Zero Dte Gamma

zero-dte-gamma

You hand over an underlying (default SPY). The skill pulls the options chain snapshot for the nearest expiry, computes per-contract gamma exposure using Black-Scholes greeks and reported open interest, aggregates by strike, and reports the net dealer gamma regime along with the top pins.

Motivated by 2024-25 research (Baltussen-Terhorst-Van Vliet 2024, Bhattacharya 2024, others) documenting that 0-day-to-expiration options now drive systematic intraday moves through market-maker delta hedging pressure. This phenomenon didn't exist meaningfully before 2022 (when CBOE expanded 0DTE availability); by 2024-25 it's a first-order intraday factor.

Interpretation

• Long gamma regime (net dealer gamma > 0): dealers hedge against price moves, compressing intraday range. Late-day chop typical. Sell-vol strategies favored.
• Short gamma regime (net dealer gamma < 0): dealers hedge with the market, amplifying moves. Trend days more likely, especially in the last hour. Buy-vol / breakout strategies favored.
• Gamma flip strike: the level where cumulative dealer GEX crosses zero. Break past it and the hedging regime changes.
• Gamma pins: strikes with the largest concentrated open interest gamma. Spot tends to gravitate toward these on expiry day.

When to invoke

• "What's the gamma regime on SPY today?"
• Pre-market prep on an SPX / QQQ options trader's watchlist
• Sizing risk for a 0DTE strategy
• The user says "gamma flip", "gamma pins", "0DTE", "dealer positioning"

Not for: single-name equity options (this is calibrated to index / ETF flow assumptions). Not for real-time (this is snapshot-based; end-of-day is fine, intraday drift can be substantial).

What you need

• Underlying ticker (--underlying, default SPY)
• MASSIVE_API_KEY exported
• Options Developer or higher entitlement. Returns a clean NOT_AUTHORIZED tier caveat without it.

Optional:

• --expiration-date (YYYY-MM-DD): pin a specific expiry. Default: nearest listed expiration to today.
• --risk-free-rate (default 0.045)
• --default-iv (default 0.15): fallback when Massive's greeks or IV field is missing on a contract.

What you get back

Two output layers.

Layer 1: canonical JSON. Per-strike call_gamma_notional, put_gamma_notional, dealer_gex, cum_dealer_gex, call_oi, put_oi. Top-level net_dealer_gex, gamma_regime, gamma_flip_strike, top_gamma_pins (top 5 by absolute notional), plus spot, days_to_expiration, and full gamma_by_strike for downstream consumers.

Layer 2: rendered note. Header + regime label + gamma flip level, top 5 pin table, one-line Take.

How it works

1. Pick nearest expiry from /v3/reference/options/contracts?underlying_ticker={U}.
2. Fetch chain snapshot from /v3/snapshot/options/{U}?expiration_date={D}. Massive returns per-contract greeks + open interest + IV.
3. Compute gamma per contract. Prefer Massive's returned gamma; fall back to Black-Scholes with the reported IV (or default_iv when missing).
4. Cash gamma per contract = gamma × OI × 100 × spot² / 100. This is dollar-gamma per 1% underlying move.
5. Dealer positioning assumption: short customer calls, long customer puts. So dealer_gex(call) = -cash_gamma, dealer_gex(put) = +cash_gamma. This is the standard 0DTE convention; not exact for any given book, but consistent across time.
6. Aggregate per strike, compute cumulative GEX walking from lowest to highest strike, find gamma-flip strike where cumulative crosses zero.
7. Top pins = strikes with the largest total notional gamma (call + put), sorted descending.

Foundations used

• massive-api-patterns for REST auth, retry, and options chain snapshot.

Output mode: note

Narrative note with a per-strike table. A single expiry chain produces 50-200 strikes; the top-5 pin table is the digestible view.

Endpoints used

• GET /v3/reference/options/contracts?underlying_ticker={U} (list expirations)
• GET /v3/snapshot/options/{U}?expiration_date={D} (chain snapshot with greeks + OI)
• GET /v2/snapshot/locale/us/markets/stocks/tickers/{U} (spot fallback chain)

Doesn't handle (yet)

• Intraday updates. Snapshot only. For live updates, wire the same aggregation onto the options WebSocket.
• Vanna and charm. Only gamma. Second-order greeks (vanna = d²/dS/dσ, charm = d²/dS/dt) are the natural next layer for a full "dealer hedging response" model.
• Actual dealer books. The short-calls / long-puts assumption is retail-flow convention. On event days (Fed, CPI, earnings), actual dealer books can invert.
• Non-index underlyings. Calibrated for SPY / SPX / QQQ / IWM where the flow assumption holds. Single-name equity gamma has different flow dynamics.
• rBergomi-consistent IV. Uses reported IV as-is; a rough-vol- consistent IV surface would be a real research extension.

These are clean PR extensions.