enabled native sleeves
Darwinex Medium V3
MSRLab investigates trading hypotheses, builds and challenges systematic strategies, combines complementary edges into portfolios, and uses live behavior as feedback for the next research cycle.
THE RESEARCH FOOTPRINT
Ideas enter from observation, code, documented strategies, external practitioners and live portfolios. They leave as rejected evidence, watchlist knowledge, or governed sleeves.
Darwinex Medium V3
2024–2026 YTD published splits
intraday timing to daily signals
idea to monitored allocation
Clear states matter more than inflated totals. These are the counts the current public repository can support.
Traceable 2024–2026 YTD published validation splits.
Current Darwinex Medium V3 portfolio footprint.
Darwinex Zero and the Deriv small-capital live test.
One watchlist and one paused program; neither is represented as deployment-ready.
Public snapshot · 19 Aug 2026. A complete lifetime count of every rejected idea and EA will be added after the historical experiment register is reconciled; until then, MSRLab will not invent that number.
Independent reconstruction, realistic costs, rolling evidence and robustness checks.

Human-directed · AI-accelerated · Evidence-led
Deployment is where expectation meets observation. What we learn returns to the front of the process.
The record includes work that survives, work that fails, and work that changes the next question.
FEATURED PROJECT
Can useful trading hypotheses be systematically extracted from the MQL5 ecosystem and survive independent MSRLab validation?
An open-code, one-chart Darwinex portfolio combining distinct strategy families under shared sleeve, symbol, and account risk caps.
MSR-003liveA broker-specific portfolio lane spanning synthetics and selected cross-broker research, with runtime exclusions made visible instead of silently ignored.
MSR-004watchlistA long-running investigation with substantial negative evidence. Simple filtered variants have not preserved the original live-style basket behavior.
Backtest, forward, and live records are different kinds of evidence. The site will never blur them.
MSRLab-curated, multi-strategy portfolio
Selected native and transferable sleeves
The book is diversified by hypothesis, market, horizon and failure mode—not simply by EA count.
Morning rangeLondon rangeDonchiancompression
Intraday · session-based · swingOxford Turtle Soupclose-extreme fadePBZ raw-fade*
Swing · positionS/R breakoutretestsupport bounceMainTL geometry*
Intraday · swingMainTL distance*WPR extension*large-candle follow
Intraday · swingSynthetic geometry*Step-family*crypto Turtle*
Intraday · swing · broker-native* MSRLab-native method, implementation, or research family. Source-inspired work is credited separately and is not presented as wholly native.
Explore the research map ↗External inspiration is credited. Native IP, curated commercial systems and unresolved ideas remain visibly separate.
Market observations, indicator reconstruction, portfolio mining and live feedback.
Structured strategy-development sessions translated into testable MSRLab research questions.
Range, trend and recovery ideas rebuilt as transparent hardened systems.
Concept pipeline that produced the FX D1 Turtle Soup sleeve.
Commercial EAs, set ecosystems and community updates evaluated separately from native IP.
Strategy concepts ingested, implemented and required to earn promotion independently.
Systematic-trading material used as hypothesis input, never as proof of an edge.
Trading concepts formalized into rules and subjected to independent testing.
Range-breakout concepts compared with existing MSRLab and Rene-style families.
A staged concept-mining funnel—not copied claims or automatic deployment evidence.
Ongoing independent validation of publicly shared strategy ideas and their robustness claims.
Darwinex, Deriv and MT5 behaviour fed back into attribution and governance.
Deployment begins a feedback loop. The objective is to detect drift, recommend action and keep replenishing the portfolio before yesterday’s edge becomes tomorrow’s unmanaged risk.
Scheduled cron jobs and on-demand pulls collect MT5 history, account state and strategy attribution.
Live frequency, execution, drawdown and sleeve contribution are compared with historical expectation.
The monitoring layer produces evidence for keep, reduce, pause, retire or investigate recommendations.
Edges are not assumed to last forever. Degradation triggers review while the lab continues developing independent replacements.
Short notes, deep dives, engineering decisions, portfolio observations, and postmortems.
FOUNDER / RESEARCH DIRECTOR