Use Gamma Curve and zero-gamma crossings

Read a repricing experiment

Gamma Curve evaluates the selected chain over hypothetical underlying spot prices. The horizontal positions are possible model inputs, not a time axis and not evidence that trades occurred there. Zero-gamma marks identify roots of the modeled net-gamma curve under those assumptions. A useful first reading asks where current observed spot sits relative to the curve and how many distinct crossings are actually reported.

Configure one experiment

A consistent selection is essential because a changed expiration set can produce a different curve even at the same current price.

1. Add Gamma Curve from Options → Exposure.
2. Choose the underlying, session, trading date and valuation minute.
3. Set the selected expirations, model, perspective and units.
4. Inspect the hypothetical price range and current-spot reference.
5. Hover the curve and read each reported zero-gamma crossing.
6. Match the selection when comparing a crossing with Key Levels.

Distinguish levels from the curve

A strike concentration aggregates exposure at a listed strike. A curve point reprices the selected chain at a hypothetical spot. Their coordinates can happen to coincide without being the same calculation. Multiple roots can exist, and a range can contain no crossing. The largest absolute curve value is also different from the largest strike-expiration cell. Use calculation names so an apparent mismatch does not become a false data-error report.

Worked interpretation: two crossings

Suppose the curve is positive below one crossing, negative between two crossings and positive again above the second. Reporting only the nearest crossing would omit part of the model. Compare current observed spot with both roots, then change the expiry selection to see which maturities contribute to the shape. If a root moves or vanishes, that is a response to the changed experiment. It does not establish an observed change in dealer inventory or a certain price path between the crossings.

Troubleshoot a flat, empty or shifted curve

Check for unavailable chain/Greek inputs, mismatched units and a narrowed expiration selection before interpreting the chart. Compare the valuation minute with the reference spot time. If Key Levels differs, verify it used the same source chain and model rather than a cached older selection. Missing inputs should produce an availability state, not an invented flip at the current price. Include the full curve and controls when asking support to reproduce it.

Model, perspective and units

Exposure is a sensitivity calculation, not an execution total. GEX means gamma exposure, DEX means delta exposure, VEX means vanna exposure, and CEX means charm exposure; charm can also be called CHEX. VEX does not mean vega in these tools. Select the metric before comparing colors or magnitudes. Per 1% move, per $1 move and per unit are distinct choices. Read the returned unit rather than assuming that every displayed dollar amount means money paid for contracts.

Classic OI applies a positioning proxy to dated open interest. Flow-adjusted modeling incorporates eligible classified intraday volume into that proxy. Neither reveals actual dealer inventory or every hedge. Dealer and customer perspectives can reverse signs while retaining the underlying sensitivity calculation. A positive bar is therefore meaningful only with its metric, model, perspective, expiration selection and valuation time. Record those settings together when sharing a view, and match them before expecting two tools to return comparable values.

Avoid turning a hypothetical sweep into a time forecast

Moving left or right on Gamma Curve changes the hypothetical underlying-price input under the selected repricing assumptions. It does not advance the clock through the session. That distinction prevents a common error: reading a curve that becomes negative above current spot as a prediction that gamma will become negative later today. The outcome also depends on which prices, times and eligible model inputs actually occur; the static sweep holds its selected experiment rather than revealing that future sequence.

Use Levels / Time when the question is how reported references changed over observed time, and Exposure Forecast when the question explicitly concerns supported future horizons. Then maintain the same chain selection when comparing them. A root on the sweep can lie between listed strikes, while a ranked wall is tied to a strike aggregate. Both can be valid without matching numerically. Report a suspected calculation issue with the full curve range, each crossing and current-spot reference so support can distinguish root-finding behavior from selection or display-coordinate differences.