Swolecharts
Live crypto data · 10 exchanges
Log inSign up

Gamma exposure (GEX) and the dealer assumption

Separate public options data from assumptions about dealer positions and hedging.

Last updated: 17 October 2026

BTC gamma exposure · model estimate

As of 04:14 UTC · live
Net GEX · model estimate
USD per 1% move
Zero-gamma level · model estimate
$81,528.3
Nearest crossing within ±20% of the index
Unsigned total · model estimate
$157.31M
USD per 1% move; no positioning assumption

Bybit and Binance; all eligible expiries. Signed GEX assumes dealers are long calls and short puts. This is not measured hedge flow.

What gamma is

Delta is an option's model price sensitivity to its underlying. Gamma is the rate at which that delta changes as the underlying price changes. A long call and a long put both have positive gamma; short positions have negative gamma. Gamma tends to be largest near the money and, near the money, grows as expiry approaches. This does not mean every far-out-of-the-money option has high gamma near expiry.

Our BTC and ETH gamma pages calculate Black-76 gamma on the forward, with interest rate r = 0. Using F for the forward, K for strike, σ for annualized mark IV as a decimal and T for remaining years:

d1 = (ln(F / K) + σ²T / 2) / (σ√T)
gamma = φ(d1) / (F × σ × √T)

Here φ is the standard normal density, φ(x) = exp(−x² / 2) / √(2π), and T uses a 365-day year. Gamma is per unit of the forward, the same for calls and puts before assigning a position sign. The model follows Black's forward-option framework. Mark IV is a pricing input, not a forecast; see implied volatility explained.

For a synthetic check with F = K = $100, σ = 0.50 and T = 1 year, d1 = 0.25 and gamma ≈ 0.00773336 per dollar. The site excludes expiries with less than 1 hour remaining to avoid the extreme near-expiry sensitivity.

From gamma to dollar exposure

A 1% forward move is 0.01 × F. Multiplying gamma by that move estimates the change in delta per coin of open interest; multiplying by OI in coins gives the change in underlying hedge units. Multiplying by F converts those units to dollars. Thus our per-option signed estimate is:

GEX_usd = sign × gamma × oi_coin × F² × 0.01

This is the approximate dollar value of the delta adjustment for a 1% move, holding the current gamma fixed. It is not option premium, profit or loss, the derivative of dollar-valued delta including its price revaluation, or observed trading volume. Across options, add the signed estimates for net GEX and the positive magnitudes for unsigned total gamma.

Synthetic two-strike example: not live data

Suppose two hypothetical options share F = $100. At strike $100 a call has gamma 0.00773336 per dollar and OI of 10 coins. At strike $110 a put has an illustrative gamma of 0.006 per dollar and OI of 4 coins. The second gamma is a supplied synthetic input, not a claim about the live chain.

  • Call: +1 × 0.00773336 × 10 × 100² × 0.01 = +$7.73336 per 1% move.
  • Put: −1 × 0.006 × 4 × 100² × 0.01 = −$2.40 per 1% move.
  • Net: +$5.33336; unsigned total: $10.13336 per 1% move.

A falling price reverses the first-order delta adjustment. For larger moves gamma itself changes, so scaling this local estimate is unreliable.

The dealer assumption

The site's explicit assumption is: dealers are long calls and short puts sold to customers. Its sign rule is calls count +, puts count −: sign = +1 for calls and −1 for puts. This is a convention applied to open interest, not an exchange report of dealer inventories.

Bybit's option tickers publish open interest and mark IV; Deribit's book summary documentation likewise describes open interest and mark IV. These fields do not identify who holds the long and short sides. Deribit documentation is a reference for field meaning, not a claim that Deribit is included in this site's live coverage; check the linked options page for eligible venues.

If dealers instead hold the opposite positions, signed contributions reverse. Mixed inventories can change both magnitude and sign. The unsigned total counts calls and puts positive and is assumption-free with respect to positioning. It still depends on the pricing model, mark IV, data quality and coverage; it is not a measurement of dealer risk.

What positive and negative net GEX are claimed to imply

The SqueezeMetrics Gamma Exposure white paper develops this sign convention and hedging argument for equity-index options. Under the assumed inventory and delta-neutral hedging, positive gamma would lead a dealer to sell into rises and buy into falls, potentially damping moves. Negative gamma would lead to buying rises and selling falls, potentially amplifying moves.

Crypto options market structure differs from equity-index options. These are hypotheses, and this page makes no prediction. Dealers may hedge elsewhere, change positions or choose not to rebalance immediately. Other flows and available liquidity matter. A positive reading does not promise calm prices; a negative reading does not establish that volatility will rise.

The zero-gamma level

The zero-gamma level is the price at which the model's aggregate signed gamma crosses zero. The site recomputes gamma over an 81-point grid from 20% below to 20% above the index price, scaling each option's forward by the same price ratio while keeping IV, time and open interest fixed. It linearly interpolates crossings and chooses the one nearest the current index.

None in range means no crossing was found within ±20%, including when all grid values are zero. It does not mean that gamma is zero everywhere or that a crossing cannot exist outside the range. A finite grid can miss crossings between samples. The level changes with the snapshot and model inputs; it is not a support, resistance or price target.

Limits of the estimate

  • Public open interest reveals neither dealer ownership nor hedge execution. The dealer sign can be wrong.
  • Only eligible public venues and usable rows are included. Missing mark IV or forward values are skipped; incomplete expiry data and expiries under 1 hour are excluded.
  • Mark IV, a zero-rate Black-76 model and fixed-input repricing simplify volatility smiles, funding and real hedging costs.
  • Snapshots are recorded every 5 minutes, can be delayed and cover only part of the market. Missing data is not zero exposure.
  • The unsigned total removes the position-sign assumption but retains all model and coverage limits. Neither view is trading advice or evidence that a signal works.

Checked against the exchanges' own documentation on 17 October 2026: Black (1976): The pricing of commodity contracts · Bybit API: option ticker fields · Deribit API: book summary fields · SqueezeMetrics: Gamma Exposure white paper. Not financial advice.

Gamma exposure FAQ

Is GEX real data?

Open interest and mark IV are exchange data. Gamma and signed GEX are model estimates; dealer positions and actual hedge trades are not observed.

Why does GEX differ between sites?

Venue coverage, open-interest units, timestamps, volatility inputs, expiry filters, pricing models and dealer sign assumptions can differ. Compare those definitions before comparing headline values.

Does negative gamma cause volatility?

Negative estimated gamma does not establish causation or predict volatility. Actual positioning, hedge choices, liquidity and other flows can overwhelm or reverse the hypothesized feedback.