nets out ahead — scores low, because the denominator (volatility) is large.
Rough field guide to the values: below 1 is generally considered weak, 1–2 is decent, above 2 is strong, and above 3 is rare enough to be treated with suspicion (backtests can produce numbers like this through overfitting, in a way live trading rarely replicates).
Why a high return with a low Sharpe ratio is specifically a warning sign
If a strategy shows a large return but a low Sharpe ratio, it means most of that return came from a small number of high-variance bets — the average outcome wasn't reliably good, the total just happens to look good because a few large swings landed in its favor. That's a strategy that could just as easily have shown a large loss under a slightly different run of luck. Total return alone can't distinguish "consistently good decisions" from "a volatile process that happened to land on the right side this time" — Sharpe ratio is specifically the number built to make that distinction, which is exactly why it's worth checking before trusting a good-looking return number at all.
A real example, not a hypothetical
Backtesting two simple strategies (an SMA crossover and an RSI mean-reversion rule) against BTC-USD, ETH-USD, and SPY over two years produced this:
| symbol | strategy | return | max drawdown | Sharpe | |---|---|---|---|---| | ETH-USD | sma_crossover | 52.8% | -47.4% | 0.71 | | SPY | sma_crossover | 21.3% | -9.9% | 1.01 |
The ETH result has the bigger headline return — more than double SPY's. But its Sharpe ratio is lower (0.71 vs 1.01), and its max drawdown is nearly five times worse (-47.4% vs -9.9%). Read as raw return, ETH looks like the better strategy. Read with Sharpe and drawdown included, SPY was the more reliable one — smaller swings, better return per unit of risk taken, even though its total number was smaller. That gap between "which one has the bigger number" and "which one is actually better risk-adjusted" is the entire reason this metric exists.
The takeaway
Never evaluate a strategy — or trust someone else's backtest — on total return alone. Return tells you the destination; Sharpe ratio tells you how rough the road was to get there, and a rough road is a real cost even when the destination looks good.
Sources: standard Sharpe ratio definition (excess return over volatility); example figures
are this project's own real backtest output (trading/results/backtest_report.json),
computed and verified while building it, not illustrative numbers.