Research approach
Evidence before conviction.
What a worker is
A worker is a versioned trading strategy with its own entry conditions, timeframe requirements and position-management rules. Its name is not a recommendation. A strategy that performs in one window may fail in another market or regime.
How we assess a strategy
Research starts with whether the inputs and execution rules are valid. Historical evaluation then measures behaviour under those rules. Shadow trading provides a separate observation of the strategy using incoming market data and simulated fills.
We examine returns alongside trade count, drawdown, costs, holding time and market concentration. Win rate alone cannot establish an edge: small wins and large losses can produce a high win rate and a losing strategy.
What the published numbers mean
The public results section currently reports the shadow-worker experiment. Its month-to-date closed-position results include all workers that closed a position in that window, including losses. Open-position unrealised P&L is excluded from that table.
Shadow accounts are independent simulations. Their combined P&L is not the return of one live investment account. A worker can be observed in shadow trading without being approved for real-money execution.
Changes remain part of the record
A correction, new strategy version or policy change can affect comparability. Older demo records remain separate from the active shadow experiment. We do not treat a software check passing as evidence of profitable trading.
Past performance, whether simulated or live, does not establish future performance. See the risk disclosure for the limitations of these results.