Backtest Results
This guide explains the performance metrics from your backtest and how the tier system works. After a successful backtest, you’ll see key metrics on your agent card in the Console and detailed results on the Agent Detail page.
Performance Metrics
After a successful backtest, you’ll see these metrics:
| Metric | Description |
|---|---|
| Portfolio Return | Total return shown as a dollar amount and percentage on $10,000 starting capital |
| Sharpe Ratio | Risk-adjusted return measure. Higher is better. |
| Max Drawdown | Largest peak-to-trough decline during the backtest |
| Win Rate | Percentage of trades that were profitable |
| Total Trades | Total number of completed trades |
| Profit Factor | Gross profit divided by gross loss |
Understanding Each Metric
Portfolio Return %
Your total return as a percentage. A 15% return means your $10,000 portfolio grew to $11,500.
Sharpe Ratio
Measures return relative to risk. The formula is: (Return - Risk-Free Rate) / Standard Deviation.
| Sharpe Ratio | Interpretation |
|---|---|
| < 0 | Losing money on average |
| 0 - 0.5 | Poor risk-adjusted returns |
| 0.5 - 1.0 | Acceptable |
| 1.0 - 2.0 | Good |
| 2.0+ | Excellent |
Max Drawdown %
The worst peak-to-trough drop. If your portfolio went from $12,000 to $9,000, that’s a 25% drawdown.
Lower is better. Professional funds typically target < 20% max drawdown.
Win Rate %
Percentage of profitable trades. Note that win rate alone doesn’t determine profitability—a 40% win rate can be profitable if winners are much larger than losers.
Total Trades
Total trades executed. More trades generally means more statistical significance in your results.
Profit Factor
Gross profit / gross loss. A profit factor of 2.0 means you made $2 for every $1 lost.
| Profit Factor | Interpretation |
|---|---|
| < 1.0 | Losing money |
| 1.0 - 1.5 | Marginally profitable |
| 1.5 - 2.0 | Good |
| 2.0+ | Excellent |
Tier System
Agents earn tier badges based on their portfolio return percentage. Tier badges appear on your agent card in the Console, on the Agent Detail page, and on the leaderboard.
| Tier | Return % | Badge Color |
|---|---|---|
| Unranked | < 7.5% | Grey |
| Bronze | 7.5% - 20% | Bronze |
| Silver | 20% - 32.5% | Silver |
| Gold | 32.5%+ | Gold |
We use portfolio return for tiers because it’s the most intuitive metric for comparing agents. A Bronze agent with 15% returns outperformed the market average. Sharpe ratio and other risk-adjusted metrics are shown separately so you can make your own comparisons.
Portfolio Performance Chart
The Agent Detail page includes a Portfolio Performance chart that visualizes your backtest results. The chart shows your portfolio’s ending value compared to the starting capital of $10,000, giving you an at-a-glance view of whether your agent made or lost money during the backtest period.
Interpreting Your Results
A Sharpe Ratio above 1.0 is generally considered good, and above 2.0 is excellent. Many quantitative funds won’t consider strategies with a Sharpe below 1.0 after transaction costs. Be cautious of extremely high Sharpe Ratios in backtests as they may indicate overfitting.
For Profit Factor, values between 1.5 and 2.0 indicate a healthy balance between risk and return, while above 2.0 is excellent. Below 1.5 is considered weak and vulnerable to slippage and transaction costs. If your profit factor exceeds 4.0, it may suggest your strategy is overfit to historical data.
Win Rate can be misleading on its own. A 40% win rate can be highly profitable if winning trades are significantly larger than losing trades, while an 80% win rate can lose money if the losses are large. Always evaluate win rate alongside profit factor and risk-reward ratio.
Trade count matters because an agent with only 5 trades has less statistical significance than one with 500 trades. Be aware of overfitting risk as well—an agent that performs amazingly on historical data may fail on live data. Simple agents often generalize better than complex ones.
Improving Your Agent
If your backtest results are disappointing, start by checking for bugs and making sure your logic is correct. Consider simplifying your approach by starting with a basic agent and adding complexity gradually. Try adjusting parameters like indicator periods or thresholds, and add risk management techniques like stop-losses to limit drawdowns. Each upload creates a new agent, so you can iterate quickly.
Next Steps
Once you have a successful backtest:
- Deploy to paper trading — Test with live market data
- Review the writing guide — Learn more advanced techniques