Can You Filter GEX by Moneyness, Trade Intent, or Strip Out Spreads? What Actually Helps

Here is the request, near-verbatim, the way it usually lands in our inbox - a thoughtful options trader building a dashboard on the exposure endpoints:

It is a sharp request, and the instinct behind it is good. Let's take the three asks in the order of how useful - and how feasible - they actually are.

1. Moneyness filtering - you can do this today

This is the most feasible of the three, and the good news is you don't have to wait for us: the data to do it is already in the response you're pulling. The /v1/exposure/gex endpoint returns, for every strike, call_gex and put_gex split out separately, plus the underlying_price. That split is the whole game.

There is one subtlety that trips people up, and it's the reason "OTM only" is not the same as "strikes above spot." Moneyness is a per-leg property, not a per-strike property. At any given strike, the call and the put have opposite moneyness:

• A call is out-of-the-money when its strike is above spot.
• A put is out-of-the-money when its strike is below spot.

So an OTM-only gamma reading keeps the call contribution from the upper half of the chain and the put contribution from the lower half:

otm_net_gex = Σ call_gex[K]  for K > spot
            + Σ put_gex[K]   for K < spot

itm_net_gex = Σ call_gex[K]  for K < spot
            + Σ put_gex[K]   for K > spot

atm_net_gex = Σ net_gex[K]   for |K - spot| / spot ≤ band   (e.g. band = 0.02)

In Python, against the live response, that is a few lines:

r    = client.get("/v1/exposure/gex/SPY").json()
spot = r["underlying_price"]

otm_gex = sum(s["call_gex"] for s in r["strikes"] if s["strike"] > spot) \
        + sum(s["put_gex"]  for s in r["strikes"] if s["strike"] < spot)

# call wall / put wall on the OTM-only set are just argmax/argmin:
otm = [s for s in r["strikes"]
       if (s["strike"] > spot and s["call_gex"]) or (s["strike"] < spot and s["put_gex"])]

What is trivial client-side: a moneyness-filtered net GEX/DEX (one filtered sum), the call wall / put wall (argmax over the filtered strikes), and the regime sign. The one part that takes an extra step is the gamma flip - it's the zero-crossing of the cumulative net-GEX curve against spot, so it's a small loop rather than a one-liner, but it's fully derivable from the same per-strike data.

Because /v1/exposure/gex is available from the Free tier up, this OTM read is open to essentially every user right now. A native moneyness= parameter is on our radar mainly so you don't have to re-derive the flip each time - and, more importantly, so the same filter can be applied to the live flow GEX, where reconstructing it by hand is much harder.

2. Opening vs closing - already baked into the flow surface

This one sounds like it requires a brand-new feature, but it's largely already shipped - just on a different surface than the settled exposure endpoints. The key is that open interest alone cannot tell you opening from closing - settled OI is a static net count with no provenance. What can tell you is the trade tape, and that's exactly what our effective-OI methodology is built on.

The /v1/flow/* suite recomputes GEX, DEX, walls, gamma flip, and max pain off effective OI: the morning's settled open interest plus an intraday simulator that side-classifies every print. A buy that opens new positioning builds OI; closing flow works it back down. That distinction - new directional intent versus position management - is precisely the "opening vs closing" signal, and it updates through the session rather than sitting frozen at the morning print.

Concretely:

• /v1/flow/summary/{symbol} returns flow_direction, intraday_oi_delta, live_gex, and flow_gex_pct_shift - the headline "how far has today's flow moved positioning" read. (Growth+)
• /v1/flow/levels/{symbol} gives the live flip and walls after the intraday flow is folded in. (Growth+)
• The scored flow-signals tape goes further, tagging each print with an explicit opening/closing bias, aggressor, and conviction score. (Alpha+)

So if the goal is "show me where today's new directional flow is concentrated," the answer is to read the flow surface, not to ask the settled exposure endpoint to do something its input data can't support.

3. Excluding multi-leg spreads - the trap

This is the ask we'd gently push back on, because the intuition - "spreads and hedges dilute the directional read" - quietly assumes GEX walls are a sentiment gauge. They aren't. Walls and flip levels are a map of dealer hedging mechanics.

A dealer has to hedge their net gamma regardless of how the customer assembled the position. A short-gamma exposure that came from a vertical spread produces exactly the same hedging flow as one from an outright. So stripping spread legs out of the walls would not "clean up" the signal - it would understate the real hedging pressure and move your levels away from where the market actually reacts.

It's worse than neutral on index and ETF names, which is where most GEX dashboards point. There, a large share of the most important strikes - the big collar, overwrite, and calendar levels - are structured positions. "Exclude spreads" risks deleting precisely the walls that matter most, handing you a cleaner-looking but less predictive chart. (For why decomposing put and call structure is its own can of worms, see our piece on directional VRP and put/call decomposition.)

And mechanically, it isn't even recoverable from the settled data: open interest carries no spread provenance. The only place you could attempt it is the trade tape, where complex orders can sometimes be detected - and even there, real multi-leg leg-by-leg decomposition is genuinely hard.

The three asks at a glance

Frequently Asked Questions

Related Reading

• Effective Open Interest: How We Compute Live GEX From Flow
• Flow vs Exposure: Which FlashAlpha GEX Endpoint Should You Use?
• Call Wall, Put Wall, Gamma Flip: Options Levels Explained
• Gamma Exposure by Strike: How to Read a GEX Chart