The gamma flip is a price level on an index or stock. It is where a model of options-dealer positioning says net gamma changes sign. The idea is that the hedging trades dealers must make behave differently above and below that level. It is a model output, not a fact reported by any exchange.
What it is
Option dealers (market makers) take the other side of customer orders and then hedge so they are not exposed to small price changes. The sensitivity of their hedge to price is gamma. When a dealer is long gamma, a rising price makes their position longer, so they sell some stock to rebalance, and a falling price makes them buy. That damps movement. When a dealer is short gamma the rebalancing flips: they must buy into rallies and sell into drops, which can amplify movement.
Nobody outside the dealers sees their real book. So models estimate it from public open interest, using an assumption about who holds what. The gamma flip is the price where that estimated total gamma crosses zero.
How it is calculated
- 1Collect open interest by strike and expiry for every listed call and put on the underlying (the OCC publishes open interest).
- 2For each strike, estimate gamma (the Black-Scholes second derivative with respect to price) using the option's implied volatility and time to expiry.
- 3Convert to dollar gamma per 1% move: gamma x open interest x 100 shares x spot squared x 1%.
- 4Apply a sign convention. A common simple model treats calls as dealer long gamma and puts as dealer short gamma (customers assumed to buy puts to hedge and sell calls for income). This is an assumption, and it is the weakest link.
- 5Re-price that sum at a grid of hypothetical underlying prices. The flip is where the summed curve crosses from negative to positive; find it by interpolating between the two grid points either side of zero.
Worked example (illustrative numbers)
One strike first. Suppose the 600 strike on an ETF trading at 600 has gamma 0.02 and 10,000 contracts of open interest. Dollar gamma per 1% move = 0.02 x 10,000 x 100 x 600 x 600 x 1% = $72 million. Repeat for every strike, sign them, and add them up at each hypothetical price.
Say the summed net gamma (in millions of dollars per 1% move) comes out like the chart below. It is negative at 590 (-1.2) and 595 (-0.5) and positive at 600 (+0.4). The zero crossing lies between 595 and 600, found by interpolation: 595 + 5 x 0.5 / (0.5 + 0.4) = 597.78.
The curve crosses zero at about 597.78: the gamma flip. Illustrative numbers, not live data.
How traders read it
- Spot above the flip: the market is assumed to be in a positive-gamma regime. Realised moves have tended to be smaller and mean-reverting; fading extremes is a common idea to test.
- Spot below the flip: a negative-gamma regime. Moves can extend and intraday ranges can widen; many traders reduce size or widen stops rather than assume a bounce.
- Distance to the flip matters. Spot 0.2% above it is a very different situation from 3% above it, and the level moves as prices and open interest change.
- The flip is a regime marker, not an entry signal. Pair it with your own levels, the calendar (CPI, FOMC) and risk limits.
Common mistakes
- Treating it as a hard support or resistance line. It is a regime boundary, and price can cross it without anything happening.
- Forgetting the sign assumption. If customers are net long calls and short puts in a given name, the sign flips and so does the story.
- Using stale data. Open interest updates once a day; positions opened or closed intraday are not in it.
- Comparing flip levels across data vendors as if they were the same number. Different volatility inputs and sign rules give different levels.
- Ignoring zero-day expiries (0DTE), which can add large gamma that open-interest snapshots miss.
Limits
The model cannot see who is long or short. It assumes a hedging pattern, it ignores futures, swaps and off-exchange positions, and it treats all dealers as one. Research on option-expiration effects (see sources) finds hedging can influence prices near large strikes, but the effect is small and uneven. Use the flip as context for risk, not as a forecast.
FAQ
Is the gamma flip the same as the zero gamma level?
Usually yes. Both names describe the price where modelled net dealer gamma is zero. Vendors differ in inputs, so their numbers differ slightly.
Does price move toward the flip?
Not reliably. It marks where behaviour is assumed to change, not a target.
How often does it change?
Every time open interest, implied volatility or price changes. On this site the board is end-of-day, so it is a snapshot for the next session.
Can I use it on any stock?
It is most informative on very liquid, heavily optioned names and indexes (SPY, QQQ, SPX). On thinly traded options the estimate is noisy.
Sources
- Options Industry Council (OIC): Options education
- Options Clearing Corporation (OCC): Volume and open interest reports
- Cboe: Learn (options Greeks and implied volatility)
- Ni, Pearson and Poteshman (2005), Stock price clustering on option expiration dates, Journal of Financial Economics
- Golez and Jackwerth (2012), Pinning in the S&P 500 futures, Journal of Financial Economics