Psymetry™ — methodology brief. DAY ONE is powered by a proprietary three-axis behavioral synthesis framework that scores IPO candidates across Psychofinance (the past behavior of the company and its founder, including financial track record and the market regime they're listing into), Psychohistory (what the outside world is saying about both — press sentiment, social signals, narrative arcs), and Psychobehavioural (an opaque founder-behavior index synthesized from publicly observable conduct, communication patterns, governance signals, and HR/litigation traces). Each axis is normalized to 0–100 and combined into a composite Pop Score. Verdict tiers — Strong Buy ≥ 60, Buy 50–59, Avoid 40–49, Strong Avoid < 40 — are calibrated against historical Day-1 outcomes.

Psychobehavioural Index. The PBI is composed of three normalized axis markers — PB-α, PB-β, and PB-γ — each on a 0–10 scale, summing to a 0–30 total. The exact derivation of each marker is proprietary. We will say only that they are computed from a weighted ensemble of publicly observable behavioral indicators. When the PBI exceeds 16, the verdict guardrail engages and any Strong Buy is automatically demoted to Buy.

Out-of-sample testing. The 102-case backtest is stratified by outcome bucket and split 67% TRAIN / 33% TEST. Auto-tune optimizes axis weights against the TRAIN set only; the headline accuracy numbers above are TEST set — cases the optimizer has never seen.

Financial-only benchmark. The strip below the headline metrics compares Psymetry against a naive financial-only baseline (Psychofinance · Financial weighted at 100%) on the same TEST set. The "behavioral lift" is the predictive accuracy that Psychohistory + Psychobehavioural add on top of pure financials.

Blind scoring · acknowledged limitation. Historical axis scores in this MVP were assigned by a single analyst with knowledge of actual outcomes. Hindsight bias possible. The forward predictions (SpaceX, Nscale) are genuinely blind — scored before any outcome exists. Treat the backtest as framework demonstration; treat the forwards as the real test.

Day-1 % = (Day-1 close − IPO offer price) / IPO offer price. MAE midpoints: Strong +45%, Moderate +20%, Mild +5%, Flat −10%. Not investment advice.