Gamma describes how an option’s delta changes as the underlying moves. Aggregating a model of option positioning gives net gamma exposure. It estimates how much hedging might respond to a small market move; it does not measure actual cash traded, future returns, or market direction.

The lower plot uses the public SqueezeMetrics GEX observations as published. Positive gamma is associated with hedging that resists a move, while negative gamma can amplify an upward or downward move. Crossing zero is a change in the estimate’s sign, not an automatic entry signal.

Units matter. The default is billions of dollars of hedge-notional change per one SPX index point. A reading of +$1B means roughly $1 billion of modeled exposure adjustment for a one-point move. The alternative view is GEX per 1% move = GEX per point × SPX close × 0.01. Both views divide dollars by one billion. This is a local gamma approximation; large price changes also change gamma itself.

The optional orange dotted line is a trailing five-observation arithmetic average in the selected units. The rank card compares the selected daily reading with the preceding 252 observations, excluding the selected day; it is a historical rank, not a probability.

The on-canvas GEX profile is a separate, dated SPX/SPXW chain snapshot. Its horizontal bars estimate gamma exposure by strike, using Cboe delayed option data, reported gamma, outstanding open interest and the 100-dollar SPX contract multiplier. Calls contribute positive gamma and puts negative gamma. Dollar exposure per SPX point is Σ signed gamma × OI × 100 × SPX spot. This gross-OI assumption does not identify actual dealer ownership, and the profile total need not equal SqueezeMetrics’ historical reading.

The one-sided teal profile extends left from the right edge of the price plot. The separate signed canvas shows positive net GEX in green left of zero and negative net GEX in red right of zero. Both use exactly the same Granularity, expiry filter, units and visible strike range, including Automatic grouping. Bin values sum calls minus puts before taking the magnitude for the teal overlay. The default is 10-point bins. Every strike draws each available strike separately, without interpolation. Preset intervals and custom widths from 1 to 1,000 points are available, and your selection is saved in this browser. Changing bins redistributes the strike profile; it does not change the historical daily aggregate below SPX. Both profile widths scale to their largest visible magnitude. The faint guide identifies the largest individual-strike GEX magnitude within the visible price range. It is a positioning concentration, not guaranteed support or resistance.

SPX price display. The default uses traditional green/red daily candles from Yahoo Finance’s SPX daily OHLC. Each candle covers one full cash session, including early-close days. Trading sessions are equally spaced: overnight, weekend and holiday gaps are removed, and candle bodies use the full session width. Actual open/high/low/close values and price gaps are retained. Space on the right keeps the latest candles clear of the teal GEX profile. Only completed sessions are admitted, with a 20-minute publication allowance. The archived candle coverage starts in October 2024 and is retained as new bars arrive. If the visible range starts earlier, SPX switches to SqueezeMetrics’ white daily closing-price line, returning to candles when you return to the covered period. The price-panel label identifies the active mode. Historical daily GEX remains daily. Downloads run hourly during U.S. weekdays. Previously validated daily bars are retained if a later response omits them, with gaps listed in the download’s metadata; missing new observations and invalid feeds preserve the last valid snapshot.

Use Focus levels to inspect the latest profile around current price. Expiry filters use calendar days from the chain timestamp. The profile stays at that timestamp when you move over old prices: the daily historical aggregate cannot reconstruct old strike profiles. Expired contracts are excluded, including same-day AM settlement and PM session closes/early closes from the exchange calendar. Open interest generally reflects prior-session positions; this is a delayed snapshot rather than live trade flow. Profile refreshes and failures are tracked independently from the historical series.

Source file: SqueezeMetrics DIX/GEX CSV. We use its date, S&P 500 closing price and GEX columns. DIX is a different indicator and is not plotted. All available observations are retained, without smoothing the daily line, calibrating it to another vendor, or filling missing values with zero. Dates identify market sessions; daily points align with the session close on the intraday time axis (16:00 New York, or the known early close), rather than implying a midnight price. Publication can occur later.

SqueezeMetrics’ original gamma-exposure paper describes the call-positive/put-negative positioning hypothesis. Its Implied Order Book research describes dollar GEX per SPX point. This page displays the provider’s published aggregate; it does not reconstruct its historical option books. Provider methodology and later corrections can affect comparisons across time.

Geeks of Finance describes its total net GEX across strikes and expirations. Matching its exact historical numbers requires its own data and calculation rules. The free source used here is a separate estimate. We have not spliced it into your existing chain-based chart.

Your Net GEX All Expirations page is a strike-level snapshot from an open-interest proxy. That model uses Σ signed gamma × open interest × 100 × spot² × 0.01 for dollars per 1% move, with calls positive and puts negative. Open interest cannot identify every dealer’s actual position. SPY option exposure is not a substitute for SPX option exposure, even after a unit conversion.

Automatic checks run three times daily and admit only completed daily observations. Failed, stale, malformed or truncated downloads preserve the last valid snapshot and show a refresh warning. Full source files are archived; accepted upstream revisions are counted. A historical download is not proof that every row was available at the original close, so this page is not a trading backtest. See the provider’s data documentation.