Research

Every validated strategy has a dossier: the method — in-sample selection, then a held-out out-of-sample test — the cost model, the reproducible numbers, and, the part no hype vendor publishes, the risks and limits we name ourselves.

ORB Breakout on MNQ: how it was validated

An out-of-sample test of an opening-range breakout on MNQ. The parameters were optimized on an in-sample block, then frozen and run once on data the search never saw — where the edge held at about 90% of its in-sample strength and cleared statistical significance on its own. The overfitting risk that optimizing creates is named plainly, not buried.

Research

ORB Breakout on MNQ: how it was validated