Gamma Exposure and Dealer Positioning
Definition
Gamma Exposure (GEX) and Delta Exposure convert publicly published Open Interest into an estimate of how much stock dealers must buy or sell to stay hedged — and therefore whether their hedging will dampen or amplify price moves.
The breakthrough is that you never need to know who is long and who is short. You only need one structural assumption about the role dealers play.
Core Ideas
The industry-standard assumption
- End customers are structurally the buyers. Retail and institutions buy more calls than they sell, and buy puts for protection.
- Market makers are the passive counterparty. They do not take directional bets; they provide liquidity and are forced into the opposite side of customer flow.
Why this is reasonable: a market maker’s business model is capturing the bid-ask spread, not predicting direction. They hedge Delta with stock or futures to stay delta-neutral. Because they are structurally short and mechanically hedged, their hedging behaviour is predictable — which is what makes the whole method work.
The nuance that matters: on the call side the assumption runs the other way. See Contradictions below.
The GEX calculation
- Take OI at each strike (public data).
- Take Gamma at each strike (from Black-Scholes — see Options Greeks).
- Assume dealers hold the customer’s opposite position, giving a signed dealer gamma.
- Aggregate:
GEX = OI × Gamma × 100 × spot² × 0.01
(calls usually signed positive, puts negative — sign convention varies by model)
Delta Exposure works the same way: OI × Delta per strike is that strike’s notional delta exposure; assume dealers hold the inverse; sum across strikes to get total dealer Delta — that is, how much stock dealers need to buy or sell to neutralize.
Why this predicts price behaviour
| Regime | Dealer hedging behaviour | Effect on price |
|---|---|---|
| Positive gamma | Sell into strength, buy into weakness | Suppresses volatility — price gets pinned, range-bound drift |
| Negative gamma | Buy into strength, sell into weakness | Amplifies volatility — sharp squeezes and fast selloffs |
This is why GEX gets used to forecast whether price will be pinned to a strike or is prone to a violent one-way move.
Dealer Delta is a snapshot, not a schedule
If total dealer Delta is very negative, dealers lose money as the underlying rises, so they must buy positive-delta instruments (stock or futures) to neutralize. Two corrections to the naive reading:
- Hedging is continuous, not an open-bell event. Dealers rebalance intraday with algorithms. Any move in price, time, or implied volatility shifts Gamma, which shifts Delta exposure, which forces more hedging. The open is just a moment you happen to observe.
- Direction of future hedging depends on Gamma’s sign, not Delta’s. Dealer Delta tells you how much they need to buy right now. Gamma tells you whether the next move makes them buy more or flip to selling.
| Dealer state | Price rises | Price falls |
|---|---|---|
| Delta negative + Gamma negative (heavy put exposure) | Must chase-buy stock to close the negative delta | Must chase-sell, widening the gap |
| Delta negative + Gamma positive | Exposure self-corrects as price rises | Exposure deepens; must buy to hedge |
Decision matrix: GEX × Dealer Delta
GEX and Delta together forecast the shape of volatility, never direction.
| GEX (Gamma) | Dealer Delta | Expected behaviour | Why |
|---|---|---|---|
| Positive (high) | Positive | Range-bound, volatility suppressed | Dealers sell rallies and buy dips; the most stable state |
| Positive (high) | Negative | Mild range with a slow upward bias | Volatility suppressed, but the negative delta still needs buying — a light bid |
| Negative (high) | Positive | Volatility expands, direction unclear, sharp moves both ways | Chase-buy/chase-sell amplifies; long delta means selling pressure is heavier on the way down |
| Negative (high) | Negative | Most dangerous — accelerating selloff or short squeeze | Amplifying hedging plus already-negative delta forces more selling into weakness; a downward spiral (volatility explosion) |
| ≈ 0 (near the flip) | Any | Most uncertain; volatility regime can switch instantly | At the gamma flip point, dealer hedging can reverse from suppressing to amplifying |
Read it as: GEX sets the size of moves. Dealer Delta sets which direction the hedging pressure leans. Whether price actually rises or falls still comes down to fundamentals, news, and macro data — things GEX cannot see.
Strike-level matrix: OI trend + Volume + Price + Gamma
| OI trend | Volume | Price action at strike | Gamma sign | Interpretation |
|---|---|---|---|---|
| ↑ Increasing | High | Approaches, held below (call strike) | Positive (high) | Resistance / gamma wall forming — hedging dampens upward moves, strike acts as a ceiling |
| ↑ Increasing | High | Approaches, held above (put strike) | Positive (high) | Support / gamma wall forming — hedging dampens downside, strike acts as a floor |
| ↑ Increasing | High | Breaks through easily | Negative | Fresh directional conviction — new positioning plus amplifying hedging; the breakout can accelerate |
| ↑ Increasing | Low | Sideways, no reaction | Neutral / small | Passive accumulation (slow institutional hedging or call overwriting) — not yet influential |
| ↓ Decreasing | High | Level near strike fails to hold | Negative (shrinking) | Unwind and breakdown — closing positions remove the hedging wall; volatility likely expands |
| ↓ Decreasing | High | Price moves away from strike | Positive (shrinking) | Normal profit-taking / roll-off; the level is losing relevance into expiry |
| ↓ Decreasing | Low | No reaction | Neutral | Quiet expiry decay — low significance |
| → Flat | High | Oscillates tightly around strike | Positive (large, static) | Established pinning point — a large existing gamma wall magnetizes price (classic max-pain behaviour into expiry) |
| → Flat | High (churn) | Whipsaws around strike | Negative (large, static) | Volatile pinning zone — heavy two-way flow but net negative gamma; expect choppy, violent moves |
| → Flat | Low | No reaction | Neutral | Inactive strike — not a key level right now |
What each dimension contributes:
| Dimension | What it reveals |
|---|---|
| OI trend | Whether positions are being built (↑), unwound (↓), or static (→) |
| Volume | How contested the strike is today — conviction |
| Price action | Whether the strike is currently working as support/resistance, or being ignored |
| Gamma sign | Whether hedging will dampen (positive) or amplify (negative) moves near it |
| Delta sign | Which direction the dealer’s net hedging pressure leans if the move happens |
Four one-line reads:
- OI↑ + high Volume + price rejected + Gamma positive → strongest support/resistance signal; the wall is holding.
- OI↑ + high Volume + price breaks + Gamma negative → breakout likely to accelerate; hedging fuels the move.
- OI flat + high Volume + large positive Gamma → classic pinning into expiration.
- OI↓ + Gamma shrinking → the strike’s influence is fading; watch for volatility expansion as the wall disappears.
Adding Delta: where the same wall is strong versus fragile
Layering dealer Delta onto the strike matrix separates walls that hold from walls that break:
| OI trend | Price action | Gamma | Delta | Read |
|---|---|---|---|---|
| ↑ | Rejected at a call strike | Positive (large) | Positive | Strong resistance — volatility suppressed and dealers already long, so rallies meet selling. Hard to break |
| ↑ | Rejected at a call strike | Positive (large) | Negative | Fragile resistance — volatility suppressed, but dealers must buy to hedge; a break can come with a sharp squeeze |
| ↑ | Held at a put strike | Positive (large) | Negative | Strong support — suppressed volatility plus continuous dealer buying; the floor is solid |
| ↑ | Held at a put strike | Positive (large) | Positive | Fragile support — dealers must sell to hedge; a break lower can accelerate |
| ↑ | Breaks up easily | Negative | Negative | Accelerating rally — new positioning, amplified moves, dealers chasing the bid |
| ↑ | Breaks down easily | Negative | Positive | Accelerating selloff — new positioning, amplified moves, dealers chasing the offer |
| ↓ | Level fails | Negative (shrinking) | Negative | Wall collapses and accelerates — hedging support withdrawn while dealers still need to sell |
| ↓ | Level fails | Negative (shrinking) | Positive | Wall collapses, direction unclear — weaker hedging, but dealer buying can catch the break |
| → | Tight oscillation | Positive (large, stable) | ≈ 0 (neutral) | Classic pin / max-pain candidate — dealers already neutral, market-making keeps price magnetized |
| → | Violent two-way swings | Negative (large, stable) | Volatile | High-churn pinning zone — heavy flow, but net negative gamma means violent whipsaws rather than calm consolidation |
Gamma and Delta must be read together. Gamma alone tells you whether moves get amplified; Delta alone tells you nothing about what happens next. Either in isolation misleads.
Evidence For and Against the Assumption
Supporting
- Gamma sign correlates with realized volatility, significantly. Comparing the highest against the second-highest GEX quantile, next-day SPX 1-day standard deviation is 0.55% versus 0.85%. At an index level of 2000 that is roughly a 4-point-per-day difference in range — reproducible, statistically meaningful support for “negative gamma → amplified volatility.”
- Cross-market validation in FX. Using DTCC data from 21 October 2017 onward, dealer gamma in EURUSD and USDJPY was negative, and option hedging measurably raised realized volatility in both pairs. The mechanism is not equity-specific.
- Partial third-party replication. Independent analysts confirm that from a trading standpoint, deep selloffs with negative or near-zero GEX are a good buy-signal cue with real long-run reference value.
Weakening
- Most dealers do not continuously hedge. A 2024 academic conference paper finds only a minority of market makers hedge continuously; most rely on fast inventory rebalancing instead. Worse, hedging is selective — aggressive against order flow that looks informed, loose against flow judged uninformative. That directly contradicts “dealers always maintain delta neutrality.”
- Raw GEX numbers overstate their own predictive power. Replication work found data errors in early analyses, and the relationship with future returns only holds up once values are normalized by realized volatility, implied volatility, or VIX.
- Limited cross-asset reproducibility. Attempts to reproduce the results on Canadian equities were mixed. The method was distilled from SPX data, so it fits SPX best.
- The originators admit the limits. SqueezeMetrics’ own research says the relationship between the options market and its underlying is still poorly understood.
Why SPX is the best-case market
| SPX property | Why the assumption fits better |
|---|---|
| Cash-settled, European | No early exercise or share delivery, so hedging shows up cleanly in futures and ETFs |
| Institution-dominated liquidity | Single-name options carry retail speculation that muddies the hedging logic; SPX participants behave closer to the model |
| Longest data history | SqueezeMetrics data goes back to 2004, versus roughly 6 months for SpotGamma — a fuller picture of the exposure effect |
| Deep hedging instruments | SPX futures and E-minis absorb large hedges, so dealers can actually do what theory says they should |
The 0.55%/0.85% volatility split and the observed exponential volatility growth under negative GEX were all measured directly on SPX, not generalized from elsewhere.
Where accuracy degrades even in SPX
- 0DTE options. Cboe’s own work notes that judging net dealer gamma requires both magnitude (bigger → larger potential hedging impact) and sign (long gamma means dealers hedge against the move — selling futures as SPX rises, buying as it falls — which suppresses volatility). But the 0DTE share has exploded, so exposure now swings hourly, and a daily snapshot has much less carry-over value.
- The call-side assumption is inverted (see below).
- Raw numbers need volatility normalization before their relationship to forward returns is trustworthy.
Contradictions
Are dealers short calls, or long them?
| Option type | Structural net seller | Structural net buyer | Driving strategy |
|---|---|---|---|
| Call | Investors (covered call writing) | Dealers (forced to absorb) | Buy-write, the short-call leg of a collar |
| Put | Dealers | Investors (buying protection) | Protective put, the long-put leg of a collar |
So “dealers are short” is accurate on the put side and backwards on the call side. Notably this was derived from data, not assumed — skew analysis, the OI distribution across strikes, and (circularly) GEX’s own effectiveness all point to overwriting and collars dominating call supply.
Practical consequence: any GEX read that treats calls and puts symmetrically inherits a sign error on the call side. Treat GEX/Delta as an auxiliary volatility indicator, not a precise readout of dealer inventory.
Relationships
- Open Interest and Volume — the public input GEX is built from, and the T+1 method for judging whether OI change is real
- Options Greeks — Delta and Gamma per contract; GEX is those scaled by OI across the chain
- Implied Volatility — the IV input to the Black-Scholes gamma, and the normalizer that makes raw GEX usable
- Market Microstructure — OTC and Dark Pools — what a dealer is versus a market maker, and why quoting obligations force the counterparty role
- Options Strategies — covered calls, collars, and protective puts are the flows that create the structural imbalance
- Portfolio Risk Management — dealer hedging is delta-neutral risk control at market scale
- Quantitative Trading — GEX as a systematic signal, and why it needs normalization before use
- Trading & Finance
References
Not investment advice — a framework for reading options market structure.