Real search · Redacted public report

Gold (XAU/USD) · H4

Qualified example. 13 local variations across 4 strategy families qualified. This preview shows the research evidence while withholding every executable strategy setting.

96Variations tested
32Nearby refinements
0Replay errors
13Qualifying variations
Leading research area

Donchian breakout · Buy

The leading candidate qualified.

Exact strategy withheldEntry conditions, indicator periods, numerical thresholds, stop and target settings, and every other qualifying variation are available only inside the customer's private Discovery report.
Fixed screening decision

What passed—or failed

Passed every fixed screening requirement.

Historical evidence

Same rules across the complete window

Modelled account returns use USD 10,000 start, 1% risk per trade, compounding. Earlier and screening-holdout periods are part of one continuous replay; full-period return is not their arithmetic sum.

PeriodModelled returnPFTradesMax DD
Full five years+79.14%2.04137-5.5R
Earlier 70%+32.88%1.6793-5.5R
Screening holdout 30%+34.82%3.2144-3.0R

Full-period annualized modelled return: 12.37%. Full-period modelled account drawdown: -5.33%.

Robustness checks

Costs and consistency

+34.86%Screening-holdout return after recorded spread was doubled · PF 3.08 · remained positive
4/4Chronological segments with a positive result
70 / 30Earlier period / screening holdout

The screening holdout is used to qualify and present results; it is not a final untouched confirmation period.

Assumptions and limits

What this does not prove

Managed benchmark results use the recorded historical bid/ask spread with no separate commission. Slippage and overnight financing are not included. The robustness check doubles the recorded spread. Actual broker costs can differ.

These cases were selected after an earlier engineering batch to illustrate three different output types. They are not a random sample and do not estimate how often a search will qualify a strategy.

Historical results do not guarantee future performance. This is research software, not a signal, personalized investment advice, or a finished algorithm.

Reproducible provenance

Audit references

Window: 2021-08-24 to 2026-08-24 UTC · Profile: managed_market_data_v2 version 2026-08-25.1 · Cost policy: managed_spread_only_v1

Run IDreplay_8101b2ef4329f0e8Run fingerprint8101b2ef4329f0e8eb6346e71f0b5b013aaec4c81b974d131e4e254d28127671Scientific result SHA-2563ff1def2de66e66788545a128b46d675e140a8255d60518e607c3f7b5d645eb7

Run your own private search.

Your completed report includes the exact rules and every qualifying variation from your search.

Start Discovery