All Posts (Page 32)

The KS statistic, derived from the Kolmogorov-Smirnov test, measures the degree to which a trading model's observed return distribution matches a reference distribution. In darwintIQ, it is one of several distribution-comparison metrics used to detect whether a model's live behaviour has shifted away from what was seen during evaluation.

Exposure measures the proportion of time a trading model holds open positions, reflecting its selectivity and the risk profile that comes with constant versus selective market presence. High exposure absorbs more overnight risk and market noise; low exposure trades less but may generate cleaner signals. In darwintIQ, exposure sits in the performance dashboard where it helps characterise a model's trading personality alongside return consistency and drawdown.

Markets do not always fit neatly into trending or ranging categories. Mixed and unstable regimes describe the transitional and erratic conditions where the structural character of price action is unclear. These are the environments where the majority of trading models lose the edge they had in cleaner conditions. In darwintIQ, regime classification distinguishes between Trend Dominant, Range Dominant, Mixed, and Unstable states, giving a realistic view of when a model is and is not operating in its intended environment.

Population Stability Index (PSI) measures how much a trading model's return distribution has shifted from its validated baseline, using a bin-by-bin comparison that detects gradual drift across the full range of outcomes. Borrowed from credit risk modelling, it provides interpretable thresholds and complements the KS statistic, Wasserstein Distance, and Jensen-Shannon Divergence in darwintIQ's distribution similarity assessment.

Mutual information is an information-theoretic metric that measures how much the return distribution of one period informs another. Unlike correlation, it captures non-linear statistical dependencies — making it a more sensitive tool for detecting whether a trading model's behaviour has remained structurally consistent. darwintIQ uses mutual information as part of a suite of distribution similarity metrics evaluated on a rolling basis.