Two different things share one name
Ask two traders what negative gamma means and you can get two answers that never touch. One describes a position: a short-options book whose losses accelerate and whose gains stall out. The other describes the whole market: dealers are short gamma, so their hedging pushes price in the direction it is already going.
Both are correct. They are the same arithmetic viewed from different distances — one applied to your account, the other to the aggregate inventory of everyone who makes markets in options. Most explainers pick a lane and never mention the other, which is why the term reads as vague. It is not vague. It just needs a scope attached before it means anything.
Negative gamma in your own book
Gamma is the rate of change of delta. Delta tells you how much a position moves per one-dollar move in the underlying; gamma tells you how fast that sensitivity is itself changing.
Long options have positive gamma. Short options have negative gamma. That is the entire rule, and everything else follows from it.
Concretely: sell a call with a delta of 0.40 and a gamma of 0.05 and you are short 40 deltas per contract. The underlying rises a dollar and delta becomes roughly 0.45 — you are now short 45 deltas, having got shorter into a market that just went up. Another dollar higher and delta is near 0.50. Your loss per dollar grows with every dollar. Run the same move downward and it reverses: delta shrinks toward 0.35, then 0.30, so the position earns less on each additional dollar of favourable move.
Losses accelerate, gains decelerate. That asymmetry is what short convexity means, and it is not a defect in the trade — it is the thing you are being paid to carry. Negative gamma travels with positive theta, and the premium collected is the compensation. Short straddles and strangles, credit spreads, iron condors and covered calls all sit on this side of the ledger. Their equity curve is a long run of small wins interrupted by moves large enough to undo several of them, which is why sizing governs the outcome more than strategy selection does.
The property that matters most: negative gamma is a path risk, not a direction risk. A short strangle does not care whether the underlying goes up or down. It cares how far, and how fast.
Negative gamma in the market
Now widen the frame. Options dealers take the other side of customer flow and hedge the resulting delta in the underlying. When their aggregate inventory leaves them short gamma, they face exactly the accelerating-delta problem above — except they neutralise it continuously, in size, by trading the underlying.
That hedging is mechanical and its direction is forced. A dealer who is short gamma gets shorter delta as price rises, so they must buy to stay neutral. They get longer delta as price falls, so they must sell. Buying strength and selling weakness is pro-cyclical: it adds to whatever move is already underway.
Positive gamma inverts it. A long-gamma dealer sells into rallies and buys into declines, draining energy out of moves and producing the tight, mean-reverting tape that characterises most sessions. Same hedging discipline, opposite sign, opposite market character. Gamma exposure is the aggregate measure of which side dealers are on.
The pro-cyclical loop running upward, on a single name with concentrated call positioning, is a gamma squeeze. The downside version has no catchy name, but it is the identical mechanism with the sign flipped, and it is the one that turns up in crash post-mortems.
Where the regime changes
Dealer gamma is not a constant. It is a function of spot, because the strikes dominating the book change as price moves through them. The level where the aggregate crosses zero is the gamma flip: above it dealers are net long gamma and damp moves, below it they are net short and amplify them. The GEX flip price covers how that level is derived and why the tape changes character on each side of it.
Two consequences follow. First, "we are in negative gamma" is a statement about where price currently is, not a standing condition — a twenty-point move can end it. Second, the flip level itself migrates as positioning changes, so yesterday's number is a stale answer to today's question.
What the evidence actually supports
This is where most negative-gamma writing stops being careful, so it is worth separating what is documented from what is asserted.
The academic case is real but conditional. Barbon and Buraschi's Gamma Fragility builds a stock-level proxy for dealer gamma imbalance across a large panel of equity options and finds that negative ex-ante imbalance interacts with illiquidity to produce intraday momentum, where positive imbalance produces reversal; more negative levels correlate with higher stock market volatility, and imbalance relates to the frequency and magnitude of flash-crash events. The effect is strongest in the least liquid underlyings, which means illiquidity is doing as much work in that result as gamma is. Separately, in the Review of Financial Studies, Ni, Pearson, Poteshman and White document a noninformational channel through which option market maker hedge rebalancing affects stock return volatility and the probability of large stock price moves.
The counter-evidence is equally real, and it comes from the exchange. Mandy Xu's Cboe review of SPX 0DTE market impact, published September 2023, estimates dealer net gamma exposure from 0DTE contracts at $170mm to $670mm through the day — potential hedging flows amounting to between 0.04% and 0.17% of daily S&P futures liquidity. It found no uptick in intraday gap moves, and a year-to-date spread between close-to-close and intraday realised volatility of 2.7 volatility points, which it notes is exactly the ten-year average. Customer flow, on that analysis, is far more two-sided than the popular narrative assumes.
The two are reconcilable. Negative gamma is an amplifier whose force scales with how thin the underlying market is relative to the hedging requirement. In an illiquid single name it can dominate the tape. In S&P futures on an ordinary session it is a rounding error against the available liquidity. Treat it as a conditioning variable — it tells you what a move will feel like once something starts one — rather than as a cause of moves.
Checking whether you are actually in it
The regime is inferred, never observed, and three things routinely corrupt the inference.
The positioning data is yesterday's. Standard gamma charts are built from open interest, a settled figure as of the prior close that does not update while the market is open. On an index where 0DTE contracts reached a record 66.2% of total SPX volume in July 2026, most of the day's gamma is created and destroyed inside the session and never reaches an open-interest chart until the next morning.
The sign is assumed. Nobody publishes dealer inventory. SpotGamma's free SPX gamma exposure page states the common convention plainly: liquidity providers treated as short puts and long calls. The SqueezeMetrics research guide instead infers direction from option volume and the subsequent change in open interest, and is explicit that rising open interest does not by itself imply dealers took on exposure. Two defensible models, two different regime calls on identical data. What to check before trusting a free gamma exposure chart covers the rest of the failure modes.
The tape is the tiebreaker. A negative-gamma regime has a signature that needs no chart: fades fail, ranges expand through the session instead of contracting, and moves that begin near the extremes keep extending. If the model says negative gamma and price mean-reverts cleanly all day, the model is describing yesterday's book.
SquawkFlow's free SPX GEX page is the live companion to all of this — the current gamma flip, the call and put walls, and how often each wall has actually held, that last figure being the only honest way to know whether a level has been worth respecting.
What it does not tell you
Negative gamma says nothing about direction. It describes the environment a move happens inside, not whether one is coming or which way it will go. The most common misreading is treating a negative-gamma reading as bearish: the loop amplifies rallies exactly as readily as it amplifies selloffs, and the sharpest short-covering days in a downtrend tend to occur in the same regime as the sharpest declines.
The usable output is smaller and duller than the commentary around it. In negative gamma: widen stops, cut size, expect trends to persist and fades to fail, and stop expecting the day to close where it opened. In positive gamma: the opposite, and premium selling stops fighting the environment. That is a position-sizing input rather than a signal — and read that way, it is one of the few pieces of market structure that reliably changes how a session should be traded.
Educational content, not financial advice. See our risk disclosure.