Even with PF1.32, the real source of profit was only ten trades—how to read a "winning EA" uncovered by breaking down 2,191 trades
Choosing a high-win-rate EA will lead to winning.
If PF exceeds 1, you can feel secure.
If a backtest is on an upward trend, operating it similarly will leave profits behind.
I believed for a long time that such numbers reflected the true ability of an EA.
However, after dissecting 2,191 trades over eight and a half years one by one, my perspective changed.
From the top 10 profit trades, only 0.46% of all trades accounted for about 57% of total profits.
In this article, using a single EA as the subject, I will read the "profit structure" that is not visible from win rate or PF alone.
This is not ACE's development story.
For those already using EAs or considering a purchase, this article helps you decide whether to stop your EA midway, continue, and what to verify.
*The figures shown are results from historical data verifications and do not guarantee future profits.
The performance table shows "averages"
Backtests display net profit, win rate, PF, maximum drawdown, and more.
These are important. However, each is the result of aggregating many trades into one figure.
Even with the same net profit of 2 million yen,
- an EA that gradually compounds each month
- an EA created by a few big wins and many small losses
the scenery visible during operation will be completely different.
If you start thinking of the latter as an EA that "steadily increases every month," you may not endure a long stagnation and may stop just before real profits emerge.
From here on, using the actual 2,191 trades, I will dissect the EA from three angles: profit concentration, exit, and trend direction.