Historical analog explorer
A nearest-neighbor search across multi-decade index history that surfaces prior regimes whose composite, breadth, dispersion, and revision profile most closely matches the current month, with the subsequent 12-month outcomes attached.
Distance(current, historical) = Σ wᵢ × (zᵢ_current − zᵢ_historical)², over the regime sub-componentsThe historical analog explorer is the model that grounds the regime read in actual outcomes rather than theory. Every month, the current values of the regime sub-components (composite score, breadth, dispersion, revisions, sector rotation pattern) are converted to z-scores and compared against the same z-scores at every prior month-end since the index history began. The five closest historical matches are surfaced, with the subsequent 12-month total return and maximum drawdown attached to each. The tool is not a forecast — small samples and changing market structure mean any individual analog is noisy — but the distribution of analog outcomes is informative. A cluster of analogs that all delivered double-digit drawdowns in the following twelve months is a meaningful prior, even if no single analog is conclusive. Analogs are weighted by inverse distance and the explorer publishes a confidence band around the analog-implied return that scales with the spread of the top-five outcomes. Pair the explorer with the scenario probability tree: when the tree's bear-branch fair value lines up with the analog cluster, the conviction on the bear case is materially higher.