Strategy explained · 0-DTE SPX

Pay a dime.
Aim for the pin.

Every morning Butterfly Guy buys one out-of-the-money SPX butterfly that expires the same afternoon, parked in the direction the market gapped. It costs about a tenth of its width. If SPX settles near the middle strike it pays ~10× what it cost. Most days it doesn't — the few that do carry the rest.

Max price
≤10% of width
Reward : risk
8–12 : 1
Expiry
Same day
spot at entry +1 upper wing −2 center +1 lower wing 10:00 4:00 max payout payoff →
Try it

Build today's butterfly

Pick the opening gap and the VIX. The lab applies the bot's selection rules to a modeled 0-DTE chain and shows what it would buy, or why it would sit out. Then run the clock to watch its value curve tighten into the payoff tent by the close.

Opening gap
0-DTE chainmodeled · mid = mark
StrikeBidMidAskOrder
10:00 4:00

Prices are modeled to show the mechanics (Black-Scholes, time from the entry window to the close, flat vol scaled from the VIX). The curves before 4:00 are the same model with less time left; the bot's trailing exit acts on that intraday value, not on the tent. They are not live quotes, and real skew shifts them. The selection mirrors the bot: center within ± pts of the VIX target, cost under the per-width cap, reward:risk of at least 8:1, closest to 10:1, wider wing on ties.

How a day runs

  1. 01 · open

    Read the gap

    Compare SPX with yesterday's close. Gapped up: buy a call fly above the market. Gapped down: a put fly below it.

  2. 02 · size

    Let the VIX set the tent

    The VIX regime picks three wing widths. Each gets a center 0.25σ / 0.5σ / 0.75σ of the VIX-implied daily move away from spot.

  3. 03 · 10:00–10:45 ET

    Only buy it cheap

    Take the fly closest to 10:1 that costs at most 10% of its width. One trade a day, one contract. Worst case: lose the debit.

  4. 04 · manage

    Trail the peak

    Once it's been green, exit if its value falls this far below its high. Otherwise it rides to the 4:00 settle.

    60%9:30–11:30
    90%11:30–1:30
    75%1:30–4:00
Best paper hit so far

Aug 3, 2026 · Call fly 7,575 / 7,600 / 7,625

Paid $209. Came back as $2,450. One day like this covers a long run of small misses — which is the whole design.

Return11.7×
PnL+$2,241

Paper trading only: no real money is behind these trades. Rules and limits are read from the bot's live configuration each time this page is built. This describes a trading experiment and is not investment advice.