What is a gamma squeeze?

Delta and gamma, in plain language

A standard listed US equity option controls 100 shares. (Contracts adjusted after a split, spin-off, or merger can carry a non-standard deliverable, so check the contract spec before assuming the multiplier.) Delta tells you roughly how many of those 100 shares the contract currently behaves like: a 0.30-delta call moves about like 30 shares of stock for a small move in the underlying. Traders often read delta as a rough probability of finishing in the money — that is a related but distinct number, and for this discussion delta is best held as exposure.

Gamma is the rate at which delta changes as the underlying moves. If a call has 0.30 delta and 0.05 gamma, a $1 rise in the stock takes delta to roughly 0.35 — the position quietly behaves like more shares without anyone trading anything. That is convexity, and it is not free: the buyer pays for it in premium and gives it back through time decay. Long gamma means exposure grows when the move goes your way and shrinks when it does not, in exchange for the option losing value if the underlying simply sits still.

Two properties of gamma drive everything below. First, gamma is largest for strikes near the current price and falls away in both directions. Second — and this is the part that gets stated carelessly — shortening the time to expiration does not raise gamma everywhere. It concentrates gamma. A 45-day call spreads modest gamma across a wide band of prices; on expiration morning the same strike carries enormous gamma in a razor-thin band around itself and almost none outside it. For a strike that price has left behind, near-dated gamma collapses toward zero rather than rising.

Why the hedger on the other side buys stock

When you buy a call, somebody sells it to you. Often that somebody is a market maker whose business is capturing the spread rather than taking a directional view. Having sold the call, they are short delta, and the standard response is to buy shares of the underlying so the combined book is roughly flat to small moves.

That hedge is not set-and-forget. A short call position is short gamma: as the stock rises, the call's delta rises, and a share hedge that was adequate at $50 leaves the book short delta at $52. Staying neutral means buying more stock on the way up and selling it back on the way down.

Notice the asymmetry in behaviour. A short-gamma hedger buys strength and sells weakness — trading with the move and adding to it. A long-gamma hedger does the reverse, selling into rallies and buying dips, which dampens moves instead. Which regime dominates in a given name is precisely what is not observable from public data.

It is also worth being clear that the market maker is not automatically the short side. When customers are net sellers of calls — covered-call writing and systematic overwriting are large, persistent flows — the dealer is long those calls and therefore long gamma, and their hedging damps the move rather than amplifying it. Nor does hedging have to happen in shares: a desk can offset an option with another option, hedge on a schedule instead of continuously, or carry the risk within limits and not hedge the increment at all.

The feedback loop that defines a squeeze

A gamma squeeze is what happens when that hedging becomes large enough, relative to normal trading in the stock, to move the very price that is driving it.

Hypothetical illustration, not a real trade and not a claim about any ticker: if 20,000 calls sit at one strike and their delta travels from 0.30 to 0.60 over an afternoon, the share hedge behind them moves from roughly 600,000 shares to roughly 1.2 million. In a name that trades 30 million shares a day, buying 600,000 shares over an afternoon is background noise. In one that trades 800,000 shares a day, it is the day's tape.

The same loop runs backward with equal force. When price falls away from the strikes, delta drops and those hedge shares have to be sold. Nothing in the mechanism makes the up-move durable — the buying that manufactured it becomes selling on the other side.

Step by step, the loop looks like this:

• Price rises toward a cluster of strikes carrying heavy call open interest.
• Delta on those calls rises, leaving anyone short them under-hedged — short delta against their book.
• Their buying can lift the price further.
• A higher price raises delta again, and gamma is largest for the strikes price is currently passing through, so each increment of the move demands a larger hedge adjustment than the last.
• New call buyers, drawn in by the move, add open interest that also has to be hedged.
• The loop stalls or reverses as soon as price stops rising, because the same mechanism sells the hedge back down.

Conditions that make a squeeze more plausible

Structure is the part you can inspect in advance. None of it is a forecast, and no single item below means much on its own — the conditions matter because they have to coincide.

One condition is deliberately absent from the list, because the public chain does not contain it: whether the dealers are actually net short those calls. That is the load-bearing assumption of the entire mechanism, and it is an assumption, not an observation.

The structural conditions worth checking:

• Call open interest concentrated at a few strikes just above the current price, rather than smeared evenly across the chain.
• Short-dated expirations, where gamma near the money is at its most concentrated and hedging adjustments are most abrupt.
• A small float and modest average daily volume, so hedging demand is material against normal turnover. A mega-cap absorbs the same flow invisibly.
• Rising call volume against existing open interest, which is consistent with the concentration being added to — though volume alone cannot confirm that.
• Thin liquidity in the shares themselves — wide spreads and shallow depth mean each hedge print moves price more.
• A catalyst capable of carrying price into the strike zone in the first place. Structure describes what could happen if price gets there; it says nothing about whether it will.

Gamma squeeze vs short squeeze

These two terms get used interchangeably, and they are different mechanisms running on different fuel.

A short squeeze is a stock-borrow story. Short sellers who need to close — because of margin, rising borrow cost, a recall, or risk limits — buy shares to cover, and their buying pressures the shorts who remain. The relevant inputs are short interest, days-to-cover, borrow fees, and available float.

A gamma squeeze is an options-hedging story. The buyers are not bearish traders capitulating; they are desks buying shares to bring a delta hedge back to neutral. Short interest is irrelevant to the mechanism itself.

They do co-occur, which is why they get conflated. A heavily shorted small float with concentrated near-the-money call open interest has both engines available at once, and from outside the book it is close to impossible to attribute a given move to one rather than the other.

What expiration does to the loop

Expiration is where the mechanism ends rather than climaxes, and it is worth understanding because so much gamma-squeeze discussion clusters around short-dated contracts.

As expiration nears, a near-the-money strike's gamma becomes extreme in a narrow price band. Small moves through the strike can flip delta a long way, which is why hedging around a big expiry can be jumpy in both directions and why price sometimes gravitates toward a heavily traded strike rather than away from it.

Then it stops. Calls that finish in the money are assigned, and the short-call side delivers shares — for a hedged desk, largely the shares already held against the position, so that hedge leaves through delivery rather than through the market. Calls that finish out of the money simply expire, and the hedge shares behind them are sold. Either way the recurring buy-pressure attached to those contracts is gone the following session. A move that depended on it has to find a different source of demand or give ground.

Why gamma squeezes are rare and mostly named in hindsight

Chains that look structurally primed are far more common than squeezes are. Price drifts sideways, the calls decay, hedges unwind gradually, and nobody writes a thread about it. The structural setup is a necessary condition, nowhere near a sufficient one.

Several assumptions have to hold at once for the loop to actually run. Dealers must be net short those calls. They must hedge dynamically in shares rather than offsetting with other options or simply carrying the risk. The position must be large against real liquidity. And price has to reach the strikes at all.

There is also no live confirmation available to anyone outside the hedging desks. You cannot watch dealer hedging happen. What you can watch is price rising on unusual options volume — which is consistent with a gamma squeeze and equally consistent with ordinary buying pressure, an index or ETF rebalance, or a single large buyer. The label usually gets applied afterward, once the shape of the move is already known.

Treat any real-time claim that a gamma squeeze is happening right now the way you would treat any unfalsifiable statement about somebody else's book.

Honest limits: what the public option chain cannot tell you

The chain publishes open interest per strike, daily volume, and greeks. It does not publish who holds which side of each contract. That gap is the ceiling on any tool built from chain data, this one included.

Gamma-exposure (GEX) models bridge the gap with an assumption: calls are dealer-short, puts are dealer-long. It is a defensible average and it is wrong in individual cases, sometimes badly — the covered-call and overwriting flows mentioned earlier are exactly the case that breaks it. When the assumption is wrong, the output points the opposite way from reality while looking exactly as confident as when it is right.

Two data facts are worth carrying with you. Open interest is a settled figure published before the next session opens, so an intraday reading describes the prior session's positioning, not today's. And option volume does not reveal whether a trade opened or closed a position, or whether the initiator was buying or selling — the same contract can be traded back and forth all day by the same participants.

That has a direct consequence for volume-over-open-interest. A V/OI above 1 means the day's volume exceeded a stale open-interest figure. It is a reasonable hint that something new is happening in that contract, and it is not proof that new positions were opened, that they were opened by a buyer, or that the buyer was informed. Tomorrow's open-interest print is the only thing that partly settles the question, and by then the move has happened.

None of this makes the structural read worthless. It makes it a filter that tells you where to look, not a measurement of what is about to happen. Any claim to the latter is describing data the public chain does not contain.

How ConvexRadar shows this — and what it does not

ConvexRadar is contract-level research software, so the structural ingredients described above appear on one scanner row instead of being assembled by hand: gamma, gamma per premium (G/P), delta, open interest, today's volume, volume-over-open-interest, estimated dollar flow, IV rank, bid-ask spread, days to expiration, and a squeeze column. There are filters for expiration window (including 0DTE), minimum gamma per premium, delta band, minimum premium, and inferred print type.

The Squeeze column is a short-dated convexity-pressure estimate built from fresh volume against open interest, gamma per premium, share and contract liquidity, estimated premium traded, and nearby expiration. It is deliberately not a dealer gamma-exposure model, and the name refers to the contract's own convexity — it is not a claim that the underlying has to go up, or a forecast of any kind.

Stated plainly, here is what ConvexRadar does not have: no licensed full options tape, no dark-pool feed, no dealer positioning or GEX derived from real inventory, and no float or short-interest data — so the float and borrow conditions listed above are checks you make elsewhere. The print labels (sweep, block, split, print) are chain-derived proxies computed from volume, open interest, and premium thresholds; they are labelled as proxies in the interface rather than presented as exchange tape, and they cannot establish who initiated a trade or on which side.

If that level of context is useful, /gamma-squeeze-scanner covers the screen itself, /options-flow-scanner covers contract-level pressure more broadly, /unusual-options-activity covers how unusual volume is identified, /0dte-options-scanner covers the short-dated end where gamma concentrates hardest, and /options-education explains the individual fields. ConvexRadar is research software, not financial advice; nothing in it is a prediction, a signal, or a recommendation, and options flow is one research input among many.