Forward-Factor Backtest — 12-Week Lookback
Walks the last 12 Mondays. For each week, picks ATM monthly front/back calls (or puts), checks whether the Forward Factor cleared your threshold, and if so simulates the calendar at Polygon daily closes, exiting at +5 business days. Closes are used as mid approximations — no spread, slippage, or commissions modeled. Results are illustrative only, not a track record. See the live signal · User manual.
Backtest config
Backtest takes 15-45 seconds depending on weeks and Polygon cache hits. Results cached for 24 hours.
Enter a ticker and click Run backtest to see 12 weeks of simulated Forward-Factor calendar entries.
How the backtest works
- For each Monday in the lookback window, fetch the underlying's daily close as the entry reference.
- Query Polygon for option contracts that were listed on that Monday, filtered to the front and back DTE windows and to strikes within ±5% of spot.
- Prefer 3rd-Friday monthly expirations; pick the front and back closest to your DTE targets.
- Find a common ATM strike present in both expirations.
- Solve implied vols from the option daily closes (BSM, risk-free 4.5%, no dividend yield).
- Compute forward IV via variance additivity, then the Forward Factor.
- If FF clears the threshold and net debit is positive, simulate the entry; otherwise skip.
- Exit at +N business days (or 1 day before front expiration, whichever sooner) using daily closes.
- Realized P&L = (back_close_exit − front_close_exit) − (back_close_entry − front_close_entry).
Backtest caveats
- No spread, slippage, or commissions: real fills will be worse than closes-as-mids by 5-20% of the structure value.
- Survivorship in option chains: Polygon's contracts list only returns contracts that existed; delisted weeks are dropped.
- One ticker, one window: 12 entries is too few to draw statistical conclusions. Use as directional evidence, not as a strategy stamp of approval.
- Past performance not predictive: rules that worked in calm regimes can fail in vol shocks (Mar 2020, Apr 2025).
- Not a recommendation: Forward-Factor calendars carry vega and term-structure risk; max loss = net debit; educational tool only.