Options and volatility
Initial option-information studies tested whether statistical associations survived later periods and transaction assumptions.
- Research status
- Early-stage options research; no standalone strategy
- Research scope
- Equity and digital options · 1 hour
Archived experiment
Question. Did option-market information add a repeatable, cost-aware forecasting result for underlying equities?
Evaluation. 161 untouched complete dates, 5 January–25 August 2026, across six underlyings and four chronological blocks. The strongest gross curiosity among the frozen tests is summarized below without its signal definition.
| Comparison | Mean per observation |
|---|---|
| Before the research cost charge | +10.10 bp |
| After a 20 bp charge | −9.90 bp |
Stress and result. Charges of 20, 30 and 40 bp were tested alongside multiplicity, date-concentration and regime checks. At 20 bp only one of four blocks was positive; the negative mean and stability checks rejected the result.
Decision. Reject this hypothesis. A signed underlying-return proxy does not demonstrate executable option-contract returns. Actual option spreads, multi-leg fills, margin, assignment and expiry require separate transaction evidence.
Research record
The redacted aggregate results retain the period, metric definitions, assumptions and decision. Signal formulas, parameters, rankings, holdings and raw data are excluded.