Part IX · Volatility in practice · Chapter 9C
Reading the tape
Classifying the day's regime
Chapters 9A and 9B applied the dealer's hedging to the VIX and to same-day options. Theory only matters if it changes what you see on a chart, so this chapter turns the framework into a way of reading the day. It gives seven heuristics, a classifier that makes them an explicit decision, and a mental model of what moves the index.
- Gamma exposure (GEX), its sign conventions and the gamma flip (Chapters 8A to 8C); the two regimes and their measured effect sizes (Chapters 8D and 8G).
- The VIX term structure and the ratio of the VIX to the VIX3M (Chapter 9A).
- Vanna, charm and the post-event unwind (Chapter 5D), and the charm clock of same-day options (Chapter 9B).
- Pin risk at expiry (Chapter 5F).
1 · Seven heuristics, ordered by impact
These heuristics follow from everything in Parts I to VIII. The Dealer Mechanics Lab lists them in declining order of importance, so the first is the one to get right before any other.
Check the sign of GEX before choosing a strategy
Positive GEX days favor premium selling, mean reversion and tight stops. Negative GEX days favor directional trades, longer holds and wider stops. Strategy selection is the single biggest leverage point: using mean reversion in a negative GEX regime is how accounts get blown up.
A GEX figure from a public vendor depends on a sign convention that may not match the flow regime of the product you trade. Chapter 8B shows the same open interest giving opposite signs under the two conventions. Verify the convention before trusting the number.
Find the gamma flip
The price at which net dealer gamma crosses zero, the gamma flip, is a regime boundary. Slow approaches to the flip rarely matter. Approaches under stress, with the VIX rising, breadth deteriorating and news pending, frequently mark the moment when feedback flips sign. SpotGamma's "Volatility Trigger" is that firm's proprietary version of this level, though the concept is the important part.
Pair GEX with the VIX term structure
Positive GEX with the VIX in contango is the stable regime, where mean reversion dominates. Negative GEX with the VIX in backwardation is the stress regime, where momentum dominates. Together the two signals capture both the positioning of risk and its pricing, which neither captures alone.
Watch the charm clock
Charm flows accumulate through the day and compress into the last 30 minutes (Chapter 9B). Trends into the 3:30 to 4:00 PM ET window frequently extend on mechanical rebalancing, especially on OPEX Fridays, the Fridays of monthly options expiration. The evidence for that comes from index and other futures. Among individual stocks the last 30 minutes tend to reverse instead (Chapter 6C). The first hour after the open is often noise, as overnight charm and vanna unwinds clear.
Expect the vanna unwind after an event
Vanna sets up before an FOMC decision, a CPI release or a jobs report (NFP). Implied volatility is bid into the event, so dealers short puts carry larger short-stock hedges than they would in calm. After the event, regardless of direction, implied volatility collapses in a vol crush. The short puts' deltas decay toward zero, and dealers buy back hedge stock. This is the post-event ramp that happens whatever the outcome. The play is long volatility going in and short volatility coming out, but only if you can size and time it correctly.
Step 1: before the event. Take the example put of Chapter 5D, struck at 95 with spot at 100, three months out and with zero rates. Its implied volatility has been bid to 25% into the event. Its delta is −0.318, so a dealer short 1,000 contracts is long 31,800 share-equivalents and holds a short hedge of 31,800.
Step 2: after the crush. The event passes and implied volatility falls back to 20%, with spot unchanged. The put's delta is now −0.287, so the dealer's long delta shrinks to 28,700. It buys back 3,100 share-equivalents of its hedge, whatever the news said.
Treat pinning as a tiebreaker
Pinning is the pull of a strike with heavy open interest on the price near expiry, the market-wide face of the pin risk of Chapter 5F. The Dealer Mechanics Lab cites Ni, Pearson and Poteshman (2005) for 16.5 basis points of return alteration on equity expiration days, with the effects concentrated near high-open interest strikes. Their study covers optionable single stocks from 1996 to 2002. On each expiration date those stocks' returns were altered by an average of at least 16.5 basis points. Closes clustered at strikes most often when the market makers who delta-hedge were net long the options. The paper does not study the index, so the figure does not convert into SPX points.
For the index, Golez and Jackwerth (2012) studied S&P 500 futures. The futures were pulled toward the at-the-money strike on days when options on them expired, shifting at least $115 million of notional futures value per expiration day, small beside that market. Right before SPX index options expired, the same futures were pushed away from the strike instead.
Pinning is real but small: useful as a tiebreaker between alternative outcomes, not as a standalone trade thesis. It is stronger in single stocks than in the index, and stronger in names with lower liquidity.
Calibrate your skepticism of GEX commentary
The empirical effect sizes are tens of basis points per standard deviation, not full percentage points (Chapter 8G). When commentators describe "massive gamma squeezes" driving 2% moves, they are usually wrong. The move came from news or from another mechanism, with gamma contributing on top. Hold this prior strongly.
Step 1: the effect. Suppose, as a round number in the range the literature reports, that dealer gamma moves the index by 20 basis points for each standard deviation of gamma imbalance.
Step 2: a large imbalance. On a day with an imbalance two standard deviations from normal, gamma contributes about 40 basis points. A 2% move is 200 basis points, so gamma plausibly explains about a fifth of it, with news or other flows explaining the rest.
2 · Where the Edge Book disagrees
The Edge Book reads the same evidence more skeptically. The disagreement is sharpest over the first two heuristics. It rejects GEX-level scalping, selling at a call wall and buying at a put wall, as never robust for a retail account (Chapter 10I). It keeps only the sign of GEX, and only as a low-confidence regime filter (Chapter 10H). In its rule, a negative sign means expecting higher realized volatility and trimming put-selling modestly, while a positive sign means sizing normally. The sign is worth 10% to 15% of one filter's weight, and is ignored when it disagrees with the VIX term structure or realized volatility.
Its first reason is that public GEX models are model-dependent: vendors disagree because they assume different holders for each strike. The second is that the explosion of 0DTE after 2022 changed the structure those models assumed. The third is a 2025 Cboe study, Mandy Xu's "0DTEs Decoded" (May 2, 2025), that puts 0DTE market-maker hedging at about 0.2% of SPX daily liquidity. The Edge Book reads that as contradicting the narrative that dealer flows dominate the day (Chapter 1G). Because the study measures same-day hedging only, reading it against GEX levels built mostly from longer-dated open interest extends the measurement.
The two sources agree more than this suggests. Both treat the GEX sign as regime context rather than a signal. Both warn that public GEX rests on assumptions and share the seventh heuristic's skepticism. They part on how much weight dealer gamma deserves: the Dealer Mechanics Lab puts strategy selection by GEX first, while the Edge Book lets it nudge sizing at most. They also part on premium selling in calm regimes, which the classifier below favors and the Edge Book, for systematic iron condors, rejects (Chapter 10I).
3 · The regime classifier
The figure below turns three readings into one regime. Net dealer GEX runs from −$10 billion to +$10 billion per 1% move, starting at +$2 billion. The VIX level runs from 10 to 60, starting at 18. Five buttons set the VIX term structure, from steep contango to deep backwardation, starting at shallow contango. Each input earns a score:
| Input | Scores |
|---|---|
| Net dealer GEX | −2 above +$4 billion; −1 above +$1 billion; 0 above −$1 billion; +1 above −$3 billion; +2 at −$3 billion or below |
| VIX level | −1 under 15; 0 from 15 to under 22; +1 from 22 to under 30; +2 at 30 or over |
| Term structure | −2 steep contango; −1 shallow contango; 0 flat; +1 shallow backwardation; +2 deep backwardation |
The three scores add up along a line from −5 to +6. A total of −2 or below reads Stable, −1 or 0 reads Range, +1 or +2 reads Caution, and +3 or more reads Stress. Each regime comes with a strategy line and four notes. Move the GEX slider from +$2 billion down through zero to −$4 billion, then press deep backwardation and raise the VIX to 35.
Shallow contango: mild upward slope; nothing structural.
Stable, score −2. Sell premium · pin trades · tight stops.
▸ Mean-reversion is the dominant intraday character
▸ Range trades preferred over breakouts
▸ Theta decay accrues smoothly; harvest it
▸ Watch for VIX term flattening as the early-warning
At the defaults the scores are −1, 0 and −1, a total of −2, which reads Stable: sell premium, pin trades, tight stops. Taking GEX to −$4 billion moves its score to +2 and the total to +1, Caution: trim premium, favor directional trades, widen stops. With deep backwardation and a VIX of 35 as well, the total reaches +6, Stress: size down, long premium only, no mean reversion.
The classifier surfaces a single question: is today's regime the one you are trading for? It gives you a structured way to override your defaults when it is not. The scoring is heuristic, not a backtest. Use it to surface the question, not as a signal in itself, and override it freely when context dictates.
Its four regimes map onto the five of Chapter 9A's table. Stable matches Stable, Range matches Normal and Caution matches Transition, while Stress covers both Stress and Crisis. The classifier also reads the VIX level, which the table leaves out.
The classifier is a points score of the kind clinical guidelines build by expert consensus before any validation cohort exists: points per factor, summed, cut into bands that each carry a recommendation. Such a score organizes judgment and makes sure the right questions get asked. Its cut points, though, were never fit to outcomes. Its inputs also overlap, as the VIX level and backwardation do, which counts the same stress twice. Read it the way you would read an unvalidated score: as a prompt, not a prognosis.
4 · A unified mental model
The S&P 500 on a typical day is the sum of three forces, in roughly decreasing order of magnitude:
- Fundamental news and information is small in the number of events on any day but large when it hits: an FOMC decision, a CPI print, a major earnings release, a geopolitical event. It is what drives markets in the strict sense.
- Mechanical positioning, from dealer hedging, volatility targeting, risk parity rebalancing and options-related flows, is large in continuous magnitude but mostly self-cancelling across participants.
- Microstructure noise, from bid-ask bounces, auction imbalances and opportunistic liquidity-seeking, is always present and never the primary mover. It is the ground on which higher-frequency strategies operate.
Most days, force two dominates by volume while force one dominates by directional impact. The dealer-hedging story of Parts VI to IX is the single largest component of force two. On most days the market is a delta-hedging machine reacting to the existing options inventory. Price does not move because "buyers stepped up" in some abstract sense. It moves because somebody is mechanically required to transact a specific size at a specific price to remain neutral.
Understanding who that somebody is and what they are forced to do is the entire game, in the Dealer Mechanics Lab's view. The Edge Book's view, above, is that the game is smaller than that.
5 · Where to go next
The Dealer Mechanics Lab suggests three directions of deeper study, in order of practical leverage.
Track your own intraday GEX before trading on it. Take Cboe data, SpotGamma, Tier1Alpha or your preferred vendor, and overlay it on your SPX or SPY chart for two weeks before trading on it. You see when it works and when it does not. This forward walk is the only way to develop a feel for the framework's effect sizes in your own timeframe and instruments.
Read the academic sources directly: Ni, Pearson and Poteshman (2005); Barbon and Buraschi (2021); Garleanu, Pedersen and Poteshman (2009); Dim, Eraker and Vilkov (2023); Adams et al. (2025); Baltussen, Da, Lammers and Martens (2021); and Demeterfi, Derman, Kamal and Zou (1999). Vendor frameworks are necessarily simplified for marketing, while the papers expose the assumptions and limitations. After the papers, the vendor commentary reads very differently.
Backtest the regime-aware version of your existing setups. Conditional on positive against negative GEX, what is your win rate? In the Dealer Mechanics Lab's experience the asymmetry is usually larger than you expect. It also persists out-of-sample more reliably than most other technical signals. The Edge Book, which grades the GEX sign as low confidence, would expect the gap to be small.
A backtest split by regime is a subgroup analysis, to which the same guards apply. Pre-specify the split before you look, test the interaction rather than comparing two separate win rates, and remember that every extra split you try is another comparison. A regime effect found on the fifth way of cutting the data is the multiplicity problem of Chapter 1D. Two weeks of forward tracking is also a small sample.
6 · Final thought
"Every option ever traded created an obligation for someone to transact in the underlying. The market is, in part, the integral of those obligations."
The Dealer Mechanics Lab
Read "integral" as accumulation: the market is, in part, the running sum of those obligations. This is the unified statement of the dealer-mechanics framework. Parts I to V built the prerequisites. Parts VI to VIII derived and aggregated the mechanism. Part IX applied it to the market as it trades in the 0DTE era. Part X then weighs each known source of return on its evidence. Parts XI to XIII translate the framework into operational discipline: strategy selection, position sizing, and the rules that keep the framework from being destroyed by behavior.
A pre-trade read takes a few minutes, starting with the sign of GEX and the convention your source uses for it. Read the VIX/VIX3M ratio and the VIX level, and put all three through the classifier. Check the calendar for an FOMC, CPI or NFP release and for OPEX. Then check the clock: the first hour and the last 30 minutes are not the middle of the day.
Then compare the regime with the trade you had planned. A premium-selling trade in a Caution or Stress regime, or a mean-reversion trade with negative GEX and backwardation, needs a reason written down before you place it. On the Edge Book's reading, the GEX sign moves your size by a little while the term structure does the real work.
Every heuristic here inherits the weaknesses of GEX. The sign is an assumption about who holds what, open interest is published a day late, and FLEX and over-the-counter books are invisible to public models (Chapter 8B). News, force one, can swamp any positioning read in minutes. The classifier's scores and cut points are heuristic, never fit to outcomes. Finally, the effects are tens of basis points, small enough for costs and bad fills to eat.
- Before choosing a strategy, write down the regime from GEX, the VIX level and the term structure. If the setup you want belongs to another regime, pass or cut its size.
- Never run mean reversion into negative GEX with the VIX in backwardation.
- Use pinning and the gamma flip as tiebreakers and boundaries, not as trades in themselves. On this both sources agree: neither backs trading GEX levels.
- Track GEX for two weeks before it touches a live decision, and judge it against your own backtest split by regime.
A 2% day after a "massive gamma squeeze" headline is almost always a news day with gamma on top. With effects of tens of basis points per standard deviation, gamma explains a fraction of such a move. Size and stop your trade for the news, not for the story.
Two vendors can publish opposite signs from the same open interest, because they assume different holders. A reading that flips your strategy should be one whose convention you know and whose flow regime matches your product (Chapter 8B).
- Choose the strategy by regime first: positive GEX favors premium selling and mean reversion, negative GEX favors directional trades and wider stops.
- The gamma flip matters when it is approached under stress, while pinning is real but small, a tiebreaker.
- Pair GEX with the VIX term structure, since either alone is incomplete. Mind the clock: charm fills the last 30 minutes, while the vanna unwind follows events whatever their outcome.
- The classifier sums three scores into Stable, Range, Caution or Stress. It is a heuristic that prompts the question, not a backtested signal.
- Gamma's effects are tens of basis points per standard deviation. The Dealer Mechanics Lab puts dealer hedging at the center of the day's volume, while the Edge Book gives it far less weight.
7 · Check yourself
Question 1
GEX reads −$2 billion per 1% move, the VIX is 26 and the term structure is in shallow backwardation. What regime does the classifier give, and what does it suggest?
Answer
The scores are +1, +1 and +1, a total of +3, which reads Stress: size down, long premium only, no mean reversion. No single reading is extreme, but together they cross the line.
Question 2
A commentator turns Ni, Pearson and Poteshman's 16.5 basis points into about 8 SPX points at 5,000. What is wrong with that? Why is pinning a tiebreaker rather than a trade?
Answer
The 16.5 basis points is a lower-bound average for optionable single stocks from 1996 to 2002. The paper does not study the index, so the figure does not convert into SPX points. The index evidence, Golez and Jackwerth (2012), is at least $115 million of notional S&P 500 futures per expiration day, small beside that market. Either way the effect is small against a normal day's range and the costs of a trade. It is also concentrated near high-open-interest strikes at expiry, so it can tip a close call but cannot carry a trade on its own.
Question 3
After an FOMC decision, implied volatility falls from 25% to 20% while dealers are short the chapter's example puts. Which way do they trade, and why does the direction of the news not matter?
Answer
They buy, because the fall in implied volatility pulls the puts' deltas toward zero, from −0.318 to −0.287. That shrinks the dealers' long delta and leaves their short hedge too big. The trade comes from the drop in implied volatility, which follows the event either way.
Question 4
A commentator blames a 2% fall on a gamma squeeze. With an effect of 20 basis points per standard deviation and an imbalance of two standard deviations, how much of the fall could gamma explain?
Answer
Gamma could explain about 40 basis points of the 200, a fifth of the fall. The rest came from news or other flows, as the seventh heuristic says to assume.
Question 5
You split your backtest by the sign of GEX and find a 64% win rate in positive regimes against 52% in negative ones. What should you check before acting on it?
Answer
Check that the split was chosen before you looked, and that the difference survives a test of the interaction rather than a comparison of two separate estimates. Count the other splits you tried, and confirm that the sign comes from a convention that matches your product. Then track it forward, as the chapter suggests, before it touches live size.