Reading backtest results
How to interpret the equity curve, key metrics, and individual trade list.
A backtest result includes an equity curve, seven key metrics, and a full trade list — together they reveal whether your strategy has a consistent statistical edge. This article explains how to read each section and which patterns indicate a healthy or overfit strategy.
Equity curve
The equity curve shows portfolio value over time, starting from 100% (your initial capital). Rises = winning trades; drops = losing trades or drawdown.
What to look for:
- Smooth upward slope — good. Consistent edge across many conditions.
- Long flat periods — the strategy has no signal and sits idle. Normal for trend-following.
- Sharp drawdowns — sudden drops. Check if they correlate with known events (flash crashes, etc.) or if the strategy needs a better stop-loss.
Key metrics
| Metric | What it means |
|---|---|
| Total return | % gain or loss over the tested period |
| Max drawdown | Largest peak-to-trough decline. Measures worst-case pain. |
| Win rate | % of trades that closed profitably |
| Average win / loss | Average profit per winning trade vs loss per losing trade |
| Profit factor | Total wins ÷ total losses. >1.5 is generally considered decent. |
| Trade count | Total number of round-trips (entry + exit). Low count → less statistical significance. |
| Sharpe ratio | Risk-adjusted return: (return − risk-free rate) ÷ volatility. >1 is reasonable. |
The trade list
Expand any trade to see:
- Entry timestamp and price
- Exit timestamp, price, and reason (TP, SL, indicator, etc.)
- P&L in USD and %
- Indicator values at entry and exit
This is the most useful debugging tool. If you see a lot of losses, open a few and check if the exit fired too early or too late.
Red flags
- Win rate > 90% with many trades: probably overfit to the backtest period. Unlikely to hold forward.
- Very few trades (< 20): not enough data to conclude anything. Extend the range.
- Suspiciously consistent returns: the strategy may be fitting to a trend that's specific to the tested period (e.g. a 1-year bull market).
- Massive drawdown: the strategy would be psychologically impossible to live-trade if a real-money version dropped 60%.
Exporting results
Click Export CSV to download the full trade list. Useful for analysis in Excel or a Jupyter notebook.