Out-of-sample results

The 2009 and 2025–2026 Balanced runs use matching reported inputs on MultiPillar v3.1.8. The 2010–2024 v3.1.8 benchmark used different risk and execution settings, including position limits and pillar risk scales. The benchmark provides context alongside the two holdouts.

Balanced mode · 12-month averages
Measure2010–2024 benchmarkTwo holdouts average
Starting balance used for comparison£10,000.00£10,000.00
Average 12-month return44.81%48.47%
Average 12-month profit on £10,000£4,481.03£4,846.62
Average ending balance on £10,000£14,481.03£14,846.62
Average monthly return3.11%3.33%
Average annual peak balance drawdown7.93%8.20%
Average annual profit factor1.331.32
Average closed trades per year433.9407.0
Average annual win rate32.24%30.75%

Profit and ending balance apply the mean yearly return to £10,000. Annual balance drawdown uses deal balances and resets the reference peak at the start of each year.

The holdouts averaged 12.20% peak equity drawdown. The benchmark’s full-period figure was 13.59%. Each holdout and each benchmark calendar year receives equal weight within its group.

View each holdout result
Separate 12-month holdout runs
MeasureBefore main testAfter main test
Tester period (end exclusive)1 Jan 2009 – 31 Dec 20091 Mar 2025 – 28 Feb 2026
12-month return12.02%84.91%
Net profit£1,201.95£8,491.30
Ending balance£11,201.95£18,491.30
Average monthly return1.02%5.64%
Peak balance drawdown7.71%8.68%
Peak equity drawdown11.16%13.23%
Profit factor1.111.53
Closed trades376438
Win rate27.93%33.56%

Sources: MT5 reports OHLC19, out of sample pre.xlsx, and OUT OF SAMPLE STANDARD.xlsx. All three use one-minute OHLC modelling.

Download 2009 monthly results ↓ Download 2025–2026 monthly results ↓ MetaTrader’s guide to forward testing ↗