Twenty years of market data, one backtest

We completed the first full historical replay across two decades of market data. The purpose was not to produce a headline return, but to test whether the reasoning layer behaved consistently across different market regimes.

The output is a research baseline the team can inspect, challenge and improve. Backtests are hypothetical and do not represent live client results.