Results

The complete record, including the bad weeks.

Every figure on this page comes from real out-of-sample studies. Nothing is selected, smoothed or rounded up, and the assumptions behind them are stated below so you can discount them yourself.

Gold

Longest record

2019-05-14 to 2026-09-08 / 192 test periods / fixed 0.1 lot

2.86Step t-statisticdistinguishable from luck

802,705 ptsNet pointsfixed 0.1 lot, no compounding

-33,428 ptsMax drawdownpeak to trough, in points

Dashed line = modelled, not real money. Cumulative points captured across 192 consecutive out-of-sample test periods at a fixed 0.1 lot, with no money management and no compounding. Spread and commission included. Hypothetical results are not a reliable indicator of future performance.

802,705Net pointsat a fixed 0.1 lot

4,181Mean test periodaverage points per period

2.86Step t-statisticacross 192 test periods

-33,428Max drawdownpoints, peak to trough

4.80Profit factorgross win / gross loss

62%Profitable test periodsof 192 periods

196,080Best test periodsingle best, points

-14,150Worst test periodsingle worst, points

Why it is a straight line, not a rocket

Traded at the same fixed 0.1 lot, with no compounding, over 7.3 years. Points measure the raw price movement captured, independent of account size or leverage. Compounding is what makes a curve go vertical - and it tells you nothing about whether the edge is real.

Japan 225 index

2023-12-15 to 2026-06-19 / 133 test periods / fixed 0.1 lot

4.68Step t-statisticdistinguishable from luck

1,645,589 ptsNet pointsfixed 0.1 lot, no compounding

-59,766 ptsMax drawdownpeak to trough, in points

Dashed line = modelled, not real money. Cumulative points captured across 133 consecutive out-of-sample test periods at a fixed 0.1 lot, with no money management and no compounding. Spread and commission included. Hypothetical results are not a reliable indicator of future performance.

1,645,589Net pointsat a fixed 0.1 lot

12,373Mean test periodaverage points per period

4.68Step t-statisticacross 133 test periods

-59,766Max drawdownpoints, peak to trough

3.60Profit factorgross win / gross loss

64.7%Profitable test periodsof 133 periods

152,571Best test periodsingle best, points

-43,710Worst test periodsingle worst, points

Why it is a straight line, not a rocket

Traded at the same fixed 0.1 lot, with no compounding, over 2.5 years. Points measure the raw price movement captured, independent of account size or leverage. Compounding is what makes a curve go vertical - and it tells you nothing about whether the edge is real.

Germany 40 index

2024-02-23 to 2026-06-19 / 123 test periods / fixed 0.1 lot

2.62Step t-statisticdistinguishable from luck

248,192 ptsNet pointsfixed 0.1 lot, no compounding

-42,117 ptsMax drawdownpeak to trough, in points

Dashed line = modelled, not real money. Cumulative points captured across 123 consecutive out-of-sample test periods at a fixed 0.1 lot, with no money management and no compounding. Spread and commission included. Hypothetical results are not a reliable indicator of future performance.

248,192Net pointsat a fixed 0.1 lot

2,018Mean test periodaverage points per period

2.62Step t-statisticacross 123 test periods

-42,117Max drawdownpoints, peak to trough

1.91Profit factorgross win / gross loss

61%Profitable test periodsof 123 periods

36,829Best test periodsingle best, points

-20,566Worst test periodsingle worst, points

Why it is a straight line, not a rocket

Traded at the same fixed 0.1 lot, with no compounding, over 2.3 years. Points measure the raw price movement captured, independent of account size or leverage. Compounding is what makes a curve go vertical - and it tells you nothing about whether the edge is real.

US Tech 100

2024-02-23 to 2026-06-19 / 123 test periods / fixed 0.1 lot

2.30Step t-statisticdistinguishable from luck

263,666 ptsNet pointsfixed 0.1 lot, no compounding

-66,007 ptsMax drawdownpeak to trough, in points

263,666Net pointsat a fixed 0.1 lot

2,144Mean test periodaverage points per period

2.30Step t-statisticacross 123 test periods

-66,007Max drawdownpoints, peak to trough

1.80Profit factorgross win / gross loss

61.8%Profitable test periodsof 123 periods

36,514Best test periodsingle best, points

-52,566Worst test periodsingle worst, points

Why it is a straight line, not a rocket

Exactly how each study is tested

Settings are selected using only price history up to a cut-off date, then scored on the following period, which the selection had never seen. The window then moves forward and the process repeats.

Consecutive test periods, no gapsMetaTrader 5High-quality dataSpread + commission includedFixed 0.1 lotMeasured in pointsNo compounding

What this does not tell you

HypotheticalNo live money. Nothing sent to a broker.

Few instrumentsEvidence, not proof. One period each.

Costs includedBut live fills can still differ.

Not live yetNo latency, gaps or real-money pressure.

Currently under test

Each instrument below gets exactly the same study as the ones above, in the same format. Nothing is published for them until their record is complete - a partial result released while it happens to look good is precisely the practice this page exists to avoid.

US30In preparation

Wall Street 30

--Net result

--t-statistic

--Max drawdown

--Profitable periods

Study in progress / results pending

US500In preparation

US 500

--Net result

--t-statistic

--Max drawdown

--Profitable periods

Study in progress / results pending