What is a good profit factor? How to read this metric in an EA backtest
Profit factor is usually displayed prominently in an MT5 report, but a high number does not by itself prove that an EA has a durable edge. Read it together with gross profit, gross loss, number of trades, test period and drawdown.
What counts as a good profit factor depends on how the number was produced. The metric compares the total money from winning trades with the total money from losing trades over the exact period tested. It tells you nothing about individual trades, trade order, recovery time or the largest single loss. Treat it as a summary that prompts the next question, not as a score for ranking EAs.
What is profit factor and how is it calculated?
The usual formula in reports is gross profit divided by the absolute value of gross loss. Gross profit is the sum of the positive results of winning trades; gross loss is the sum of the magnitudes of the negative results of losing trades. The ratio shows how many units of gross profit came with each unit of loss in the sample. Both fields must come from the same report and the same period. It is a historical ratio, not a forecast for future trades.
| Profit factor | Cautious interpretation |
|---|---|
| Below 1 | Gross loss exceeds gross profit in the test sample. |
| Around 1 | The margin is thin; real costs can flip the result. |
| 1.5–2 | May be worth a closer look if the data and sample are broad enough. |
| Above 3 | Not automatically good; check trade count, test period and risk. |
| Very high on a short sample | Easily driven by a few trades or by fitting to the past. |
Example from a published EA report
M1 REGIME backtest report, 1 Jan – 2 Oct 2026, initial deposit 300 USD: 268 trades, net profit 324 USD, profit factor 1.68 and maximum balance drawdown 14.53%. These are published Strategy Tester figures, not live account results. The sample is still short and does not guarantee future results. See the report and test conditions on the EA page.
Why does the number of trades change the reading?
Profit factor is calculated from all trades, so one or two large trades can have a strong effect when the sample is small. For instance, a system with a few very large winners and many small losers can still show a high ratio, but you need to know whether those large trades were rare events or depended on one favorable stretch. Conversely, a short run of small trades can also leave the result unstable. Trade count does not make a result certain; it only helps you judge how exposed the result is to chance.
The 200-trade mark is sometimes used as a signal for reviewing the sample, not as a statistical standard that guarantees anything. Two hundred trades over a few months in one market regime differ from 200 trades spread over many years and many conditions. Check the start date, end date, symbol, timeframe and data quality as well. A parameter set chosen after many attempts also risks over-fitting, even if the final report shows hundreds of trades.
Is a profit factor above 3 suspicious?
You cannot conclude anything from the threshold of 3 alone. However, a very high value on few trades is a reason to examine how the result was produced. Compare gross profit and gross loss, the number of winning and losing trades, the largest loss, equity drawdown, and the spread and commission settings. If there are a few outlier wins, ask what the result looks like when each of them is set aside in turn; this is sensitivity analysis, not editing the report figures.
Also ask whether many configurations were tested and only the best-looking one was kept. The more parameters you try on the same history, the higher the chance of finding a result that looks excellent by luck. Testing on data not used for optimization, then following a demo on new data, adds perspective; neither guarantees profit.
Trading costs and companion metrics
Profit factor only means something if costs are simulated realistically. Spread, commission, swap and slippage reduce the net result. How much depends on trading frequency, position size and asset type. For a strategy with a thin edge, ignored costs can make the real ratio far lower than the report shows. Check the test configuration and compare several assumed cost levels if the tool allows it.
Put profit factor next to drawdown, net profit, expected payoff and number of trades. Drawdown shows the historical decline; net profit shows the end-of-period difference under specific conditions; expected payoff describes the average per trade. No single metric replaces reading the equity curve, the trades and the test period. If the report is missing important information, treat any conclusion as incomplete.
A backtest simulates results on past data under certain assumptions. It is not a live trading result, it does not confirm that an EA will be profitable and it does not guarantee future results.
To see other metrics and EA profiles with published data, continue with the links below.
Frequently asked questions
There is no guaranteed threshold. Above 1 means gross profit exceeded gross loss in the sample; around 1.5–2 may be worth a closer look when the data, trade count and costs are clearly presented.
It may be a notable result, but you need to check trade count, time span, drawdown, costs and parameter optimization. On a short sample, a few trades can push the ratio up.
No. It describes the gross profit and loss of one specific simulation. New market conditions, spread, commission and slippage can differ.
Divide gross profit by the absolute value of gross loss, using both fields from the same report and the same period.
It depends on the settings and the report. Check how spread, commission, swap and slippage were simulated before comparing.
DinexSpace labels the source of every number in its research profiles: backtest, forward, demo or live. You can check them against the points covered in this article.