99-Delta Lab

Does the “99-delta ITM covered call” beat its put-spread twin — and does the Keltner entry filter do anything?

SPXW / TNA / TQQQ / UPRO · 2018-01-02 → 2026-07-31 · ~7 DTE · short leg at 20Δ, long leg at the 99Δ-call strike · ThetaData EOD NBBO, Black-76 deltas, parity settlement.

1 — Put spread vs ITM call spread

Identical strikes, identical expiry, identical payoff. Any difference is execution, not edge.

Positive = the put spread earned more. At mid the two should be within noise of each other (parity); at real fills the gap is the cost of crossing the deep-ITM legs.

2 — Exit rules

Every rule marked daily against the same entries. hold is his stated “one and done” baseline.

Cumulative P&L per 1 contract, mid fills. Curves are downsampled for display; tables below carry the exact figures.

avg/worst is the column that decides the stop question — average P&L divided by the worst single trade. Stops nearly always lower the average; the question is whether they buy back more tail than they cost.

3 — VIX regime at entry

Hold-to-expiry only, so the regime effect isn’t confounded with the stop effect.

Each dot is one trade: VIX at entry (x) against realised P&L (y).

4 — Where the tail lives

The 85% win rate is not the story. The left tail is.

Distribution of per-trade P&L, hold-to-expiry, mid fills.

5 — Entry-filter arms

Read this one carefully. The Keltner arm overlaps heavily in time with the daily arm — many of its trades share the same week. A rising curve there is not yet evidence; it needs a non-overlapping test before it means anything.