What GEX measures

Gamma measures how an option’s delta changes per $1.00 underlying move. GEX combines gamma, contract counts, contract size and an assumed position sign.

The result depends on the selected contracts, inputs and timestamp. It models exposure; it does not reveal dealer inventory.

Sources: Options Industry Council (OCC): Gamma.

REAL SESSION · HISTORICAL CASE STUDY

A gamma flip under review

Compare QQQ’s completed closes with its short-horizon flip.

QQQ · Sep 9, 20261-minute source bars15-minute delayed model source
Full-session retrospectiveCompleted through 4:00 PM ETModel first recorded at 2:44 AM ET.
Full-session view
3:55 PM ET–4:00 PM ETO 716.01H 716.37L 715.53C 716.28
QQQ, 2026-09-09. 5-minute historical candles completed by 4:00 PM ET. Visible 9:30 AM ET to 4:00 PM ET. Later source minutes are hidden.713.31715.16717.02718.87720.72Call wall: $720.00 · Native primary display · first recorded 2:44 AM ETCall wall 720.00Gamma flip · ≤1 DTE: $717.91 · Native intraday: at most 1 DTE · first recorded 2:44 AM ETFlip ≤1 DTE 717.91Gamma flip · ≤7 DTE: $717.98 · Native weekly: at most 7 DTE · first recorded 2:44 AM ETFlip ≤7 DTE 717.989:30 AM ET12:45 PM ET4:00 PM ET
Call wallGamma flip
Outside the price range shown: Put wall $705.00.
Review the historical session
390 / 390
MOMENTS TO INSPECT

Choose a checkpoint for 1-minute detail. Later prices stay hidden.

HOW TO USE WHAT WAS VISIBLE

A flip crossing alone is not an entry.

Historical review. Plans are hypothetical; fills, delivered signals and a trading edge are not proved.

Context
Preopen walls: call $720.00, put $705.00. Flips: ≤1 DTE $717.91; ≤7 DTE $717.98. ≤1 DTE is not 0DTE-only. The 15-minute-delayed model was stale; alerts were suppressed. Pause until model timing and a fresh broker quote are verified. No order was authorized.
Wait for
The 09:55–09:59 block closed $718.88 above the short-horizon flip, known at 10:00 ET. This is an observation, not an automatic entry.
Execution
No order is specified or executed. Define entry, invalidation and risk separately using fresh quotes and acceptable spreads. Shorts require borrow; size must fit buying power. No option or futures trade is inferred.
Invalidation
A completed close back below the same flip challenges the hold-above idea. Keep the horizon fixed.
Management
Follow completed closes. Wait for any separately required wall touch. If no condition qualifies, staying out is valid.
What followed

Above at 09:56 ET; back below at 10:17. The session had 12 strict adjacent-close crossings; neither primary wall was touched. No entry, alert, fill, edge or P&L is established.

Source, timestamps & selection
Candles
Databento XNAS.ITCH · venue OHLCV, not consolidated SIP volume. These are venue bars, not a consolidated market tape.
Model source
Tradier sandbox · original archived native model. Source delay: 900 seconds.
Model built
2026-09-09T06:40:02.007023+00:00
First recorded
2026-09-09T06:44:08.998656+00:00
Open interest timestamp
Unknown. No point-in-time open-interest clock was supplied.
Evidence limit
A local producer cycle recorded this model before the cash open. The producer log is mutable and has no independently sealed receipt. The first and selected records report stale=true and alerts_suppressed=true; no enabled alert, trader broadcast or execution is proved. Exact OI availability is unknown. The chart uses a different price provider. Bar-end times are the earliest theoretical observation times after a minute closes, not proven client delivery times.
Why this day
Four editorial examples selected after reviewing all 12 SPY/QQQ/IWM sessions from September 8–11, 2026. Hindsight selection is explicit. Untouched levels, recrossings and the invalidation-first case remain visible; no hit rate or predictive performance is claimed.

The producer record is mutable. “First recorded” describes the recorded archive; it is not proof that a trader received the same levels at that instant.

Archive fingerprints
  • resource-case-study-census.jsonaf9379d9eb6802044356957fd6cf9a96b203eb57bddafd844c2d56ebf679de85
  • cycles.parquet774e0b67dd965b652f12ef1f7a5e90a49a81ba14871c05f2851144a24731b7a2
  • contracts-QQQ.parquet0ad36606577cb7693b97c4538fcfd12b70c99d7e1e2831c7a10d427fc86822db
  • QQQ-preview.json61590fe909b65c86fcac3ef1642b45a9032f2cc6558cff5d812cd4d44520faa7
Explore the concept with a simplified model
TRY THE MODEL

What changes the exposure?

Move the contract count, then change the assumed position sign. The option’s gamma stays the same.

Illustrative example · not market data
DOLLAR GEX PER 1% UNDERLYING MOVE+$200,000.00
−$1.00M0+$1.00M
Spot $100.00Gamma 0.02Multiplier 100 shares
1,000 contracts
Assumed position sign

0.02 × 1,000 × 100 × $100.00² × 0.01 × (+1)

A positive assumed position sign gives positive modeled gamma. That does not mean the next price move will be up.

Gamma is held constant. The result approximates the dollar value of a delta-equivalent share change for a 1% move, valued at the starting price. Open interest alone does not reveal the position sign.

The dollar GEX formula

Signed GEX = gamma × open interest × multiplier × price² × 0.01 × assumed sign. Add the contract contributions to get net GEX.

This convention expresses the dollar change in delta-equivalent shares for a 1% underlying move, valued at the starting price. It is not premium, volume or P&L. A per-$1.00 measure uses different scaling.

Standard ETF contracts generally represent 100 shares. Verify adjusted deliverables or exclude them.

Sources: Cboe: Equity Options Product Specifications; Cboe DataShop: Frequently Asked Questions.

A quick example

Fictional ETF: price $100.00, gamma 0.02, open interest 1,000, multiplier 100. A 1% move is $1.00: 0.02 × 1,000 × 100 × $1.00 = 2,000 delta-equivalent shares, worth $200,000.00.

The assumed sign makes that +$200,000.00 or −$200,000.00 per 1% move. Contract counts do not establish dealer direction. This illustration holds gamma constant; actual gamma changes with price, time and volatility.

Positive does not mean bullish

A hedged long-gamma position may sell rises and buy falls; a short-gamma position may do the reverse. Actual price effects depend on inventory, hedging and liquidity.

Long calls and long puts both have positive gamma. Assigning puts a negative sign is an inventory assumption. Cboe’s SPX 0DTE study also distinguishes gross activity from net market-maker risk; its finding is sample-specific.

Sources: Options Industry Council (OCC): Gamma; Cboe — Mandy Xu: Much Ado About 0DTEs: Evaluating the Market Impact of SPX 0DTE Options.

Read it beside price

Use concentrations, sign changes and expiry differences as reference points.

  • Compare price with exposure concentrations.
  • Compare the nearest expiry with a wider window.
  • Keep the model timestamp visible.

How this sits in a plan you already have

Prism plots modeled exposure next to price so you can locate entry and invalidation against the current book. It does not choose the trade, predict direction, or produce a return.

  • Wait for price at an exposure concentration instead of chasing through the middle of the range.
  • Place the stop beyond that structure instead of an arbitrary tick count.
  • Recheck the expiry window before treating one concentration as the only one that matters.
  • If the data clock is not current, stand down. A demo or stale book is not a live book.

What remains unknown

Open interest does not show who owns each contract, when it opened or how it is hedged. A level cannot guarantee a reversal, volatility or direction.

Check price structure, liquidity and event risk before applying your own risk plan.

Sources and editorial notes

Read the primary references and editorial policy

Published by GEXfreak. Prism by GEXfreak is the product described in these guides. Definitions reference the primary sources below; examples and tool-selection checklists are GEXfreak’s own illustrations. This is product education, not independent investment research.

These sources explain options mechanics and data. They do not endorse Prism or validate the predictive value of its levels. Undated pages are marked below; the checked date is when the source was reviewed.

For the product’s calculation and data boundaries, read Prism’s methodology.

All resources
IWMSep 11, 2026Historical
One wall. Repeated tests.
Put wall 290.00
IWM, 2026-09-11. 1-minute historical candles completed by 4:00 PM ET. Visible 9:35 AM ET to 10:15 AM ET. Later source minutes are hidden.9:35 AM9:55 AM10:15 AM
1m closes · XNAS · recorded model
QQQSep 11, 2026Historical
Above, then back below
Flip ≤1 DTE 716.60
QQQ, 2026-09-11. 1-minute historical candles completed by 4:00 PM ET. Visible 9:45 AM ET to 10:10 AM ET. Later source minutes are hidden.9:45 AM9:57 AM10:10 AM
1m candles · XNAS · recorded model
QQQSep 11, 2026Historical
XNAS prices · Tradier model
Model built 3:03 AM ETFirst recorded 3:08 AM ET
QQQ, 2026-09-11. 15-minute historical candles completed by 4:00 PM ET. Visible 9:30 AM ET to 4:00 PM ET. Later source minutes are hidden.9:30 AM12:45 PM4:00 PM
15m model delay · OI time unknown

Open the paid workspace.

Membership is required for live SPY, QQQ and IWM books. The chart on this page is a labeled demo or a recorded session.

Demo values are fictional. Create an account to save your layout; review current access conditions on the pricing page.