Full-Year Win Rate Moved 6.4 to 36.9 Points Inside 9 EA Backtests; the Weakest Year's Profit Factor Was Below 1 in 4

A backtest report prints one win rate and one profit factor for the whole test. We split the published MT5 backtests of 9 of our EAs into calendar years and looked at how far those two figures moved from one full year to the next inside the same test. This article uses backtests only. It contains no forward-account figures. Figures are as of 2026-10-24.

- Between its weakest and its strongest full year, the win rate of one EA moved by 6.4 points (ATLAS PORTFOLIO, 29.5% to 35.9%) up to 36.9 points (ETHEREUM TREND, 16.7% to 53.6%).

- The profit factor of the weakest full year was below 1 in 4 of the 9 tests: GOLD NEURON 0.20, ETHEREUM TREND 0.43, MEGAMAX DONCHIAN 0.91 and DONCHIAN ENGINE 0.92. All 4 tests still end with a positive total.

- In ETHEREUM TREND and GOLD NEURON the strongest full year had more than 12 times the profit factor of the weakest (5.25 against 0.43, 2.53 against 0.20).

- In 8 of the 9 tests the year with the lowest win rate was also the year with the lowest profit factor. GOLD KING is the exception (2024 and 2025).

- The widest win-rate range comes with the fewest trades per full year (ETHEREUM TREND, 12 to 28). The two narrowest come with 95 to 185 trades (ATLAS PORTFOLIO) and 1,165 to 1,567 (GOLD VIPER).

- Each closed trade is assigned to the calendar year of its close time in the Strategy Tester. Commission and swap are included. Open positions at year end are not valued.

- Only full calendar years are compared. A year in which the test period covers less than 12 months is marked partial and left out; that is 2026 in 8 of the 9 tests. BITCOIN COMET ends on 2025-12-31.

- The win rate and the profit factor of a year are computed from the closed trades of that year. The range in points and the ratio of the two profit factors are our own arithmetic.

- The drawdown in this article is the balance drawdown, measured on closed-trade balance only. The yearly figures are the largest fall from a balance peak to the next low inside that year, with the opening balance of the year as the first peak. The last column is the balance drawdown from the MT5 report for the whole test. We do not put an equity drawdown, which includes open trades, in this article.

- A yearly balance drawdown can be larger than the whole-test figure of the report (MEGAMAX DONCHIAN: 52.2% in 2020 against 15.19%). In a compounding test the largest fall in money comes late, when the balance is large, and the report's percentage is the one for that fall; an earlier, smaller fall in money can be deeper in percent. We keep the two in separate columns.

- TRINITY and AIFW are not in this article: the original report files behind their published backtests are not in our breakdown, so we could not split them by year.

EA | Test period and data | Full years | Lowest full-year win rate (year) | Highest full-year win rate (year) | Range (points) | Lowest full-year profit factor (year, balance drawdown inside that year) | Highest full-year profit factor (year, balance drawdown inside that year) | Trades in a full year, fewest to most | Balance drawdown (MT5 report, whole test)

ETHEREUM TREND | H4, 2019.01.01-2026.06.28, 100% real ticks | 7 | 16.7% (2019) | 53.6% (2025) | 36.9 | 0.43 (2019, 6.3%) | 5.25 (2024, 7.3%) | 12 to 28 | 6.7%

GOLD NEURON | M30, 2004.01.01-2026.09.01, 100% real ticks | 22 | 38.5% (2004) | 65.7% (2013) | 27.2 | 0.20 (2004, 24.4%) | 2.53 (2013, 9.2%) | 39 to 229 | 22.84%

DONCHIAN ENGINE (free) | H1, 2017.01.01-2026.06.01, 100% real ticks | 9 | 23.3% (2020) | 48.3% (2021) | 25.0 | 0.92 (2020, 9.2%) | 3.19 (2021, 2.6%) | 56 to 76 | 7.31%

MEGAMAX DONCHIAN (USDJPY) | H1, 2017.01.01-2026.06.01, 100% | 9 | 25.0% (2020) | 48.3% (2021) | 23.3 | 0.91 (2020, 52.2%) | 3.33 (2021, 7.8%) | 56 to 76 | 15.19%

GOLD TRIPWIRE | M15, 2020.01.01-2026.10.14, 11% real ticks | 6 | 50.5% (2023) | 70.1% (2025) | 19.6 | 1.49 (2023, 6.3%) | 4.42 (2025, 1.8%) | 82 to 109 | 6.28%

BITCOIN COMET | H4, 2018.01.01-2025.12.31, 98% | 8 | 32.0% (2022) | 45.1% (2018) | 13.1 | 1.36 (2022, 14.5%) | 2.65 (2018, 16.6%) | 51 to 115 | 14.06%

GOLD KING | M1, 2019.01.01-2026.06.01, 5% real ticks | 7 | 79.7% (2024) | 91.0% (2019) | 11.3 | 4.8 (2025, 7.5%) | 13.56 (2021, 1.0%) | 215 to 655 | 7.45%

GOLD VIPER | M30, 2019.01.01-2026.09.01, 8% real ticks | 7 | 52.5% (2022) | 59.4% (2025) | 6.9 | 1.18 (2022, 15.1%) | 2.48 (2020, 10.3%) | 1,165 to 1,567 | 19.6%

ATLAS PORTFOLIO | H4, 2019.01.01-2026.06.28, 100% | 7 | 29.5% (2020) | 35.9% (2022) | 6.4 | 1.16 (2020, 11.6%) | 1.84 (2022, 19.2%) | 95 to 185 | 16.29%

Sorted by the win-rate range, largest first. The yearly figures and the whole-test balance drawdown come from the same MT5 Strategy Tester report. The tests cover different periods and data quality, so compare the numbers inside one row, not the rows with each other.

ETHEREUM TREND had a full year with a 16.7% win rate and a profit factor of 0.43 (2019) and a full year with a profit factor of 5.25 (2024). It makes 12 to 28 trades in a full year, so a handful of trades can move a year's win rate by many points. GOLD NEURON has the longest test in the table, 22 full years, and its weakest year (2004, profit factor 0.20, balance drawdown 24.4% inside the year) is far from its strongest (2013, 2.53).

For both, the single profit factor in the report is an average over years that look nothing like it.

MEGAMAX DONCHIAN and DONCHIAN ENGINE trade USDJPY on H1 over the same test period. Their weakest and strongest full years fall on the same years, 2020 and 2021, their win rates in those years are within 1.7 points of each other, and both make 56 to 76 trades in a full year. The difference sits in the drawdown: in 2020 the balance drawdown inside the year was 52.2% for MEGAMAX DONCHIAN and 9.2% for DONCHIAN ENGINE.

ATLAS PORTFOLIO stayed between 29.5% and 35.9% in every full year, GOLD VIPER between 52.5% and 59.4%. A narrow win-rate range did not mean a narrow drawdown: the balance drawdown inside a single year reached 19.2% (ATLAS PORTFOLIO, 2022) and 19.6% (GOLD VIPER, 2023).

- One weak year is not by itself a sign that an EA has stopped working. 4 of these 9 tests contain a full year with a profit factor below 1 and still end positive. Compare a weak year with the weakest full year of the EA's own backtest.

- The fewer trades an EA makes in a year, the more its yearly win rate can move. Judge a low-frequency EA over several years, not one.

- Read each year's profit factor next to that year's drawdown. A weak year came with 6.3% in one test (ETHEREUM TREND, 0.43) and 52.2% in another (MEGAMAX DONCHIAN, 0.91).

- A backtest value is a measurement of the past on one history, not a limit. A future year can be weaker than any year in the test. Nothing here is a promise of future results.