COMPLETED RESEARCH · 28 AUG 2026
Mining MQL5
Can useful trading hypotheses be systematically extracted from the MQL5 ecosystem and survive independent MSRLab validation?
CompletedThe program did not produce a new deployable strategy. It did produce a permanent negative-evidence record, a reusable intelligence and broker-validation system, and a much sharper understanding of why public trading information rarely exposes the complete edge.
The completed funnel
Counts describe different research stages and are not added together as independent strategies.
What was accomplished
- Public ideas were separated from utilities, duplicate families, and hidden-tail designs.
- Downloadable public source was audited, compiled in isolation, and challenged through frozen broker campaigns.
- Commercial popularity and attributable public records were tested as research signals rather than accepted as proof.
- Reusable provenance, safety, chronology, cost, and stopping infrastructure was retained.
Why no strategy was promoted
- Many public descriptions exposed only the outer shell of the decision process.
- Apparent edges failed chronology, realistic execution, stability, or cross-market replication.
- Some designs contained unacceptable recovery, concentration, or operational tails.
- Interesting information did not consistently survive as an executable portfolio edge.
A completed negative result is research capital.
The broad search is closed so future work does not repeatedly mine the same evidence under new names. A bounded question can reopen only when materially new evidence exists.
Read the full public report
The detailed report documents all five phases, their audited scale, the conclusions, and the conditions that could justify reopening the program—without publishing code, configurations, or portfolio IP.