The datasets, tested
against real prices.

Stock Dilution

A point-in-time record of S-1 registration statements, flagged dilutive vs. non-dilutive. The dilutive flag is a supply-shock signal — most violent in the smallest caps, where a secondary offering is enormous relative to float.

Median -14.9% at 1 month (-15.5% vs. SPY), 71% of filings lower (z = 8.9). Monotonic in size: nano-caps 74% down, large-caps 43%.

De-SPAC Events

Completed SPAC-to-operating-company mergers, parsed from SEC filings. A structurally challenged cohort that tends to drift down after completion — and harder when shareholders redeem heavily.

Median -29% at 3 months, -47% at 6 months; 74% lower at 3M (z = 9.3). Heavy redemptions roughly triple the damage (-43% vs. -16%).

Corporate Default Events

Public-company default events labeled from SEC filing text — missed payments, covenant breaches, accelerations, bankruptcies. A hard credit signal that keeps hurting the equity after disclosure.

Median -34% at 3 months; 61% lower a month later (z = 4.6). Severity-ordered: missed payments -37%, soft events near flat.

Earnings Implied vs Realized

A per-name track record of how each earnings move was priced (pre-earnings ATM straddle) versus how it actually moved. A direct read on the volatility risk premium. Gross of costs; the premium is market-wide, not name-specific.

Straddles overprice the realized move 64% of the time (z = 16.3), +1.1% of spot per event (t = 10.4); +4,024 points captured cumulatively.

Next-Day Movers

Each afternoon, the equities the model rates most likely to make a large move next session. A magnitude signal (it makes no directional claim), built for non-directional vol trades and movement screens.

Selected names realize a 4.8% median move vs. 2.4% on a normal day (2.0x); 5%+ moves hit 48% vs. 24%.

IV/HV Rank (52-Week)

Where a name's implied and realized vol sit inside their own trailing 52-week range. Extremes mean-revert, which makes the rank a clean timing signal for buying vs. selling premium.

Top-of-range IV falls -24.9 vol points over the next month; bottom rises +11.2. Monotonic across deciles (corr -0.36 on 141k obs).

S&P 500 0-DTE Strike Band

The morning strike range SPX is expected to hold through the same-day 0-DTE close. A calibrated expected-move for strike selection. (SPX is proxied here by SPY x10; tests are run basis-free.)

Predicted width contains the day 91.5% of the time; 80% close containment; band width correlates 0.80 with the realized range.

Quant Galore Momentum Index

A rules-based, 10-stock, monthly-rebalanced cross-sectional momentum basket. It compounds far past the benchmark — real, significant outperformance, but high-beta momentum risk, not a smooth ride.

$1 grows to $30.4 vs. $2.9 for SPY since 2017; beats the benchmark 61% of months, +2.84% mean monthly excess (t = 2.7).

IV/HV Premium

Per name and day, implied vol measured against realized — as a spread and a ratio. A conditioner: it tells you when implied is genuinely rich, which is the one regime where selling vol pays (selling blindly in this universe loses).

Seller edge sorts monotonically -9.4 to +5.1 vol points across IV/HV-ratio quintiles (14.5-pt long-short spread); win rate 27% to 53%.

Dividend Capture

Ex-dividend events with drop ratios, net capture, and recovery odds. The mechanical claims hold tightly — and that is the alpha: the dataset lets you avoid a trade that looks like free yield but is near-zero EV.

Ex-day drop averages 0.87x the dividend; net capture is +0.04% with 45% of events underwater; recovery 55% (1d) to 91% (20d).

S&P 500 Risk Regime

A daily risk-on / risk-off flag. Read honestly: it is a strong, forward-looking volatility signal for sizing exposure and timing vol trades — not a market-direction call. Returns are the same in both regimes, and we show the flat chart deliberately.

Risk-off days realize 24.7% vol vs. 12.8% (2x); forward 10-day vol 19.0% vs. 12.3% (t = 15.0). Returns: 13.0% vs. 12.7% — not directional.

Vol-of-Vol Index

How unstable a name's volatility has been over the trailing month. The weakest signal in the suite, reported as such: it is not a clean cross-sectional factor — it earns its keep only as a top-decile tail flag.

Correlation with forward vol is just 0.05 and non-monotonic. At the extreme it works: the top decile realizes 61% vol and 11% max moves vs. ~8% elsewhere.