Skip to main content

Analyzing Results

This page explains how to read BananaEA backtest and optimization results.

No single number tells you whether a setup is suitable. Review profit, drawdown, trade count, costs, and stability together.

First Check: Did The Test Run Correctly?​

Before reading performance metrics, check:

  • The EA version is correct.
  • The symbol and timeframe are correct.
  • The preset or inputs loaded correctly.
  • The test period is correct.
  • Spread and commission assumptions are realistic.
  • The Journal and Experts tabs do not show repeated errors.
  • Trade count is not extremely low.

If the test setup is wrong, the metrics are not useful.

Net Profit​

Net profit is the total profit or loss after closed trades and costs.

It answers:

  • Did this test period end positive or negative?
  • How much did the account change in this historical simulation?

Limitations:

  • It does not show how much risk was taken.
  • It can be dominated by one large trade.
  • It can improve simply because risk was increased.
  • It does not prove future performance.

Always review net profit with drawdown and trade count.

Profit Factor​

Profit factor is:

Gross profit / Gross loss

A value above 1.0 means gross profit was larger than gross loss in the test.

Limitations:

  • A high value can be unreliable if there are few trades.
  • One large winner can inflate it.
  • It can collapse out-of-sample.
  • It does not show whether drawdown is tolerable.

Use profit factor as one comparison metric, not as a pass/fail rule.

Maximum Drawdown​

Maximum drawdown measures the largest decline from a peak to a trough during the test.

It matters because a setup can make money overall while still producing a drawdown the trader cannot tolerate.

Review:

  • Drawdown as a percentage of account balance.
  • Drawdown in money terms.
  • How long drawdown lasted.
  • Whether drawdown increased sharply during one market period.

If the backtest drawdown is near your personal limit, live trading may exceed it.

Win Rate​

Win rate is the percentage of trades that closed profitably.

It is easy to understand but often misleading.

A high win rate can still lose money if losses are much larger than wins. A lower win rate can be acceptable if winning trades are larger than losing trades.

Review win rate together with:

  • Average win
  • Average loss
  • Reward-to-risk
  • Profit factor
  • Consecutive losses

Average Trade And Expected Payoff​

Average trade is:

Net profit / number of trades

Expected payoff is a similar way of thinking about average outcome per trade.

This matters because very small average trade values can be erased by spread, commission, slippage, or execution differences.

If average trade is small, be more cautious with the result.

Trade Count​

Small samples are unreliable.

Use extra caution when:

  • The test has fewer than 30 trades.
  • One or two trades explain most of the result.
  • The setup trades only during one narrow market condition.
  • Different date ranges produce very different results.

More trades do not guarantee reliability, but too few trades makes analysis weak.

Consecutive Losses​

Maximum consecutive losses show the longest losing streak in the test.

Use this to ask:

  • Could I continue following this setup after that many losses?
  • Would my account remain within risk limits?
  • Would a larger live losing streak be tolerable?

Live losing streaks can exceed the historical backtest.

Equity Curve​

A useful equity curve is not only upward. It should also be understandable.

Review:

  • Is growth gradual or driven by one spike?
  • Are drawdowns clustered in one period?
  • Does the curve go flat for long periods?
  • Does the curve become unstable after a market regime change?

Red flags:

  • Perfectly smooth growth.
  • One trade explains most profit.
  • Deep unrecovered drawdown.
  • Sharp collapse after a strong start.
  • Very different behavior across years.

Comparing Optimization Results​

Do not choose the top row automatically.

Review the top group:

  • Are similar values appearing repeatedly?
  • Are profit and drawdown balanced?
  • Is trade count adequate?
  • Are nearby settings also acceptable?
  • Does the selected result survive out-of-sample testing?

A slightly lower-profit setup with more stable behavior can be preferable to the highest historical profit result.

Common Red Flags​

Pause or reject the setup if you see:

  • Very high profit with very few trades.
  • Perfect or near-perfect win rate.
  • Zero or near-zero drawdown.
  • Profit factor that is extreme compared with nearby tests.
  • Out-of-sample result much worse than in-sample.
  • Large sensitivity to tiny setting changes.
  • Unrealistic spread, commission, or slippage assumptions.
  • Repeated trade errors in the Journal.

Practical Decision Framework​

Before using a setup further, confirm:

  • Test ran without setup errors.
  • Trade count is large enough to review.
  • Profit factor is acceptable for the risk taken.
  • Drawdown is within your tolerance.
  • Average trade is not too small for costs.
  • Results are not dominated by one trade.
  • Out-of-sample test is acceptable.
  • Parameter sensitivity is acceptable.
  • Demo forward testing is planned before live use.

What To Save​

Save enough information to reproduce the test:

  • EA version
  • Preset or .set file
  • Symbol and timeframe
  • Broker
  • Date range
  • Spread and commission assumptions
  • Risk settings
  • Main metrics
  • Screenshot or exported report
  • Notes about errors or unusual behavior

Next Steps​