Two numbers, two different questions
Traders often put options flow and gamma exposure (GEX) side by side and ask which one is "right." That framing misses what each one is. They are not competing forecasts of the same thing. They answer two different questions about the options market:
- Options flow asks: what traded today? It is a running record of transactions: which contracts changed hands, how many, at what price, and where in the bid/ask spread each trade landed.
- GEX asks: what is already open, and how might dealers hedge it? It is a model built on open interest, the count of contracts outstanding at the last settlement, combined with each strike's gamma and an assumption about who holds which side.
One is activity, the other is inventory. A busy trading session can leave open interest almost unchanged if most trades close existing positions, and a quiet session can still sit on top of a very large existing book. Keeping that distinction straight resolves most of the confusion people run into when the two seem to disagree.
What options flow actually measures
Flow starts with the trade record. Each transaction has a symbol, strike, expiry, size and price. What it does not have is a label saying whether the buyer or the seller started the trade. That side has to be inferred, usually by comparing the trade price to the quote at that moment: a fill at or near the ask is classed as buyer-initiated, a fill near the bid as seller-initiated.
That inference is imperfect, and the research on it is specific. Using an exchange dataset that recorded the true direction of each trade, Savickas and Wilson found that the standard rules signed option trades correctly 83% (quote rule), 80% (Lee-Ready), 77% (Ellis-Michaely-O'Hara) and 59% (tick rule) of classifiable trades, and that the legs of complex index trades not executed through the exchange's automated retail system were misclassified close to half the time by every method (Savickas and Wilson, Journal of Financial and Quantitative Analysis, 2003). The Lee-Ready rule itself was designed for stocks, not options (Lee and Ready, Journal of Finance, 1991).
So every "bullish call buying" label is a probability statement, not an observation. Two more limits sit on top of the side problem:
- Opening and closing look identical. A trader buying calls to close a short position produces the same record as a trader opening a new long. You only learn which it was after the next open interest update.
- Hedges and spreads hide intent. A large call purchase may be one leg of a spread, a hedge against a short stock position, or a covered-call buyback. The single trade does not tell you.
That is why flow is most useful in aggregate (net premium by strike or expiry across a session) and least useful as a single headline trade. Our guide to what options flow is covers how flow is grouped and filtered, and reading options flow for direction goes through the habits that keep a single trade from being over-read.
What GEX actually measures
Gamma exposure takes the open interest at each strike, multiplies it by that option's gamma (how fast delta changes as the underlying moves) and by the contract multiplier, and adds the results across the chain. The concept and the arithmetic are set out in what gamma exposure is.
Two inputs make GEX a model rather than a measurement:
- The sign is an assumption. The usual convention treats customers as net long puts and net short calls (often because of call overwriting), so dealers are assumed short puts and long calls. That makes call gamma positive and put gamma negative from the dealer's side. Nobody observes the dealer's actual book; the convention is a reasonable average, and it can be wrong at individual strikes. The dealer positioning guide explains where the assumption comes from and where it breaks.
- Open interest is daily. Outstanding contracts are counted once per session, overnight. Spot price and implied volatility move during the day, so the gamma of each strike can be repriced, but the positions underneath are as of the prior close. Contracts opened and closed within the same session never appear in that count at all. The timing is covered in detail in when open interest updates.
So GEX describes the shape of an existing book under a stated convention. It is a map of where hedging pressure might sit, not a record of anything that happened today.
Why the mechanics connect them
The two numbers are linked by hedging. When a dealer takes the other side of a customer's option trade, it typically hedges the delta in the underlying. As the price moves, gamma changes that delta, and the dealer rebalances. If dealers are net long gamma, rebalancing means selling into rallies and buying into dips, which tends to damp moves. If they are net short gamma, rebalancing goes with the move.
This is not only a trader's story. Ni, Pearson, Poteshman and White report evidence that option market makers' hedge rebalancing affects stock return volatility and the probability of large stock price moves, a channel separate from any information in the option trades themselves (Review of Financial Studies, 2021).
Flow feeds the book; GEX describes the book. Today's flow becomes tomorrow's open interest (net of closes), and that is the open interest the next GEX estimate is built on. The two are not rivals. They are the same market seen at two speeds.
Where flow and GEX agree
The readings line up when today's activity reinforces the existing structure. A few common patterns:
- Put buying into a large put-heavy strike. If flow shows sustained put demand at strikes that already carry large put open interest, the activity and the inventory point the same way: more downside protection at a level the market already cares about.
- Call selling near a call-heavy strike. Overwriting flow at a strike that is already a large concentration of call open interest (a call wall in GEX terms) is consistent with the usual sign convention, so the model's assumption looks better supported there.
- Quiet flow, stable map. When flow is light and spread across many strikes, the prior session's GEX shape is less likely to have changed much overnight.
Agreement does not make an outcome more certain. It means two different inputs are describing the same structure, and neither is contradicting the other's assumptions.
Where they conflict, and what the conflict means
The more informative case is disagreement. The standard example is heavy call buying into a call wall.
GEX, on the usual convention, treats a large call strike as a place where dealers are long gamma and their hedging tends to slow the price as it approaches. Now suppose flow shows aggressive buying of those same calls. If those trades open new long positions, customers are now long calls that dealers sold. Dealers at that strike may be less long gamma than the model assumes, or even short gamma there, and the damping story weakens.
Other conflicts follow the same logic:
- Flow says opening, GEX still uses yesterday's book. A large one-sided session can shift positioning before the next open interest count. The GEX map you are looking at may describe a book that no longer exists in that form.
- Flow is mostly closing. A burst of activity at a strike that ends with open interest falling the next morning was largely positions coming off, not new exposure. Flow looked dramatic; the map barely moved.
- Flow is spreads. Heavy volume at one strike can be the hedged leg of a structure whose other leg sits elsewhere. Read alone, it can make a strike look more important than it is.
When the two clash, the honest reading is that one of the model's inputs (the sign convention, or the age of the open interest) is under strain at that strike. That is useful information. It is not a signal to act on.
A practical way to read both
A sequence that keeps each number in its lane:
- Start with the map. Note the as-of date of the open interest behind the GEX levels. Identify the strikes with the largest concentrations and the approximate level where net gamma changes sign.
- Then read the day's activity. Look at flow in aggregate by strike and expiry. Ask whether it is concentrated at the strikes the map highlights, and whether it looks like opening or closing activity.
- Look for tension. Where flow and the convention disagree, mark that strike as lower confidence in the GEX model, not as a trade idea.
- Check the next morning. The following session's open interest change tells you how much of yesterday's flow became positions. That is the only way to confirm what flow could only suggest.
SquawkFlow shows both views: its GEX page states the settlement date of the open interest it uses, and its options flow pages summarize aggregate session activity rather than individual executions. Neither is offered as a prediction.
The short version
- Options flow is today's transactions, with the trade side inferred and the open-or-close question unanswered until the next count.
- GEX is a model of the existing book, built on daily open interest and an assumed dealer sign.
- They connect through dealer hedging: flow changes the book that GEX describes.
- Agreement means two inputs describe the same structure. Conflict means a model assumption is under strain at that strike.
- Neither is a signal alone, and reading them together mostly tells you how much to trust each one.
Educational content, not financial advice. See our risk disclosure.