All Posts (Page 22)

Survivorship bias distorts how traders evaluate strategies by making only successful models visible. When you study strategies that worked, you're looking at a filtered subset — the failures have been quietly removed from the sample. This article explains why survivorship bias matters in trading model evaluation, how it leads to overconfidence, and how darwintIQ's evolutionary evaluation framework naturally counters its worst effects.

Volatility in forex and other liquid markets isn't distributed evenly across a 24-hour day. It follows the rhythm of global trading sessions — London, New York, Tokyo — and the overlaps between them. This article explains how session-based volatility affects the regimes trading models operate in, and why darwintIQ's continuous evaluation framework accounts for this naturally.

SP500 currently reads bullish, with regime and model leadership broadly aligned.

DAX currently reads constructive bullish, with regime and model leadership broadly aligned. Main risk: Expanding volatility can reduce robustness quickly.

Wasserstein distance — sometimes called the Earth Mover's Distance — measures how much one probability distribution would need to change to match another. In trading model evaluation, it detects when a model's return distribution has drifted from expectations. This article explains what it measures, how to interpret it, and how it fits into darwintIQ's broader model evaluation framework.