Understanding Backtest Results: Sharpe Ratio, Sortino, and Max Drawdown

You have built a trading strategy, fed it historical data, and hit the backtest button on your backtesting platform. Now you are staring at a wall of numbers: Sharpe ratio 1.34, Sortino 1.87, max drawdown -23.4%, win rate 58%, profit factor 1.62. What does it all mean? More importantly, are these numbers any good?

Understanding backtest results is just as important as running the backtest itself. A strategy with a high total return but terrible risk-adjusted metrics is a ticking time bomb. Conversely, a modest-return strategy with excellent risk metrics might be exactly what you need for consistent, sustainable growth.

This guide breaks down every key backtest metric, explains what each one tells you, and provides concrete benchmarks for crypto trading strategies.

Why Metrics Matter More Than Total Return

Total return is the number every beginner fixates on. "My strategy made 300% in a year!" Impressive, right? Maybe not.

Total return tells you nothing about:

• How much risk you took to achieve that return
• How deep the drawdowns were along the way
• Whether the returns were consistent or came from one lucky trade
• Whether the strategy would survive a worst-case scenario

Two strategies can both return 100% annually, but one might achieve it with smooth, consistent growth and -10% max drawdown, while the other swings between +200% and -60% before ending at +100%. The second strategy would psychologically destroy most traders and would be one bad day away from ruin.

Risk-adjusted metrics tell the real story. When you use Sentinel Bot's comprehensive backtesting engine, all of these metrics are calculated automatically. Your job is to understand what they mean.

Sharpe Ratio Explained

The Sharpe ratio is the most widely used risk-adjusted return metric in finance. It measures how much excess return you earn per unit of risk (volatility).

The Formula

Sharpe Ratio = (Strategy Return - Risk-Free Rate) / Standard Deviation of Returns

• Strategy Return: Your annualized return
• Risk-Free Rate: The return of a "riskless" investment (often set to 0% in crypto or the prevailing stablecoin yield rate, roughly 3-5%)
• Standard Deviation: The volatility of your returns, measuring how much they fluctuate

What the Number Means

A Sharpe ratio of 1.5 means you earn 1.5 units of return for every unit of risk you take.

| Sharpe Ratio | Interpretation |

|---|---|

| < 0 | Losing money on a risk-adjusted basis |

| 0 - 0.5 | Poor, not worth the risk |

| 0.5 - 1.0 | Below average but may be acceptable |

| 1.0 - 1.5 | Good, typical of solid strategies |

| 1.5 - 2.0 | Very good, above average |

| 2.0 - 3.0 | Excellent, top-tier strategy |

| > 3.0 | Exceptional (verify it is not overfit) |

Sharpe Ratio Pitfalls

It penalizes upside volatility. The Sharpe ratio uses total standard deviation, meaning big winning trades increase volatility and can actually lower the Sharpe ratio. A strategy that earns 1% most days but occasionally earns 20% will have a lower Sharpe than one that earns 1% every day, even though most traders would prefer the occasional windfall.

It assumes normal distribution. Crypto returns are famously non-normal, with fat tails and skewness. The Sharpe ratio underestimates risk in crypto because it does not account for the probability of extreme events.

It is time-period sensitive. A strategy tested during a calm bull market will show a higher Sharpe than the same strategy tested during a volatile bear market, even if the underlying edge is identical.

For a more nuanced view, use the Sharpe ratio alongside the Sortino ratio, especially when evaluating strategies on multiple exchange pairs.

Sortino Ratio

The Sortino ratio addresses the Sharpe ratio's biggest flaw: penalizing upside volatility. Instead of using total standard deviation, it uses only downside deviation.

The Formula

Sortino Ratio = (Strategy Return - Risk-Free Rate) / Downside Deviation

The only difference is the denominator. Downside deviation measures the volatility of returns below a target return (typically 0%), ignoring positive returns entirely.

Why Sortino is Better for Crypto

Crypto strategies often produce asymmetric returns: many small losses offset by fewer but larger wins. The Sharpe ratio penalizes these large wins by increasing the standard deviation. The Sortino ratio does not.

Example:

A trend-following strategy produces daily returns of: -0.5%, -0.3%, -0.2%, +0.1%, +5.0%

The 5% winning day dramatically increases total volatility, lowering the Sharpe ratio. But the Sortino ratio only considers the three losing days, giving a much more accurate picture of the strategy's risk profile.

| Sortino Ratio | Interpretation |

|---|---|

| < 0 | Losing money |

| 0 - 1.0 | Poor to average |

| 1.0 - 2.0 | Good |

| 2.0 - 3.0 | Very good |

| > 3.0 | Excellent |

When reviewing results from Sentinel's backtesting dashboard, compare both Sharpe and Sortino. If the Sortino is significantly higher than the Sharpe, your strategy produces asymmetric returns (big wins, small losses), which is generally desirable.

Max Drawdown

Max drawdown (MDD) is the largest peak-to-trough decline in your portfolio during the backtest period. It answers the question: "What is the worst pain I would have experienced?"

How It is Calculated

Max Drawdown = (Trough Value - Peak Value) / Peak Value * 100

For example, if your portfolio grows from $10,000 to $15,000, then drops to $11,000, before eventually rising to $20,000:

• Peak: $15,000
• Trough: $11,000
• Max Drawdown: ($11,000 - $15,000) / $15,000 = -26.7%

Even though the strategy ended at $20,000 (100% total return), at one point you were sitting on a -26.7% drawdown from your highest portfolio value. Could you stomach that?

Why Max Drawdown Matters

Psychological survival. Research consistently shows that traders abandon strategies during drawdowns. If you cannot emotionally handle a 30% drawdown, do not deploy a strategy that showed a 25% drawdown in backtesting, because live drawdowns are almost always worse than backtested ones.

Capital survival. The math of recovery is brutal:

| Drawdown | Required Return to Recover |

|---|---|

| -10% | +11.1% |

| -20% | +25.0% |

| -30% | +42.9% |

| -40% | +66.7% |

| -50% | +100.0% |

| -75% | +300.0% |

A -50% drawdown requires a 100% return just to break even. This is why drawdown management is critical, especially when backtesting leveraged strategies where drawdowns are amplified.

Max Drawdown Benchmarks for Crypto

| Max Drawdown | Assessment |

|---|---|

| < -10% | Excellent, very conservative |

| -10% to -20% | Good, suitable for most traders |

| -20% to -30% | Acceptable for aggressive strategies |

| -30% to -50% | High risk, only for experienced traders |

| > -50% | Dangerous, reconsider the strategy |

Remember: live drawdowns typically exceed backtested drawdowns by 1.5-2x due to slippage, execution delays, and market conditions not captured in historical data.

Win Rate

Win rate is the percentage of trades that are profitable. It is intuitive and easy to understand, but deeply misleading when viewed in isolation.

The Win Rate Trap

A 90% win rate sounds fantastic. But if your average win is $10 and your average loss is $100, you need 10 winning trades to offset one loss. Your expected value per trade is:

EV = (0.90 * $10) - (0.10 * $100) = $9 - $10 = -$1

Despite a 90% win rate, you lose money on average. Conversely, trend-following strategies often have win rates below 40% but are highly profitable because their average win is 3-5x their average loss.

Win Rate in Context

Win rate must be evaluated alongside average win/loss ratio:

| Win Rate | Required Avg Win/Loss Ratio | Strategy Type |

|---|---|---|

| 30% | > 2.33:1 | Trend following |

| 40% | > 1.50:1 | Momentum |

| 50% | > 1.00:1 | Breakeven at 1:1 |

| 60% | > 0.67:1 | Mean reversion |

| 70% | > 0.43:1 | Scalping |

When building strategies with Sentinel's strategy blocks, consider which style your strategy follows and set expectations accordingly. A 35% win rate with a 3:1 reward-to-risk ratio is an excellent strategy.

Profit Factor

Profit factor is the ratio of gross profits to gross losses. It is one of the cleanest metrics for evaluating strategy quality.

The Formula

Profit Factor = Gross Profits / Gross Losses

• Profit factor = 1.0: Breakeven (profits exactly equal losses)
• Profit factor > 1.0: Profitable
• Profit factor < 1.0: Losing money

Benchmarks

| Profit Factor | Assessment |

|---|---|

| < 1.0 | Unprofitable |

| 1.0 - 1.2 | Marginal, likely unprofitable after real-world costs |

| 1.2 - 1.5 | Acceptable but thin |

| 1.5 - 2.0 | Good |

| 2.0 - 3.0 | Very good |

| > 3.0 | Excellent (verify not overfit) |

A profit factor below 1.2 in backtesting is a red flag. Real-world conditions (slippage, partial fills, execution delays) will erode that thin margin. Target a profit factor above 1.5 to have a meaningful buffer.

Putting It All Together

No single metric tells the full story. Here is how to evaluate a strategy holistically.

The "Good Strategy" Profile

A solid crypto trading strategy should exhibit most of these characteristics:

• Sharpe ratio > 1.0
• Sortino ratio > 1.5
• Max drawdown < -25%
• Win rate appropriate for strategy type (see table above)
• Profit factor > 1.5
• At least 50+ trades in the backtest (statistical significance)
• Consistent performance across different market regimes

The "Too Good to Be True" Profile

Be suspicious if you see:

• Sharpe ratio > 4.0
• Win rate > 80% with high average win/loss ratio
• Max drawdown < -3%
• Profit factor > 5.0

These numbers usually indicate overfitting, look-ahead bias, or a data error. See our guide on common backtesting mistakes to diagnose the issue.

The "Reject" Profile

Do not deploy a strategy that shows:

• Sharpe ratio < 0.5
• Max drawdown > -40% (especially with leverage)
• Profit factor < 1.2
• Fewer than 30 trades in the backtest period
• Significantly worse performance in the second half of the test period

Good vs Bad Metric Values: Quick Reference

| Metric | Bad | Acceptable | Good | Excellent |

|---|---|---|---|---|

| Sharpe Ratio | < 0.5 | 0.5 - 1.0 | 1.0 - 2.0 | > 2.0 |

| Sortino Ratio | < 0.5 | 0.5 - 1.5 | 1.5 - 3.0 | > 3.0 |

| Max Drawdown | > -40% | -25% to -40% | -10% to -25% | < -10% |

| Profit Factor | < 1.2 | 1.2 - 1.5 | 1.5 - 2.5 | > 2.5 |

| Recovery Factor | < 1.0 | 1.0 - 2.0 | 2.0 - 4.0 | > 4.0 |

Frequently Asked Questions

What is more important, Sharpe ratio or max drawdown?

For most traders, max drawdown is more important because it determines whether you will psychologically and financially survive the strategy's worst period. A high Sharpe ratio with a devastating max drawdown means the strategy is risky despite good risk-adjusted returns on average. Always evaluate both together.

How many trades does my backtest need for reliable metrics?

At minimum, 30 trades for basic statistical significance, but 100+ is strongly recommended. Below 30 trades, metrics like win rate and profit factor can be heavily skewed by a few outlier trades. The more trades, the more confidence you can have in the results.

Why is my Sortino ratio much higher than my Sharpe ratio?

This indicates your strategy produces asymmetric returns: large winning trades that increase total volatility (hurting Sharpe) but do not contribute to downside volatility (preserved in Sortino). This is generally a positive sign, as it means your big moves are in the right direction.

Should I compare metrics across different trading pairs?

Yes, but with caution. Different pairs have different volatility profiles. A Sharpe ratio of 1.5 on BTC/USDT is more impressive than 1.5 on a low-cap altcoin because BTC is less volatile. When comparing strategies across pairs using Sentinel's multi-exchange support, normalize for each pair's volatility.

Ready to analyze your strategy's performance with professional-grade metrics? Sentinel Bot calculates Sharpe, Sortino, max drawdown, profit factor, and more for every backtest. Sign up free and start making data-driven trading decisions.

Disclaimer: This article is for educational purposes only and does not constitute financial advice. Past performance metrics from backtesting do not guarantee future results. Cryptocurrency trading involves significant risk of loss. Always do your own research.