GammaMonster — Methodology

Important: This document describes a hypothetical model and its backtested behaviour. All statistics reference historical data and past observations. No trades were executed. Nothing in this document constitutes investment advice, a recommendation to trade, or a guarantee of future performance. The model is presented for educational purposes only.

Section 01

The Core Observation

Options markets generate large, predictable flows. Institutional programmes — covered call ETFs, structured products, protective put strategies — create the same types of order flow on a regular schedule. These aren't discretionary bets. They're mandated by fund rules and rebalancing calendars, and because the flows are so heavily one-sided, dealers and market makers step in to provide the liquidity.

When a dealer takes the other side of an options trade, they inherit exposure they need to hedge. If we can estimate what dealers are likely holding across every strike, we can map out where their hedging pressure is concentrated — and whether it's working with the market or against it.

That's what GEX (Gamma Exposure) attempts to do. By running the full options chain through the Black-Scholes gamma formula, we estimate dealer positioning at each strike price. These are estimates based on assumptions — we don't know exactly who is on each side of every trade. But because the dominant flows are so systematic and well-documented, historical data suggests those assumptions are close to the mark.

When the resulting gamma profile is net negative, dealers who have taken on that exposure typically need to hedge with the move — selling into drops and buying into rallies. This tends to amplify moves in both directions. When it's net positive, the opposite: they buy dips and sell rallies, dampening moves like a shock absorber. The sign of gamma at the current price tells you which regime you're in.

The GammaMonster model focuses on a specific pattern within the negative-gamma regime: when a stock trades below its put wall (the strike where negative dealer gamma is most concentrated), the mechanical selling pressure can exhaust itself. IV compresses, the options driving the hedging decay, and the pressure fades. In the backtest, stocks in this condition posted positive 5-day returns more frequently than the same stocks did on randomly selected dates outside of Event D conditions.

What this is not: This isn't a claim that dealer gamma causes reversals, or that any specific event will be positive. It's an observation that, in backtested data, a specific set of measurable conditions coincided with a higher-than-baseline rate of positive short-term stock returns.

Section 02

Data Inputs

Everything the model uses is publicly available market data. Four inputs:

1. Dealer Gamma Exposure (GEX)

Estimated dealer positioning at each strike, calculated from open interest and implied volatility using the Black-Scholes gamma formula. We calculate this for the SPX (broad market) and for individual stocks.

2. VIX & Term Structure

The VIX measures 30-day implied volatility on S&P 500 options. VIX3M measures the same thing over 93 days. The ratio between them tells you the shape of the curve — above 1.0 means near-term implied volatility is higher than long-term (backwardation).

3. Put Wall

The strike price where negative dealer gamma is most concentrated on a given stock. Think of it as the level where dealer hedging pressure is heaviest.

4. Distance-to-Wall

How far the stock's current price sits from its put wall, as a percentage. Closer = more intense hedging pressure at the current price.

Section 03

Event Classification

The model flags an Event D (Put Wall Snap) when all four of these conditions line up at the same time:

Condition	Requirement	Data Source
Stock price vs put wall	Below put wall	GEX profile + price
Single-stock dealer GEX	Confirmed negative	Options open interest
SPX aggregate GEX	Negative regime	SPX options chain
Distance to put wall	<5% below	Price / put wall

Each observation gets a composite score from 0.00 to 1.00, based on four things: how close the stock is to its put wall, how much negative gamma is sitting at that wall strike, whether single-stock GEX confirms the setup, and the broader SPX gamma profile. Higher score = the conditions are more tightly aligned with what the backtest flagged.

What the score means: A higher score doesn't predict higher returns. It means the observable data more closely matches the conditions that, historically, coincided with a higher positive rate. Scores below 0.50 were barely above a coin flip in the backtest.

Section 04

Regime Classification (Tier System)

Not all market environments are equal. In the backtest, the volatility regime at the time of each event corresponded to meaningfully different 5-day return distributions. The model sorts regimes into three tiers using two data points — VIX level and term structure shape:

Tier	VIX Level	VIX / VIX3M	5D Pos Rate (close-to-close)	Avg 5D (close-to-close)
Tier 1 — Monster	>25	>1.0 (backwardation)	74.2%	+4.7%
Tier 2 — Elevated	>20	Any	66.1%	+3.2%
Tier 3 — Baseline	≤20	Any	57.4%	+1.4%

These boundaries came from splitting the backtest data by volatility environment and seeing what the return distributions looked like in each bucket. They describe what happened historically — not what will happen next.

Section 05

Backtest Overview

We ran the model's rules across historical data and found 1,867 Event D observations. For each one, we measured the stock's close-to-close return (the percentage change in closing price from the event date) at three horizons: 1 day, 5 days, and 10 days. No trades were modelled — just publicly available closing prices.

Metric	Value
Total Event D observations	1,867
Date range	Jan 2020 – Dec 2025
Universe	US equities with listed options (liquid names)
5-day positive rate (close-to-close)	66.8%
Average 5-day close-to-close return	+3.4%
Median 5-day close-to-close return	+2.1%
Execution assumptions	None — no trades modelled
Transaction costs modelled	None

Backtest limitations: These are observations, not trading results. No transaction costs, no slippage, no portfolio effects. The average return is higher than the median, which tells you a few big winners are pulling the average up — the typical outcome is smaller than the headline number suggests. Past patterns don't guarantee future results.

Section 06

Market Microstructure Rationale

This is the most important section. The data tells you what happened in the backtest. This section explains why there's a structural reason behind it.

The Two Regimes

Contango (normal market). VIX curve slopes up, dealer gamma is typically positive. Dealers tend to buy dips and sell rallies — dampening moves in both directions. Historically, this environment has been associated with lower realised volatility. This is the regime most traders are used to.

Backwardation (inverted curve). VIX curve inverts, dealer gamma tends to go negative. Now dealers typically need to hedge with the move — selling into drops, buying into rallies. This tends to amplify moves instead of dampening them. The range of outcomes widens in both directions.

The Mechanical Chain

1. Client flows are predictable. Covered call ETFs, structured products, and protective put programmes all generate regular, systematic order flow. Because these flows are so one-sided, dealers step in to provide liquidity — and in doing so, they take on exposure they need to hedge. Since the flows follow mandated schedules (not gut feelings), we can estimate dealer positioning at each strike using the Black-Scholes gamma formula. These are assumptions, but ones that historical data supports.

2. Inverted curve steepens the put skew. When near-term IV is elevated (backwardation), the puts that drive dealer hedging sit at higher volatility levels than normal. The premium for downside protection gets steeper.

3. Vanna builds the pressure. Vanna is how an option's delta changes when IV moves. On an inverted curve, as each option rolls toward expiry it moves into higher vol. That increases the delta on dealers' puts, meaning they need to sell more shares to stay hedged. The selling pressure doesn't stay flat — it tends to grow day by day. Layered on top of negative gamma, this compounds the mechanical selling pressure near the put wall.

4. Exhaustion. Dealer hedging is finite — bounded by the size of the options positions driving it. Once the bulk of the selling is done, the pressure fades and price stabilises near the put wall.

5. Vanna reverses. If the move stabilises and IV starts to compress — especially on the near-term puts that were most inflated — falling IV decreases delta on those same positions. Dealers may then be over-hedged — holding more short stock than they need. To rebalance, they would buy shares back. In theory, the same vanna that amplified selling on the way down can create buying pressure on the way back up.

6. Theta accelerates the unwind. Near-term options decay fastest. Each day that passes reduces their gamma and their grip on dealer hedging. The pressure has a built-in expiry date.

The theoretical cycle: steep put skew → dealer selling (gamma + vanna) → exhaustion near the wall → IV compression → vanna reversal → dealer buyback → snap toward the put wall. Not every event follows every step, and the sequence can break down at any point.

What The Model Watches For

The model isn't just watching negative gamma. It's watching the flip between regimes. In contango, dealer hedging tends to dampen moves. In backwardation, the range of outcomes widens — bigger drops, but also bigger rallies. The model identifies the specific conditions where, historically, this mechanical sequence played out at the single-stock level and coincided with a higher-than-average rate of positive short-term returns in the backtest.

Caveat: A structural explanation isn't proof, and a historical pattern isn't a guarantee. This all works as long as the underlying plumbing — structured product flows, systematic overwriting, dealer hedging mechanics — stays consistent. Markets evolve, dealer behaviour adapts, and regulation changes.

Section 07

What This Model Is Not

Clarity about what this model does not do is as important as understanding what it does.

It is not a trading signal.

The model identifies observable market conditions. It does not generate buy or sell recommendations, and no part of the dashboard or educational materials should be interpreted as instructions to trade.

It is not a portfolio strategy.

The backtest does not model a portfolio. There is no position sizing, no capital allocation, no risk management framework, and no compounded return track record. The statistics describe individual stock return observations in isolation.

It is not a guarantee of edge.

Historical statistical patterns are descriptive, not predictive. The positive rates and average returns documented in the backtest describe what happened in the past. They are not a statement about what will happen in the future.

It is not investment advice.

GammaMonster is an educational product. It teaches the concepts of dealer gamma exposure, volatility regime classification, and market microstructure using a rules-based screening model as a teaching framework. Members are responsible for their own investment decisions.

Section 08