# 02 — Concepts (ELI5) The whole project rests on one mental model: every option you trade has a dealer (market maker) on the other side, and that dealer doesn't want directional risk, so they hedge with stock. The hedging is **forced**, **predictable**, and **size-able**. Mapping where the dealer's hedge pressure concentrates = mapping where price has structural support/resistance. --- ## Gamma > **Gamma = how fast the dealer's hedge needs to change as price moves.** Think of an option's delta as "for every $1 the underlying moves, how much does the option change in value." Gamma is the *derivative* of that — how fast delta itself changes. - ATM (at-the-money) options have the highest gamma. The dealer's hedge has to change *fast* around the current price. - Deep OTM options have tiny gamma. They barely move, so no hedge adjustment needed. Mathematically, gamma is the same Black-Scholes quantity finance people learn first; the dealer-positioning angle is just "scale it by open interest and dollar-value to get total $ exposure." --- ## GEX (Gamma Exposure) > **GEX = how much dollar buying or selling the dealer is forced to do at a strike, summed across every contract there.** ``` GEX_per_strike = sign × gamma × open_interest × contract_size × spot² × 0.01 └──┬───┘ ── per-share ── └ contracts/share ┘ -1 for calls, +1 for puts (SpotGamma convention) ``` **The sign convention is critical** ([[03 - Math & Sign Conventions]] explains why): - **Calls → negative GEX (resistance).** Dealers are net short customer calls. Their hedging buys stock on the way up, but past a heavy call strike the hedge is saturated; further rises force them to *sell* — that pushes back down. - **Puts → positive GEX (support).** Dealers are net long customer puts. Heavy put strikes are levels where dealer hedging actively supports price. **Net GEX = call_GEX + put_GEX.** When net GEX is large and positive, dealers are net long gamma → market is *stabilizing* (mean reversion, pinning to OI clusters). When net GEX is large and negative, dealers are net short gamma → *destabilizing* (momentum, trend extension, vol spikes). The regime can flip based on where price is relative to the **gamma flip strike**. --- ## Gamma Flip > **The price level where the regime changes from positive-gamma (stabilizing) to negative-gamma (destabilizing).** Implementation: integrate per-strike GEX from low to high strikes, find the strike where the running total crosses zero. That's the flip. - **Spot > flip** → positive gamma regime → expect mean reversion, range-bound trade, pinning to strike clusters - **Spot < flip** → negative gamma regime → expect momentum, sharper swings, trend extension The dashboard banner reflects this. A move *through* the flip is itself a tradeable signal — regime change. > **Note on methodology:** SpotGamma defines flip as "the spot price at which total dealer gamma *would be* zero" (model-based, recomputing gamma at each hypothetical spot). My implementation is the simpler cumulative-integration definition (find K where Σ_{strikes ≤ K} GEX = 0). Numbers differ by a few points typically. Both signals capture the regime. --- ## Call Wall (Resistance) > **The strike with the most aggressive dealer call-hedging.** When price approaches a call wall, the dealer is forced to *buy* stock to stay hedged (delta is rising on the calls they're short, so they need to own more shares). That dealer buying initially helps price grind into the wall. Once price *touches* the wall, the dealer's already maxed out their delta hedge — any further rise has to be met by *selling* stock back (because gamma is highest near the strike). That selling is the resistance. **Ranking:** by absolute call-side GEX magnitude — strikes where each contract carries the most dealer-hedging weight. Plain OI ranking gets fooled by deep-OTM tail-hedge positions; gamma-weighted ranking surfaces the strikes that actually matter. --- ## Put Wall (Support) > **Same idea but for puts and the downside.** As price falls toward a put wall, dealer hedging activity is buying stock (positive put GEX = support). Below the wall the hedging cushion thins out and price can accelerate. Ranking is by absolute put-side GEX magnitude, parallel to call walls. --- ## Vanna > **Vanna = how the dealer's delta-hedge changes with volatility.** Specifically: ∂Delta / ∂σ = ∂Vega / ∂S. Total vanna exposure tells you how much the dealer's stock position will shift if implied vol rises or falls by 1 point. - **Positive vanna** → rising IV forces dealers to buy stock → supportive on vol expansion - **Negative vanna** → rising IV forces dealers to sell stock → destabilizing on vol expansion Less commonly featured than gamma but useful for understanding why markets sometimes rally on bad news (vol spikes cause dealer-driven dealer buying via vanna). --- ## Max Pain > **The expiration price that causes the maximum loss to option buyers** (and conversely, max profit to option *writers* — typically dealers). ``` total_payout(K) = Σ call_OI × max(0, current_price - K) × contract_size + Σ put_OI × max(0, K - put_strike) × contract_size max pain strike = argmin K ``` **Folk theory:** options expire toward max pain because the writers prefer it that way. Empirically there's a weak gravitational pull on big expiration days (third Friday of the month, end-of-quarter); on regular days it's barely detectable. Treat as soft signal, not a hard magnet. **Our implementation:** finds the literal min across all strikes in the bucket. When OI is dominated by far-OTM tail hedges (portfolio insurance), max pain can land deep OTM (e.g., ES 0DTE max pain at 6370 when spot is 7540 = 15% below spot). That's mathematically correct but practically interesting only as "where the deep-OTM put hedges live." --- ## Zero-GEX Zones > **Strikes where cumulative GEX crosses zero (besides the main flip).** The gamma flip is *the* primary regime boundary — but cumulative GEX can wobble across zero several times within a chain. Those additional crossings are minor "regime micro-pivots" — strikes where dealer hedging direction inverts. The dashboard surfaces them as a sorted-by-distance-from-spot list at the bottom. They're useful as secondary levels to watch on a trending day. --- ## L1-L4 — SpotGamma's "Large Levels" > **The top 4 strikes ranked by raw `call_oi + put_oi`** within a strike-distance window of spot. Different from call/put walls (which are gamma-weighted) and different from `Abs Gamma Strike` (which is gamma-magnitude). L1-L4 are the simple **OI-cluster locations** — where the most contracts sit, regardless of which side or how sensitive they are. Empirically reverse-engineered 2026-05-26 by comparing against SG's published L1-L4 in their indicator string. Matches **3 of 4 strikes** on ETFs (IWM, QQQ); degrades to 1-2 of 4 on indices/futures because of deep-OTM tail-hedge OI dominance (see [[09 - SpotGamma Tracking & Reverse Engineering]]). L1-L4 serve a different purpose than gamma walls: they tell you **where the option contracts are**, not **where the dealer-hedging force is strongest**. Both are useful for different trades. ## C1-C4 — SpotGamma's "Combo Levels" > Same idea as L1-L4 but **merged across the index AND its ETF** (SPX+SPY, NDX+QQQ, RUT+IWM, DJX+DIA). The merged OI distribution is then converted to a common axis and peaks are found — but the **off-grid output values** (e.g., SPX C1 = 7503 when SPX trades in 5-pt increments) prove SG uses some form of interpolation/centroid finding, not raw strike selection. We don't compute these yet — see [[09 - SpotGamma Tracking & Reverse Engineering]] for the open research question and [[07 - Future Work]] item 11. ## TL;DR cheat sheet | Term | What it tells you | How we rank | |---|---|---| | **Gamma** | Sensitivity-of-sensitivity per option. ATM strikes peak. | n/a, per-contract | | **GEX (per strike)** | Dollar dealer-hedging force at that strike. | sign × γ × OI × M × S² × 0.01 | | **Net GEX** | Total dealer hedge force across all strikes. | sum | | **Gamma flip / Vol Trigger** | Strike where regime changes. | cumulative-GEX zero crossing nearest spot | | **Call wall** | Resistance — top dealer call-hedge concentration. | by absolute call_GEX (gamma-weighted) | | **Put wall** | Support — top dealer put-hedge concentration. | by absolute put_GEX | | **Vanna $/1pt IV** | Dealer delta-shift per IV-point move. | sum across strikes | | **Charm $/day** | Dealer delta-shift per trading day passing. | sum across strikes | | **L1-L4** | Top 4 OI clusters (SG style). | top 4 by `call_oi + put_oi`, ±10% spot | | **Abs Gamma Strike** | Single biggest gamma magnet (call+put combined). | argmax `|call_gex|+|put_gex|` | | **Max pain** | Strike that minimises option-writer payout at expiry. | argmin payout over all strikes | | **Zero-GEX zones** | Secondary regime micro-pivots within the chain. | all cumulative-GEX crossings | | **Implied 1-day range** | Expected daily ± move at 1σ. | spot × ATM_IV × √(1/252) | ## See also - [[03 - Math & Sign Conventions]] for the actual formulas - [[05 - Dashboards & Run Guide]] for how to see all this on screen - [[09 - SpotGamma Tracking & Reverse Engineering]] for the L1-L4 / C1-C4 reverse-engineering analysis