Part VIII · The whole market · Chapter 8C

The map

Gamma exposure by strike and expiry

1 · From one number to a map

Chapter 8A added every line of the option chain into one number. A single number answers one question: are dealers, as a group, long or short gamma? It hides where their gamma sits. Two books with the same total behave differently if one holds its gamma at the strike next to spot and the other spreads it over strikes 200 points away.

Vendors therefore spread the total back out. There are three levels of aggregation, and every display shows one or more of them:

• A cell: one strike and one expiry, the dealers' GEX from the calls and puts at that strike and date.
• A strike profile: one strike summed across all expiries, usually drawn as bars against strike.
• A total: every strike and every expiry, the GEX of Chapter 8A.

A GEX heatmap draws the cells as a grid, with strikes as rows and expiries as columns, and colors each cell by its value. Vendors call a cell a node, and most keep the word for the large ones.

One cell is Chapter 8A's formula applied to one strike and one expiry. The worked example takes the largest cell of this chapter's first figure: the 6,000 strike, expiring the same day, with spot at 6,000.

2 · How to read a map

Every heatmap has the same few elements, and each hides a choice the vendor made:

• Rows are strikes, usually with high strikes at the top, so up on the page is up in price.
• Columns are expiries, near to far, often with the same-day column first.
• The hue of a cell is its sign: whether hedging there trades against a move (dealers long gamma) or with it (dealers short gamma). The sign comes from the vendor's convention, not from data (Chapter 8B).
• The depth of a cell is its size relative to the largest cell on the chosen scale. A "big node" is big against the biggest cell displayed, not against the market's liquidity.
• Lines mark spot and the gamma flip, where total GEX changes sign, as the next section explains.

Units come first. This book states GEX in dollars of delta per 1% move. Vendors use at least four units: shares per $1 of spot, dollars per index point, dollars per 1% move, and share or futures equivalents. Some state none. A figure in dollars per point converts to dollars per 1% by multiplying by $0.01\,S$, so at SPX 6,000, $10 billion per 1% is about $167 million per point. Divided by one E-mini contract, $50 times the index, it is about 33,000 contracts per 1% move. A number without its unit cannot be set against anything.

The figure below draws a made-up SPX chain this way: strikes every 25 points from 5,700 to 6,300, against ten expiries from the same day to 45 days. Each cell is $\Gamma \cdot \text{OI} \cdot 100 \cdot S^2 \cdot 0.01$, signed by the chosen convention. Green marks a cell where hedging opposes a move and rose one where it joins a move. The right panel adds each strike across all expiries, rings number the three largest cells, and the dashed line is the flip. The open interest is fixed at a spot of 6,000 and never changes, whatever you do.

Three sliders set spot (5,850 to 6,150), implied volatility (8% to 40%) and the hours to the close (0.25 to 6.5). Three switches set the sign convention, what the cells show (GEX, size only or VEX), and whether colors scale to the whole map or to each expiry. Read the first state first. Then run the hours to the close down to 0.25, and move spot to 6,010 and on to 6,015. Put spot and the hours back, and switch the convention. Last, try size only, VEX and each expiry.

spot

6,000

IV

16%

hours to the close

6.50 h

sign convention JPM-era 0DTE-era

cells GEX size only VEX

color scale whole map each expiry

total GEX = +$5.0B per 1%

largest cell: 6,000 · 0d

same-day share of size = 10%

flip = 5,980

At the start the readouts give total GEX of +$5.0 billion per 1%, the largest cell at 6,000 on the same-day column, a same-day share of size of 10%, and the flip at 5,980. Ring 1 is the cell of the worked example. Rings 2 and 3 are the 5,800 and 5,900 strikes of the 30-day column, where this chain's monthly expiry holds heavy put open interest. Each is about −$0.97 billion. Under the JPM-era convention the puts below spot are rose and the calls above green. The strike profile on the right is rose up to 5,950 and green from 5,975.

Run the hours down and nothing trades, yet the map changes. With 15 minutes left the same-day cell reads +$12.15 billion, 5.1 times its size at the open, and the total climbs to +$14 billion. The same-day share of size rises to 28%. Move spot 10 points to 6,010 and that cell falls to +$1.82 billion, and at 6,015 to +$0.17 billion. The largest cell is then 6,100 on the 30-day column.

Switch the convention and every color flips. The total reads −$5.0 billion, while the rings stay on the same three cells and the flip stays at 5,980. Size only shows what the rings were already ranking: $35 billion of hedging per 1% in total, in one hue, with nothing to say which way any of it trades. The VEX view moves the largest cell to 5,800 on the 30-day column and drops the same-day share of size to 1%. Each expiry recolors every column against its own largest cell. The 1-day column's largest cell, +$0.23 billion, then looks as deep as the same-day cell's +$2.38 billion.

3 · The gamma flip, and what it is not

The level of spot at which total dealer GEX crosses zero is the gamma flip. To find it, a vendor recomputes every cell at a range of hypothetical spot levels, with open interest held fixed, and looks for where the total changes sign. On one side of the flip, dealer hedging amplifies moves, because dealers are short gamma in aggregate. On the other side it damps them, because they are long gamma.

Which side is which depends on the convention. With the JPM-era convention and a book like the figure's, puts below spot and calls above, the amplifying side is below the flip and the damping side above it. The 0DTE-era convention swaps the two sides. The flip itself stays put, because reversing every sign moves no zero.

The flip moves as spot moves, because each strike's gamma depends on how far spot is from it. It also moves with implied volatility and time when nothing trades. Raise the figure's IV from 16% to 35%, and the total turns from +$5.0 billion to −$0.57 billion while the flip climbs from 5,980 to 6,006, above spot. At 40% the flip reaches 6,014. Higher volatility shrinks the gamma at the money, where this book is long calls, and swells it on the wings, where its puts outnumber its calls. The same open interest now reads short gamma.

The flip is not SpotGamma's Volatility Trigger. The Dealer Mechanics Lab also calls the flip the volatility trigger, taking SpotGamma's name for it. SpotGamma itself defines two different levels. Its Zero Gamma is "the modeled spot price where aggregate GEX crosses zero", which is the flip. Its Volatility Trigger is a level that "identifies where dealers have their last major concentration of positive gamma support, which is different from Zero Gamma". SpotGamma calls it "a proprietary SpotGamma level derived from its estimated distribution of gamma rather than a generic zero crossover", which "often sits several points above Zero Gamma".

Other vendors draw their own regime lines. MenthorQ's High Volatility Level "Marks the price where the cumulative gamma curve reverses slope, separating positive from negative gamma regimes". The description does not say whether that is the same calculation as the flip. Nomura's strategy notes, as others quote them, speak of a "Gamma Neutral" line. Before you compare two vendors' regime lines, check that they compute the same thing.

What does theory say happens at the flip? Nothing sudden. In the effective-volatility formula of Chapter 8D, damping turns smoothly into amplification as dealer gamma passes through zero. The formula's special point is where the feedback reaches its critical value, which depends on market depth, not on the zero of gamma. The research this book draws on found no test of the flip as a price level.

So the flip is a regime boundary, not a hard mechanical barrier. Crossing it slowly during a calm session is uneventful. Crossing it during stress, when dealers themselves are under-positioned to absorb flow, is where the destabilization shows up, as the case studies of Chapter 8H show.

4 · The walls

Vendors also mark the strikes where dealer gamma is most concentrated. The call wall is the strike with the largest call gamma, and the put wall the strike with the largest put gamma. SpotGamma, which says it "originated and popularized" the term, defines the call wall as "the strike with the largest net call gamma" and "the upper bound of the expected range". Its put wall is "the strike with the largest net put gamma" and "the lower bound". MenthorQ gives the same definitions the names Call Resistance and Put Support.

The Dealer Mechanics Lab reads a call wall above spot as potential resistance and a put wall below spot as potential support. The market dashboard of Chapter 11A uses them that way.

The mechanics add a condition. A wall slows a move toward it only if dealers are long gamma there, which means long those options. If they are short them, their hedging runs with the move, and a break through the wall accelerates. The sign at a strike is exactly what open interest does not show (Chapter 8B).

SpotGamma's own explainers state the hedge backwards. Its call-wall page says "dealers who sold calls at that strike must sell shares as price approaches". It then says the same dealer is "short gamma" and must "buy shares as the underlying rises toward the strike". Its put-wall page says dealers "who sold puts at that strike must buy shares as price falls toward it". By the sign chain of Chapter 8A, a dealer short a call buys as price rises, and a dealer short a put sells as price falls. Resistance at a call wall needs dealers long those calls, and support at a put wall needs dealers long those puts.

The Edge Book disagrees about trading these levels. It rejects scalping around them, selling at the call wall and buying at the put wall, for retail accounts (Chapters 1G and 10I). It gives three reasons. Vendor models disagree, because each assumes different positioning. The rise of 0DTE options since 2022 changed who holds what. And a Cboe study from 2025 estimated 0DTE market-maker gamma hedging at only about 0.2% of SPX daily liquidity.

That study is Xu's "0DTEs Decoded" (Cboe, May 2, 2025), whose words are "at best, just 0.2% of the SPX daily liquidity". Cboe's 2023 note by the same author put net market-maker 0DTE gamma at $170 million to $670 million through the day, 0.04% to 0.17% of S&P 500 futures liquidity. Both concern 0DTE market-maker hedging only. The walls most maps draw come mostly from longer-dated open interest, so extending the figure to every wall is an inference, one the Edge Book makes.

The Edge Book keeps only the sign of GEX, as low-confidence regime context for sizing other trades (Chapter 10H). Both sources agree that the sign is regime context, not a trade by itself. No academic test of trading a wall was found in this book's research. The only published wall statistics are a vendor's own, and they lack a base rate. Chapter 8J shows what a base rate does to them.

5 · How the map moves when nothing trades

A live map combines three inputs, and only one of them is exact:

1. Yesterday's open interest, with an assumed or estimated sign. The Options Clearing Corporation counts open interest once a day, overnight, and shows it on its site the next morning. The industry-wide file arrives the next business day. It counts contracts, not sides.
2. Today's trades, whose side, and whether they open or close a position, must be inferred by trade classification.
3. Today's Greeks, recomputed from spot, implied volatility and time. Only these are exact.

The third input moves the map on its own. At the money, gamma is proportional to $1/\sqrt{T}$ (Chapter 5C). From 6.5 hours to 1 hour before the close, $\sqrt{6.5/1} = 2.55$, so the same-day cell of the worked example grows from +$2.38 billion to +$6.07 billion with no contract changing hands. Away from the money the opposite happens: as expiry nears, the gamma of a strike 25 points from spot collapses toward zero.

The second figure draws that change for one expiry through one session. Each column is a quarter hour from the open to 3:45 PM and each row a strike, 5 points apart. A cell is the dealers' GEX at that strike and time, green where hedging opposes a move and rose where it joins one, and the black line is spot. The made-up book is long the round strikes near 6,000 and short a few strikes 20 to 30 points from it.

Its time slider picks a quarter hour, marked by a solid box, with a dashed box half an hour earlier. The book is either fixed at the open or changed by the day's trades. In the second case customers buy options at the strike nearest spot after a move of more than 3 points in a quarter hour, and sell them after a quiet one. The readouts split the change in the largest node over the last half hour into what spot and the clock did and what new trades did. Start with the book fixed, drag the time to 3:30 and then 3:45, and then switch on the day's trades.

time

2:30 · spot 5,992

new day

book fixed at the open with the day's trades

color scale whole day each time

largest node: 6,000 at +$0.88B

since 2:00: −$0.09B

Where the change came from: −$0.09B from spot and the clock, $0.00B from new trades.

At the start the time is 2:30 with spot at 5,992. The largest node is 6,000 at +$0.88 billion, down $0.09 billion since 2:00, all of it from spot and the clock. At 3:30 the same node reads +$1.0 billion, about twice its size at 10:00. At 3:45 the largest node jumps to 5,990 at +$0.93 billion. With the book fixed, every change on the map is Greek arithmetic: no position moved.

Switch on the day's trades and the 3:45 reading becomes +$0.49 billion, down $0.46 billion since 3:15. That is +$0.69 billion from spot and the clock and −$1.2 billion from new trades. Vendors' displays show only the net. It cannot tell you that the clock added more than half a billion dollars of gamma while customers took away more than a billion.

A change in a node between two refreshes therefore has three sources: Greek drift at fixed positions, estimated new positions, and, overnight, the reset to the clearinghouse's reconciled count. They call for different readings. None of the vendors in this book's research separates them in its display. The inputs that get the most attention are the least reliable: Chapter 8B gives the evidence, and two findings set the scale. Ardia and Vaudescal (2026), a working paper, find that gamma rebuilt from public open interest correlates at 0.99 with a dealer-inventory benchmark for longer-dated options, but at 0.53 for same-day options. The 0.99 is an upper bound, since the two share a starting point. And Adams, Fontaine and Ornthanalai (2024) find that market makers "match most 0DTE order flows", so their same-day gamma comes mostly from older positions that have aged into expiry.

6 · VEX and charm views

The same grid can show other Greeks. Vanna exposure, or VEX, puts each cell's vanna in place of its gamma: how many dollars of delta dealers gain or lose when implied volatility rises one point. The first figure's VEX view gives it in dollars of dealer delta per volatility point. It looks nothing like the GEX view. Vanna is zero at the money and peaks about one standard deviation of the remaining move away from it, and unlike gamma it does not grow as expiry nears. So the largest cell moves to 5,800 on the 30-day column. The same-day column falls to 1% of the total size.

Vendors say the two views answer different questions. Skylit tells users: "You can choose GEX for today's levels, or VEX for a multi-day view and days when volatility spikes". It states a tenor rule: "For options 0–5 DTE, gamma is dominant; for 7–180 DTE, vanna is dominant". Its slogan is "GEX governs the range; VEX governs the drift". It also warns that a map of the front expiry can hide back-month vanna: "When vol spikes, back-end +VEX creates significant sell flow that's entirely invisible in a front-month gamma map". Chapter 8F checks these rules against Black-Scholes and builds the market-wide vanna and charm this section only names.

One test sets a limit on what the VEX view adds. FlashAlpha, a vendor testing its own product over 1,972 SPY days from 2018 to 2026, found VEX correlated +0.72 with the VIX and +0.76 with at-the-money implied volatility. Once those were controlled for, VEX's correlation with next-day realized volatility was −0.01 (p = 0.77). The vendor calls the test pre-registered. Charm views, the drift of dealers' delta with time alone, come from vendors such as OptionsDepth and Volland. Chapter 8F draws charm into expiry.

7 · Skylit's Heatseeker, a worked example

Skylit sells a heatmap called Heatseeker. Its published vocabulary turns a GEX map into a full set of trading rules, which makes it a good worked example. This book uses it to show theory applied in practice. It is an example, not an endorsement. The sections below describe what the product shows and translate each of its terms into the book's mechanics. They credit what the theory supports, and say what is untested, inconsistent in Skylit's own material, or contradicted by evidence.

Skylit publishes its vocabulary on undated documentation and education pages. The quotations below are its words as those pages give them. Third-party review pages add that it has "over 2,300 members since launching in 2025" and plans at $99.99, $299 and $699 a month. None of that was checked on Skylit's own site.

What the map plots

The layout is the book's: "Rows are strikes, columns are expirations, and each cell is a node". Each column is one expiration, "such as the current expiration, next week, two weeks out, or later periods". Skylit warns that "A major node several expirations out may not be a major node for the current expiration". Each cell carries "a value (positive or negative)", with "positive exposure shown in green to yellow and negative exposure shown in blue to purple".

Around the grid sit a GEX and VEX switch, a star on the largest cell, "the King", and a highlight on "the strike nearest price". Filters set "the number of expirations shown, the number of nodes shown, the selected exposure type, and current versus expanded expiration views". The map "updates live through the session". A Velocity Mode shows node changes over "1 min to 1 day". A Trinity Mode shows SPX, SPY and QQQ side by side.

What Skylit discloses, and what it does not

Skylit's one statement of method is this: "Market data is licensed directly from OPRA, NASDAQ and CBOE. Dealer positioning is derived by Skylit from that data, not reported by the exchanges". The calculations, it says, "run on custom inference models and proprietary intelligence built from years of research into dealer microstructure".

No page found gives the formula, the unit of a cell, the sign rule for calls and puts, the implied volatility surface, or how trades are classified. Its examples are in dollars ("$2 billion", "$200 million") with no unit of move. Its sister product, Flowseeker, "streams options flow with the side it filled on", so the platform does classify trades, but Skylit does not say whether Heatseeker's nodes use that classification. The silence is typical of the trade: a 2026 working paper by Sen finds that commercial GEX services "disclose little of their methodology".

This matters for everything that follows. Every term below depends on a cell's sign and size, and a reader can check neither.

Size and sign: nodes, the King, floors and ceilings

Nodes. Skylit says every node "acts like a magnet in the market, where price is attracted to these zones due to dealer positioning". Yet "these same areas can also act as walls that repel price and create reversals". In the book's terms a node is a cell: dollars of hedging per 1% move at one strike and expiry.

Theory supports a narrow version of magnets and walls. Near expiry, a large position held long by hedgers at one strike pulls price toward it and damps moves around it. The same position held short pushes price away and amplifies moves (Chapter 8E). Away from expiry only a volatility effect remains: price lingers where hedgers are long gamma and moves fast where they are short. Theory does not support nodes as barriers that price bounces off, effects that ignore the sign, or a fixed effect per node. The same node matters more on a thin day than on a deep one. A claim that a node may attract or repel, with no rule for which, fits every outcome and cannot be tested as stated.

The King Node. Skylit defines it plainly: "The King Node is the strike carrying the largest absolute exposure value on the board. Not the largest positive, not the largest negative. The largest absolute". It is "the most influential node on the map", where market makers "typically prefer to settle price by the end of day (EOD) or end of week (EOW)". The rule is to "Treat it as your primary destination target for the session", and "King Nodes are not defined by color. They can appear as yellow or purple".

On the first figure the King Node is ring 1, the same-day 6,000 cell, which sits at spot: the destination is where price already is. With 15 minutes left and spot at 6,015 it jumps to 6,100 on the 30-day column, and at 6,020 to the same-day 6,025 cell. Move spot to 5,950 at the open and the King becomes 5,800 on the 30-day column, 150 points below spot and short gamma under the JPM-era convention.

"Settle price" is pinning. Pinning needs three things: hedgers long gamma at the strike, a position large against the market's depth, and expiry close (Chapter 8E). A purple King, short gamma, should repel. The measured effects are real and small. Closing prices of optionable stocks cluster at strikes on expiration days, altering returns by at least 16.5 basis points per expiry (Ni, Pearson and Poteshman, 2005). S&P 500 futures are pulled to the at-the-money strike on some expiries and pushed away from it on others, at least $115 million of notional per expiry (Golez and Jackwerth, 2012). No test of the largest absolute node as a daily or weekly target was found. SpotGamma's nearest equivalent, Absolute Gamma, "the strike with the largest total gamma", is likewise called "a strong support/resistance level and a magnet" without a test.

Floors and ceilings. In Skylit's words, "A floor is a large node sitting beneath current price. As price declines toward it, dealer hedging from that level becomes active. The larger the node, the more mechanical buying gets activated on approach". And "The strongest floor is always the largest exposure node below spot, not the nearest node, the biggest one". A ceiling is "a large node above spot, and dealers at that level are positioned to sell as price rises into them". Skylit calls "the relationship between spot and its nearest floor and ceiling" "the single most useful piece of context you can have for any intraday trade".

Mechanical buying on approach is what dealers long gamma do below spot. A floor defined by size alone may be purple. Then hedging sells into the fall toward it. On the first figure the strongest floor by Skylit's definition is 5,800 on the 30-day column, −$0.97 billion under the JPM-era convention: an accelerator. Switch the convention and the same cell is +$0.97 billion, a cushion. Its size, and so its rank as a floor, never changed.

Pika and Barney nodes. Here Skylit's words match the book's mechanics exactly: "When dealers are net long gamma, shown as yellow Pika nodes on Heatseeker, their hedging works against the market's direction". And "When dealers are net short gamma, shown as purple Barney nodes on Heatseeker, their hedging runs with the market's direction". In its own phrase, "Positive gamma creates friction, while negative gamma creates fuel". The evidence of Chapter 8G agrees on direction: negative gamma goes with intraday momentum and positive gamma with reversal. Two cautions remain. The color rests on Skylit's unpublished sign rule. And vendors do not share colors: SpotGamma's TRACE map uses purple for positive gamma, the opposite of Skylit's purple.

The Absolute Value Rule. Skylit's central rule is that "size determines influence, regardless of sign". Its example: "A yellow Pika node with $2 billion in exposure has more structural weight than a purple Barney node with $200 million, and the larger node is going to dominate". Its advice is "don't filter by color first" but to "look at the largest nodes across the whole structure, then assess whether they're positive or negative".

The rule is half right. Size sets how much hedging a node generates for each 1% move. Sign sets which way it trades: a $2 billion long-gamma node damps a move, and a $2 billion short-gamma node amplifies the same move. In every model with price impact, the sign decides attraction against repulsion. Size matters only relative to liquidity (Chapter 8D). The measured effects depend on sign: momentum with negative gamma and reversal with positive (Barbon and Buraschi, 2021; Baltussen, Da, Lammers and Martens, 2021). Bitcoin's expiry reversals are strongest when estimated dealer gamma is negative (Weiss and co-authors, 2026). The first figure's size-only view is the map the rule asks you to read. It ranks the same three cells as the GEX view and cannot say which way any of them trades.

A charitable reading exists. Size is the better-measured half of a cell, because open interest is counted while the sign is assumed (Chapter 8B). That is a reason to distrust a map's colors. It is not a reason to expect a node to act the same whatever its sign. No test of the rule was found.

Shape: Gatekeepers, Pika clouds and air pockets

Gatekeepers are "defensive levels that often block price from reaching the King Node", which "act like bouncers at the door of a nightclub". Skylit explains: "A Gatekeeper sits between two larger structural nodes. Its job is not to be the target. Its job is to determine whether price can reach the target cleanly or gets stopped in transit". One that holds means price "stalls in the region it's already in", and a broken one means price "typically accelerates into the cleared zone".

In the book's terms a Gatekeeper is a node between spot and a larger one. If dealers are long gamma there, hedging slows the approach; if short, it speeds a break. Both outcomes are named in advance, so the rule describes what happened rather than predicting it. No test was found.

Pika clouds are "a dense cluster of multiple positive gamma nodes sitting in close proximity on the map". Inside one, Skylit says, "the trend halts, movement becomes inefficient, and price sticks, rotates, and struggles to move cleanly in either direction". A band of strikes where dealers are long gamma is the damping zone of the book's mechanics. There effective volatility is lowest, so price moves least and lingers longest (Chapter 8D). Skylit adds that "A large Pika Cloud below spot is a strong mechanical floor". The barrier half of that sentence goes beyond what theory supports, and no test was found.

Air pockets and liquidity vacuums. An air pocket is "a zone with low or weak GEX exposure. Minimal structural resistance. Minimal structural support. When price enters an air pocket, movement becomes faster with no mechanical drag". A liquidity vacuum is "an extended air pocket. Multiple consecutive zones of low GEX exposure stacked together. In a liquidity vacuum, price doesn't just speed up. It can gap. Levels don't work".

Low local gamma means little hedging per point: less damping where dealers are long, less amplifying where they are short. In the book's feedback formula, gamma near zero leaves volatility near what news alone produces (Chapter 8D). Price moves faster there than in a long-gamma zone and slower than in a short-gamma one. A gap needs something else, thin order books, which the name "liquidity vacuum" borrows but the map does not measure. "Low GEX" is not quantified. On the first figure the thinnest strikes near spot are 5,950 and 5,975, with $0.68 billion and $0.44 billion of size across all expiries. They sit around the flip at 5,980, where the chain's open interest changes from mostly puts to mostly calls.

Time: velocity, lifecycle and reshuffles

Velocity Mode "tracks the rate of change in dealer positioning, measuring how quickly nodes are growing, shrinking, appearing, or disappearing across recent updates". Skylit's formula for a move is "Space (an air pocket) plus fuel (high rate of change) equals acceleration". Its rule: "When Velocity Mode shows a rapidly growing node adjacent to an air pocket, that combination is one of the highest-conviction setups in the framework".

As the second figure showed, node change mixes Greek drift, estimated flow and the overnight reset. Near spot on the same-day column, drift alone makes nodes grow all afternoon. The part that is flow rests on classifying trades, the least reliable input on the map (Chapter 8B). Nothing in Skylit's material separates drift from flow. No test of the setup was found.

The node lifecycle has four stages. A fresh node is "untested and at full strength", and fresh nodes "produce the strongest reactions (~80% probability when spot approaches)". A tested node has had "one price interaction and a portion of its positioning has been absorbed". A delivered node has "already produced a reaction and its mechanical obligation is partially or fully satisfied". A decaying node's "Influence weakens over time with no further price interaction". The summary: "Growth equals intent. Decay equals protection".

The 80% comes with no method, sample, or definition of a reaction. A hit rate without a base rate cannot be read, and Chapter 8J shows how much chance alone produces. Size changes from day to day come from spot, implied volatility and time as well as from positions. A node at the money grows toward expiry and one away from it shrinks, with no new intent at all. Nor is hedging used up by a touch. A long-gamma dealer re-hedges at every pass until the position is closed or expires.

Reshuffles "happen when dealer exposure shifts quickly, often without warning". One "often follows large volatility events like CPI, FOMC, or Power Hour when market makers rebalance risk". A review adds that "power hour liquidation can reshuffle maps entirely". All three have mechanical causes. An event that moves implied volatility moves every cell and the flip, as the first figure's IV slider showed. The last hour is where at-the-money gamma grows most. And each morning the estimate built from yesterday's flow is replaced by the clearinghouse's count.

The late-day window is also where hedging shows in prices. Baltussen, Da, Lammers and Martens (2021) find, across more than 60 futures markets from 1974 to 2020, that the return of the last 30 minutes is predicted by the rest of the day's. They link it to gamma hedging demand. The effect reverts over the following days. So a map at the close need not describe the next session.

Regimes and patterns

Range, trend and whipsaw days. "Positive gamma creates range days, negative gamma creates trend days, and mixed signals create whipsaw days". These are the two regimes of Chapter 8D in vendor language, and the direction is supported. The size on a typical day is modest (Chapter 8G). Much of what the sign says about tomorrow's volatility is already in the VIX (Chapter 8J). For whipsaw days Skylit says "The only edge in a Whipsaw environment comes from fading the extreme ends of ranges rather than chasing directional momentum", a claim nothing tests.

Rug and Reverse Rug. A Rug is "a bearish configuration with positive gamma stacked above spot and negative gamma sitting below spot, with spot positioned below the positive node". Skylit calls it "a ceiling that not only blocks price but pushes it down faster once rejected". The Reverse Rug is "the exact mirror, with negative gamma above spot and positive gamma below spot", which Skylit calls "one of the higher-conviction bullish setups on the map".

Mechanically a Rug damps rallies and amplifies falls, the slow-grind-up, fast-crash-down asymmetry of the JPM-era book in Chapter 8B. The positive node above does not push price down. The short gamma below amplifies whatever fall occurs. The Reverse Rug's short gamma above spot amplifies a rally. Skylit publishes no frequency, hit rate or sample for either.

The Beach Ball is "when price overshoots a node, punches through it, and then reacts. Like pushing a beach ball underwater: the deeper it goes, the stronger the release". Skylit adds: "The overshoot is not the trade. The reaction that follows the overshoot is the trade". You identify it by "the failure to follow through", when "The tape stalls".

Nothing in hedging stores energy that grows with depth. Near expiry, a long-gamma strike pulls price back toward it from either side (Chapter 8E). That pull fades with distance rather than growing. A pattern recognized by a failure to follow through is recognized after the fact. No test was found.

The negative GEX pit is "a price zone where dealer positioning turns fully pro-cyclical, displayed in purple on the heatmap, where the crowd holds a heavy concentration of puts and dealers are net short gamma". It ends in "hedge exhaustion": "When the ball hits the bottom of the pool (ITM saturation), there's nothing left to push against". Its entry rule pairs "a failed continuation attempt into the pit" with "a VIX curl down from its peak".

The mechanism it gestures at is real. Deep in the money a put's gamma is small, so dealers' selling slows. Customers closing puts also make dealers buy back hedges. Neither is shown to release "all at once". The VIX half of the rule is a vanna argument, as is Skylit's "squeeze setup", "when GEX and VEX are pointing in opposing directions". Chapter 8F takes both up, and no test was found.

Trinity Mode is "a combined heatmap view of SPX, SPY, and QQQ displayed simultaneously". Skylit's rule: "A node that stands alone on one instrument is a signal worth noting, not a setup worth sizing into. Full alignment (all three agree) represents high-probability conditions". The three books share many holders, so agreement is partly the same fact seen three times, the trap Chapter 11C calls multicollinearity. SPY's flow is also shaped by wholesalers who internalize retail orders (Chapter 2B), a caveat Skylit does not raise. No test was found.

Where Skylit's own material disagrees with itself

Five inconsistencies appear in Skylit's pages:

1. Two Kings. Skylit's reading guide, documentation and product page define the King Node by "largest absolute exposure value". Two other pages say "The strongest +GEX node across the heatmap is called the King Node, it acts as the primary structural gravity center for price". On the first figure under the JPM-era convention both definitions pick the same-day 6,000 cell. Under the 0DTE-era convention the largest absolute cell is still 6,000, at −$2.38 billion, while the strongest positive cell is 5,800 on the 30-day column: two Kings 200 points apart.
2. Negative gamma above spot. One page says "In a low-VIX environment, the most common configuration is positive GEX sitting below spot as a mechanical floor, with negative GEX above acting as a ceiling". The Reverse Rug page says negative gamma above spot "amplifies any bounce". Short gamma cannot both cap a rally and amplify it. The mechanics side with the second page.
3. Floors by size, mechanics by sign. A floor is the "largest exposure node below spot", which may be purple, yet "mechanical buying gets activated on approach", which only a positive node implies.
4. Vanna's sign. One page states "A dealer who is short options carries positive vanna". Black-Scholes vanna is positive for out-of-the-money calls and negative for out-of-the-money puts, so a dealer short options has positive vanna only on the put wing. The page's own example, short puts and falling implied volatility, is the put case.
5. Refresh and coverage. The product page claims "a one-second refresh", while a third-party listing says its chat bot "streams SPXW, SPY, QQQ, IWM, and VIX maps every 15 seconds". The top plan covers "300+", "3,500+" or "5,000+" tickers, depending on the page.

The claims against the evidence

Skylit states few numbers about performance. One is the ~80% for fresh nodes. The other is an FAQ figure, reported by third-party pages, of 85% to 90% accuracy "during the creator's personal usage", with "your results may vary". Neither comes with a definition of success, a sample or a base rate. Third-party reviews are mixed. One reports the tool "wrong as much as it is right" and blames "treating Heatseeker as a directional signal service". Another claims "~50% win rates" for users who treat large nodes as "buy signals", with no data. Star ratings of 4.8 to 4.9 on a review site measure satisfaction, not performance.

Chapter 8G gives the full review of this evidence, and Chapter 8J turns each rule into a hypothesis you can test.

What Skylit gets right

Much of Skylit's framing matches this book's. Its Pika and Barney language states the sign rule correctly. Its limitations page lists "Not 100% Accurate", "Low volume periods can skew heatmap efficacy", "Map Reshuffles" and "Context Matters", which it glosses as "Don't trade off a single node or isolated event". It says "Heatseeker is contextual intelligence, not a trading signal", and tells readers to "Treat these as zones of interest, not precise levels, and use price confirmation before entering". Its best-practice page asks for three confirmations before acting: a nearby King or Gatekeeper, a "confirmation or rejection wick", and SPX, SPY and QQQ "in agreement".

Its "Charts First" rule, "the workflow rule that chart structure must precede exposure data in every trade decision", is the book's own. "You form the thesis from the chart, then you check Heatseeker to see if the structure underneath supports it or kills it": positioning confirms a thesis and never writes one (Chapter 11A). The tension is inside Skylit's material. Its pattern pages hand out entries, targets and "highest-conviction" setups, while its documentation calls the map context, not a signal.

8 · Other maps, other names

Vendors draw the same quantity with different designs, names and colors:

The deepest split is between fixed levels and moving nodes. SpotGamma sets its levels each morning, and "they hold for the trading session", while its other tools show the day's flow against them. Skylit's nodes move all day. Fixed levels can be logged and tested. Moving nodes answer "what now?" but leave no record of what you saw when you acted, the testing problem of Chapter 8J. Even one vendor's two methods disagree: GEXBot's own example response puts its major positive strike by volume and by open interest at two different strikes, 7,135 and 7,200, at the same moment.

9 · Check yourself