H1 vs H2 in 82 Full Years of 9 EA Backtests: One Half Lost, the Other Gained in 22; GOLD NEURON H1 Median −2.5%

Every calendar year has two halves: January to June (H1) and July to December (H2). We split each full year inside the published MT5 backtests of 9 of our EAs into these two halves and measured the return of each half. We wanted to see how often the two halves of the same year tell different stories. This article uses backtests only. It contains no forward-account figures. Figures are as of 2026-11-08.

- Across 82 full years in 9 EA backtests, one half lost money while the other half made money in 22 years. In 14 of those 22 the full year still closed positive, so the yearly figure hid a losing half.

- H1 returned more than H2 in 47 of the 82 years, H2 in 35. MEGAMAX DONCHIAN and the free DONCHIAN ENGINE share one engine and the same 7 H1 years; without the free engine the split is 40 to 33 of 73. We do not read this as a seasonal rule.

- GOLD NEURON has the largest gap between the halves: its median H1 return is −2.5% and its median H2 return is +16.9% over 22 full years. 20 of its 44 halves lost money.

- GOLD KING and GOLD TRIPWIRE have no losing half at all (0 of 14 and 0 of 12). The same GOLD KING test has a maximum relative equity drawdown of 91.3%, so a clean half-year record says nothing about the drawdown inside it.

- Each closed trade is assigned to the month in which it closed. Commission and swap are included. Open positions are not valued.

- Only full calendar years are used. 2026 is not a full year in 8 of the tests and is left out. BITCOIN COMET's test ends on 2025-12-31.

- H1 return = net result of January to June ÷ the balance at the start of the year. H2 return = net result of July to December ÷ (start balance + H1 net result). Several tests compound, so a half is compared by percent of the balance it started with, not by money.

- "Higher half" counts the years in which H1 or H2 had the larger return. There were no ties.

- The last column is the maximum relative equity drawdown of the whole test (site settings): the largest fall as a percent of the equity peak. It is not a half-year figure and is shown for scale.

- 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 month.

EA | Test period and data | Full years | Higher half: H1 / H2 (years) | Median return: H1 / H2 | Years one half lost and the other gained | Weakest half (year, return) | Max relative equity DD (whole test)

GOLD NEURON | M30, 2004.01.01-2026.09.01, 100% real ticks | 22 (2004-2025) | 10 / 12 | −2.5% / +16.9% | 6 | 2007 H2, −25.4% | 55.5%

DONCHIAN ENGINE (free) | H1, 2017.01.01-2026.06.01, 100% real ticks | 9 (2017-2025) | 7 / 2 | +5.5% / +4.1% | 5 | 2020 H2, −7.0% | 10.8%

MEGAMAX DONCHIAN (USDJPY) | H1, 2017.01.01-2026.06.01, 100% | 9 (2017-2025) | 7 / 2 | +19.4% / +7.4% | 4 | 2020 H2, −41.3% | 58.6%

ATLAS PORTFOLIO | H4, 2019.01.01-2026.06.28, 100% | 7 (2019-2025) | 5 / 2 | +51.6% / +24.9% | 3 | 2025 H1, −3.8% | 26.3%

ETHEREUM TREND | H4, 2019.01.01-2026.06.28, 100% real ticks | 7 (2019-2025) | 3 / 4 | +3.2% / +3.6% | 2 | 2021 H2, −8.1% | 17.1%

BITCOIN COMET | H4, 2018.01.01-2025.12.31, 98% | 8 (2018-2025) | 5 / 3 | +16.6% / +15.4% | 1 | 2018 H1, −6.2% | 31.2%

GOLD VIPER | M30, 2019.01.01-2026.09.01, 8% real ticks | 7 (2019-2025) | 3 / 4 | +45.2% / +42.7% | 1 | 2022 H1, −4.9% | 30.3%

GOLD KING | M1, 2019.01.01-2026.06.01, 5% real ticks | 7 (2019-2025) | 4 / 3 | +34.2% / +23.8% | 0 | 2019 H1, +14.0% | 91.3%

GOLD TRIPWIRE | M15, 2020.01.01-2026.10.14, 11% real ticks | 6 (2020-2025) | 3 / 3 | +17.9% / +20.0% | 0 | 2020 H1, +2.8% | 12.7%

- ATLAS PORTFOLIO, 2025: the full year closed at +25.7%, but H1 was −3.8% and the whole gain came in H2 (+30.7%). Someone who started the EA in January would have been at −3.8% on closed trades at the end of June, in a year that ended well.

- MEGAMAX DONCHIAN, 2020: H1 made +38.6%, then H2 lost −41.3%, and the year closed at −18.6%. A strong first half was followed by the weakest half in its test.

- GOLD NEURON: 7 of its 22 full years lost in both halves (2004, 2005, 2007, 2009, 2015, 2017, 2018). In 2023 H1 was −15.2% and H2 +39.0%; the year closed at +17.9%.

- BITCOIN COMET, 2018: H1 −6.2%, H2 +125.1%, full year +111.3%. GOLD VIPER, 2022: H1 −4.9%, H2 +28.9%, full year +22.6%. In both cases one losing half sits inside a strongly positive year.

- Years with both halves negative are rare outside GOLD NEURON: ETHEREUM TREND 2019 (−4.1% and −0.7%) is the only other one.

- A yearly return table can show a positive year whose first or second six months lost money. Ask for monthly or half-year figures, and for the maximum drawdown of the same test.

- Do not judge an EA after its first six months. In 22 of 82 years in these tests, the two halves of the same year had opposite signs.

- Do not pick a start month from these counts. 47 against 35 over 82 years is a small difference, and two of the EAs share one engine.

- This article is about past tests. It does not promise any future result.

- All forward accounts we publish are demo accounts, and none is used in this article.

The two EAs discussed most in this article: GOLD NEURON https://www.mql5.com/en/market/product/187329 · MEGAMAX DONCHIAN https://www.mql5.com/en/market/product/181838