Understanding Results
After running a backtest, sweep, or walk-forward test, you'll see a results table and a chart. This page explains what every number means and how to interpret it.
The Metrics Table
Total P&L (Profit & Loss)
What it is: The total amount of money the strategy made or lost, in dollars.
How to read it:
- Positive = the strategy was profitable overall
- Negative = the strategy lost money
Context: A positive P&L is necessary but not sufficient. A strategy could make money overall but have terrifying drawdowns along the way.
Win Rate
What it is: The percentage of trades that ended in profit.
How to read it:
- 60% means 6 out of every 10 trades made money
- A win rate above 50% isn't always required — some strategies win rarely but win big
Common misconception: A high win rate doesn't guarantee profitability. A strategy could win 90% of the time but lose so much on the 10% that it's unprofitable overall.
Sharpe Ratio
What it is: How much return you get for the risk you take. It measures consistency — not just profit, but how smooth the ride is.
How to read it:
| Sharpe | Interpretation |
|---|---|
| Below 0 | Strategy loses money |
| 0 to 1 | Positive but not great — high volatility relative to returns |
| 1 to 2 | Good — decent return with manageable risk |
| Above 2 | Excellent — strong returns with relatively smooth performance |
Why it matters: Two strategies can both make 10%, but if one does it smoothly and the other swings wildly between +30% and -20%, the smooth one has a better Sharpe.
Profit Factor
What it is: Total money won divided by total money lost.
How to read it:
| Profit Factor | Interpretation |
|---|---|
| Below 1.0 | Losing money (losses > wins) |
| 1.0 to 1.5 | Marginal — barely breaking even after costs |
| 1.5 to 2.0 | Good |
| Above 2.0 | Very strong |
Simple way to think about it: For every $1 lost, how many dollars did you win? A profit factor of 2.0 means you won $2 for every $1 you lost.
Expectancy
What it is: The average profit (or loss) per trade, in dollars.
How to read it:
- Positive = on average, each trade makes money
- Negative = on average, each trade loses money
Why it matters: This is the simplest "is it worth it?" number. If expectancy is $5, then over 100 trades you'd expect to make about $500.
Max Drawdown
What it is: The largest peak-to-trough drop in your account balance during the test period. It answers: "What was the worst losing streak?"
How to read it:
- Shown as both a dollar amount and a percentage of capital
- A drawdown of 15% means at some point, the account fell 15% from its highest value
Why it matters: This tells you how much pain you'd need to endure. Even a profitable strategy can have periods where you're significantly down. If you can't stomach a 20% drawdown, don't use a strategy that has one in testing.
Rule of thumb: Expect real-world drawdowns to be 1.5-2x worse than what you see in backtesting.
Initial Capital / Final Capital
What it is: The starting balance and ending balance.
How to read it: Final capital minus initial capital equals Total P&L. These are included so you can see the scale relative to your investment.
Total Commission
What it is: The total fees paid across all trades.
Why it matters: Strategies that trade frequently can eat their profits in fees. If Total P&L is $500 but Total Commission is $400, the strategy only really made $100.
The Equity Curve
The chart below the metrics table shows your account balance over time.
How to read it:
- Upward slope = making money during that period
- Downward slope = losing money during that period
- Flat areas = no trades happening or trades are breaking even
What to look for:
- A steadily rising curve is ideal (consistent profits)
- Large drops followed by recoveries are drawdowns
- A curve that goes up then crashes down might indicate a strategy that worked for a while then stopped working
Sweep Results
When you run a parameter sweep, results are presented as a ranked table.
How to read it:
- Results are sorted by the metric you chose (Sharpe Ratio by default)
- The top row is the best-performing parameter combination
- Each row shows the parameters used and the key performance metrics
What to look for:
- Is the top result much better than the rest? (might be overfitting)
- Are there clusters of similar good results? (more robust — nearby settings also work)
- Is the top result suspiciously good? (probably overfitting — validate with walk-forward)
Walk-Forward Results
Walk-forward results show two levels of detail:
Aggregate Metrics
These combine all the "exam" periods together. This is the most important number — it tells you how the strategy performed on data it had never seen.
Per-Window Breakdown
Each row shows one learning/exam pair:
| Column | Meaning |
|---|---|
| Window | Which time period (1, 2, 3...) |
| Best Params | Settings that won during the learning period |
| Train P&L | How much the strategy made during learning (expected to be good) |
| Test P&L | How much it made during the exam (the real test) |
| Test Sharpe | Consistency during the exam period |
What to look for:
- Test metrics should be positive (strategy works on unseen data)
- If train metrics are amazing but test metrics are terrible, the strategy is overfit
- Consistent performance across multiple windows is a strong sign
Rules of Thumb
What 'good' looks like
- Sharpe > 1.0 on walk-forward test periods
- Profit Factor > 1.5 consistently
- Max Drawdown < 20% of capital
- Win Rate matters less than Sharpe and Profit Factor
- Consistency across walk-forward windows matters more than any single number
Red flags
- Amazing backtest results that collapse in walk-forward
- Only one or two parameter combinations work (fragile)
- Very high trade count with tiny average profit (vulnerable to fee changes)
- Max drawdown larger than you're comfortable losing in real life