Options dealer hedging & gamma positioning

Definition

Options dealers (market makers in ms-option-chain terms) run hedged books: they take the other side of customer options flow and neutralize direction with stock (mgmt-delta-hedging at industrial scale). The consequence traders care about is that dealer hedging generates MECHANICAL stock flow whose direction depends on the sign of dealers’ aggregate GAMMA — and Ni-Pearson-Poteshman-White (2021, RFS) documents that this rebalancing measurably moves underlying prices. LONG-gamma dealers trade AGAINST moves (sell rallies, buy dips — dampening); SHORT-gamma dealers trade WITH moves (amplifying). This sign is the most important single fact about how the options market feeds back into the stock market.

How it works / structure

• The mechanics: a dealer short calls gets longer delta as price rises and must SELL stock into the rally to stay hedged only if long gamma — signs matter: LONG gamma (customers net sold options to dealers) → hedging is counter-trend, pinning and dampening; SHORT gamma (customers net bought options) → hedging chases price, adding momentum and gap risk; greek-gamma carries the per-position math this aggregates.
• The documented effects (NPPW): stock return patterns consistent with hedge rebalancing — including expiration-week pinning toward high-open- interest strikes (event-opex clustering evidence traces to the same paper lineage) — survive controls; the effect is pervasive, not anecdotal.
• The practitioner overlay (labeled): “GEX” dashboards estimate aggregate dealer gamma from open interest by ASSUMING who holds what (customers long puts/short calls, typically) — the estimates are assumption-heavy inference, not disclosure; useful as regime context (positive-GEX regimes empirically quieter, negative-GEX regimes jumpier), unreliable as precision signals; the KB carries them labeled practitioner throughout.
• The stress form: short-gamma cascades — dealers forced to sell into falls — are documented episode mechanics (episode-volmageddon-2018; the Feb 2018 and various 0DTE-era intraday accelerations); opt-0dte-mechanics concentrates the same feedback intraday.

When it applies

Regime context for index trading (dealer-gamma sign estimates condition expected realized volatility and intraday behavior — mean-reverting vs trending sessions); expiration tactics (event-opex pinning and the post-expiration “unclenching”); options-flow interpretation (indicator-options-flow — large customer prints CREATE dealer hedge flow, the second-order effect often exceeding the first); understanding volatility events (short-gamma anatomy).

Risk profile & failure modes

• Assumption fragility (the GEX trap): aggregate gamma estimates flip sign under different who-holds-what assumptions — treating a dashboard number as ground truth imports its unverifiable assumptions; use as coarse regime tint, sized accordingly.
• Reflexive crowding: gamma-level trading is now itself crowded — documented decay pressure on the patterns (philosophy-adaptive-markets); pinning effects weakened as they became consensus.
• Sign regime flips: the dampening regime trains mean-reversion habits that the amplifying regime executes — the transition (usually via large put buying or expiration roll-off) is exactly when learned behavior inverts.
• Single-driver narrative: dealer flow is ONE force among macro, earnings, and positioning — post-hoc “gamma did it” explanations are mostly unfalsifiable; the KB requires the OI and flow evidence attached.

Evidence & limits

NPPW (2021, RFS) is the peer-reviewed anchor for hedging impact and pinning; episode records document the stress form. Aggregate dealer positioning is not disclosed — all real-time gamma maps are estimates, and the KB’s labeling rule is strict about it.

Falsifiable-thesis examples

Illustrations only, not signals:

• “Index realized volatility is lower on days estimated positive-gamma vs negative-gamma, same VIX level (regime-conditioning check)” — falsified by the conditional realized-vol spread.
• “Stocks close within 0.5 strikes of a high-OI strike on expiration Fridays more often than non-expiration Fridays (pinning replication)” — falsified by the clustering comparison.

Cross-references

• The per-position math: greek-gamma, mgmt-delta-hedging, strategy-gamma-scalping
• The calendar node: event-opex; the intraday concentrate: opt-0dte-mechanics
• The flow lens: indicator-options-flow; the stress exhibit: episode-volmageddon-2018