What a call wall and a put wall are
Call wall and put wall are the two strikes where options gamma is most concentrated around the current price. The call wall is the strike above it carrying the largest concentration of call gamma; the put wall is the strike below it carrying the largest concentration of put gamma. Together they are read as the upper and lower boundaries of the range the options market is currently structured around, and they get drawn as resistance and support because of the mechanical hedging activity of options market makers, not because of sentiment or chart history. See today's SPX GEX chart with the call wall and put wall for those two levels.
The usual published definitions agree in substance: the call wall is the strike with the most significant overhead call concentration, commonly read as resistance, and the put wall is the strike with the largest net put gamma for a given underlying.
Neither wall is simply the strike with the most open interest: open interest gets weighted by gamma first, so a nearby strike with moderate size routinely outranks a distant strike with more contracts. And neither is a level anyone observes. Both are outputs of a gamma exposure model, inheriting every assumption it makes.
Two definitions of the put wall circulate, and they do not always point at the same strike: the stricter one weights each strike by gamma multiplied by open interest, the looser one just takes the largest put open interest below spot. SquawkFlow uses the first for both walls. When two services disagree about where a wall sits, check which quantity each is ranking before concluding one of them is broken.
For the individual definitions, start with what a call wall means or what a put wall means. This comparison follows the different hedging assumptions behind the two walls and why the usual support and resistance stories do not mirror each other.
The call wall, derived from the delta
The mechanism is worth deriving rather than memorising, because the version circulating on most explainer pages contains a sign error that inverts the conclusion. A call's delta rises from near zero toward one as spot rises. Everything follows from that.
Take a dealer who is short that call. As the index climbs, their position delta becomes more negative, and the offsetting hedge is long underlying, so staying delta-neutral means buying more as price rises. That is a short-gamma book, and buying into strength amplifies the move. It is the machinery behind a gamma squeeze, the opposite of a ceiling.
Now take a dealer who is long that call. Position delta grows as spot rises, the offsetting hedge is short underlying, and staying neutral means selling more into the rally. That is a long-gamma book, and it produces exactly the deceleration people mean when they say the market is running into a wall.
So the call wall story has a precondition most write-ups skip: it acts as resistance only if the hedging dealer is long the calls at that strike. Pages asserting that dealers are short the calls and therefore sell into the rally weld together two common clauses that describe no hedging book that exists.
The convention that resolves it is explicit in the sources. The arithmetic is plain: long calls and long puts have positive gamma; short calls and short puts have negative gamma. FlashAlpha's key levels write-up spells out the positioning the wall assumes: "at the call wall, dealers are modeled as net long calls (long gamma), reflecting structural exposure rather than a claim about the dealer's literal book."
SquawkFlow uses the same convention: call gamma signed positive and put gamma negative from the dealer perspective, which is the same statement as modelling dealers long the calls, and resistance at the call wall is what that implies. The GEX formula walkthrough covers the arithmetic term by term.
Why the put wall does not mirror the call wall
The put wall is usually described as the mirror image of the call wall. The delta math does not mirror, and this is the most commonly repeated error in the subject.
The standard story runs like this: customers buy puts for protection, dealers take the other side and are therefore short those puts, and as price falls toward the strike the dealers must buy the underlying to stay delta-neutral. Mechanical buying appears under the market, the decline slows, the wall holds.
Half of that is right. Customers do predominantly buy index puts, and dealers are predominantly short them. The SqueezeMetrics gamma exposure white paper, the 2017 document that put this framework into circulation, lists it as an explicit assumption: "Put options are bought by investors; sold by market-makers." The hedging direction is the part that is wrong.
A long put has negative delta, and as spot falls toward the strike that delta moves toward −1.00. A short put, which is the dealer's side, has positive delta, so the dealer's position delta becomes more positive as price drops. A delta-neutral desk offsets positive delta by selling the underlying, so the dealer who is short puts sells into the decline, and sells more the further it goes. That is not support but amplification, and it is precisely what short gamma means. The Options Industry Council states the underlying fact plainly in its gamma primer: "Long options, either calls or puts, always yield positive Gamma. Short calls and short puts will have negative Gamma."
This matters for internal consistency, not just pedantry. The same vendors who describe the put wall as a dealer-buying floor also describe the region below the zero-gamma flip as dealer-selling amplification, and both stories cannot run off the same short-put inventory. For the put wall to produce genuine dip-buying through delta hedging, the desk has to be long the puts, not short them. That happens, but it is a different assumption from the one the calculation is built on.
So why does the put wall hold as often as it does?
Because delta hedging on short puts is not the only flow at that strike. Four mechanisms survive scrutiny.
Hedge exhaustion below the strike. Once price trades through the wall those puts go deep in the money, delta saturates near −1.00, and gamma collapses, so the dealer who was selling all the way down stops needing to sell. Removing a headwind is not the same as adding a bid, but on a chart the two look identical.
Position closing. Put holders who bought protection out of the money reach the strike with most of the convexity they paid for already spent, and when they close the dealer buys back the stock shorted against those puts. Some practitioners treat this as the primary mechanism rather than the delta story: closing all those positions at once mechanically prompts dealer buying.
Vanna. When a test of the wall fails and implied volatility comes back in, put deltas shrink and dealers short those puts buy back part of their hedge. Practitioners call this the bullish side of the vanna effect. It is second-order flow, not the first-order delta hedge, and our piece on vanna and charm covers it.
Distance. A put wall typically sits well below spot, and a level several percent away goes untested on most days for reasons that have nothing to do with options positioning. Any honest hold rate has to account for this before claiming predictive power.
The positioning is an assumption, not a measurement
This is what separates a usable level from a decoration. Open interest is a cleared count of contracts outstanding. It tells you a contract exists, not which side of it a dealer is on, and no exchange publishes dealer inventory.
The convention rests on a structural argument: in index products, institutions buy downside protection and sell upside calls against long portfolios, so the dealer ends up long calls and short puts. It is a reasonable prior, not a fact. A call-positive, put-negative formula built from public data is a simplifying inventory convention, not a rule of option mathematics and not a direct observation of any dealer book.
There is genuine evidence on how often the prior is right. Amaya, Garcia-Ares, Pearson and Vasquez used proprietary exchange trade records, which flag the trading capacity on each side, to reconstruct the actual aggregate options market maker position in every SPX and SPXW series at one-minute frequency from July 2020 through June 2023. Their finding, in 0DTE Index Options and Market Volatility: How Large is Their Impact? (January 2025): "The results show that the gamma of the aggregate OMM position, while typically positive, is often negative."
That sentence is the honest summary of the whole category. The convention is right on average in SPX, and wrong often enough that a wall drawn on a day when dealers are actually short gamma is not merely weak, it points the wrong way.
The SPX call wall and put wall specifically
SPX is where dealer gamma has the best claim to moving the tape, for reasons of scale. The exchange reported that SPX options traded 970.6 million contracts in 2025 at an average daily volume of 3.9 million, with a 0DTE average daily volume record of 2.3 million contracts, "representing 59% of total SPX volume". At a contract multiplier of 100, per the SPX contract specifications, a book that size hedged in the most liquid futures market in the world is capable of being the marginal offer.
Scale cuts both ways, and the 59% figure is the reason. Open interest settles overnight, so a wall built from it describes positioning carried into today. Most of the session's volume trades in contracts that expire that day, and positions opened and closed inside the day never reach an open interest file. Treat both SPX walls as a map of overnight-carried exposure that same-day flow can overwhelm.
One practical consequence: the positions behind the walls do not change during the session. SquawkFlow's SPX levels come from a full-chain snapshot built each morning before the open, across roughly 21,000 listed contracts, and the published morning walls stand for the session. Intraday refreshes repoint spot, the implied range and the 0DTE magnet. A separate repriced read revalues the same settled open interest at a later spot, and because gamma depends on spot, that can move which strike ranks as a wall even though no position has changed. Any page showing a call wall ticking with every trade is showing you something other than settled open interest.
If you trade SPY or ES the SPX walls still apply, because SPY strikes sit near one-tenth of the index level: a 7,750 call wall lands around 775 on SPY. The SPX gamma levels piece covers the rest of the board.
When a call wall holds, and when it fails
Three outcomes, not one. Price can decelerate into the strike and reverse, the textbook case. It can pin near the strike as hedging flow cuts both ways. Or it can break, and breaks tend to be fast, because the dealer selling that was capping the move disappears above the strike and the hedging flow there can invert.
The careful reading treats a wall as a level that can act as a magnet, a resistance reference or an acceleration point depending on who owns the options, time to expiry and new flow. It is not an automatic ceiling.
Walls fail in patterned ways. Scheduled macro such as CPI and FOMC injects directional flow that dwarfs hedging. Thin walls with modest gamma behind them give way to ordinary momentum. Monthly and quarterly expirations retire large blocks of open interest at once, so a wall that framed the range all month can cease to exist the following Monday. And underneath all of it sits the sign assumption: if dealers are short rather than long the calls at that strike, there was never a wall there to break.
A put wall break is more informative than a call wall break. The gamma that was present collapses, and if price is also below the zero-gamma flip you are in the regime where dealer hedging amplifies moves rather than dampening them. The put wall failing and the tape getting faster are the same event described twice.
How the walls shift over time
Neither wall is static. New flow moves them: a large institutional call buyer at a higher strike can pull the call wall upward, and heavy put buying at lower strikes pushes the put wall down. Expiration removes them: as near-term options expire the gamma they contributed disappears, and after monthly OPEX the entire landscape can reset, which is why the first few days after major expirations often feel different. Rolling shifts them: traders moving positions from one expiration or strike to another change where the next morning's walls land.
What a published hold rate actually tells you
None of the mechanism above tells you how much weight to put on today's level, and that is the question that decides position size. That answer comes from a track record, not from theory.
Some level vendors publish hold rates for their SPX walls, typically a headline share of sessions in which the intraday low stayed above the put wall. The limitation is that such a number is unconditional. A put wall 3 percent below spot and one 0.4 percent below spot are entirely different propositions, and blending them produces a figure that describes neither. The distant wall inflates the average; the near wall is the one you were actually going to trade against.
The honest way to use any published hold rate is to ask three questions of it: how far away was the wall when the rate was measured, what counted as a touch, and how many sessions are in the sample. A rate that cannot answer all three is marketing. SquawkFlow's free SPX gamma exposure page publishes the walls themselves with their capture time and the overnight open interest change behind them, and as of August 2026 it deliberately publishes no hold rate, because the one it previously showed failed that test. The Start Here path walks through building your own touch-conditioned count in a levels log, which is the only version of this number whose assumptions you fully know.
Trading with call and put walls
The first read is where the call wall and put wall sit relative to spot. A session that opens between them is inside the range the model draws; one that opens near or beyond a wall is at a level whose behaviour the model does not decide on its own, since a test can resolve as a reversal, a pin or a break.
Three reading notes. First, a wall comes with its distance from spot attached; the level alone is half the information. Second, what matters at a test is the resolution, not the touch. A low that pierces the put wall and closes back above it is a different event from a close below. Third, any hit rate worth weighing is one measured at the distance in question, not an unconditional headline percentage.
None of this makes a wall a barrier. Each is a single strike, derived from settled open interest, resting on an assumption about who owns what that nobody can verify. Used as one input among several it earns its place; used as a place to stand in front of a moving market it does not. Today's SPX call wall and put wall sit on the free SPX gamma exposure chart with the zero-gamma flip, each published with the time it was captured and the settlement open interest behind it, and no hold rate attached.
Checking the walls against what happened
Three records on this site carry the dates behind the levels above, and all three are free to read without an account.
The session record is the resolved account of each closed session: the levels that were published that morning, and where the index actually finished relative to them. It is a record of what was published and what happened, not a hit rate, and it is the document to read if you want to build the touch-conditioned count this article argues for rather than take anyone's headline percentage.
The overnight open interest change is the net change in contracts open across the chain, one settlement to the next, from SquawkFlow's own archive of daily settlements. Walls move when the open interest behind them moves, so this is where a wall relocating overnight becomes visible as a cause rather than a surprise.
The corrections log lists every published figure that was later amended, with the date of both. A source that never corrects anything is not a source that is never wrong.
Each of those pages has a plain-text copy at the same address with .md on the end, which is the version to cite or hand to a model.
Educational content, not financial advice. See our risk disclosure.