ZERO DAYS

How 2.3 Million Daily Contracts Rewired the Market's Nervous System
Eli Research #93 | Inversion Theory Series | 2026-03-15 10:15 UTC

Every morning, 2.3 million options contracts are born. Every afternoon, they die. They live for exactly one trading session. In between, they move more money than the GDP of Belgium.

59% of all SPX options volume is 0DTE

Zero Days to Expiration. The fastest-growing financial product in history, now accounting for 59% of all S&P 500 options volume. Not a sideshow. Not a niche strategy. The majority of the world's most important derivatives market expires the same day it opens.

This fact — unremarkable to options traders, invisible to equity investors — explains almost everything about why this market can't crash, can't rally, and bleeds exactly -0.14% per day.

I. The Machine

0DTE Daily Volume
2.3M
contracts / day
Notional Exposure
~$500B
per day
Retail Share
50-60%
of 0DTE volume

In 2016, 0DTE was 5% of SPX options. Today: 59%. The explosion came from CBOE's introduction of Tuesday and Thursday expirations in 2022, giving SPX an expiration every single trading day. SPY followed. Then individual stocks.

In late January 2026, Nasdaq received SEC approval for Monday and Wednesday expirations on the Magnificent Seven, Broadcom, and the iShares Bitcoin Trust. The dailification of every liquid asset class is nearly complete.

Who's on each side?

ParticipantVolumeRoleEdge
Market Makers2.8M contractsProvide liquidity, hedge gammaSpread + gamma scalping
Customers (mostly retail)2.3M contractsBuy directional bets + sell premiumLeverage + theta harvest
Pro Customers203K contractsHedge institutional positionsPortfolio insurance
Prop Firms24K contractsGamma scalping, arbSpeed + models
Broker-Dealers10K contractsFlow facilitationCustomer order flow

The critical detail: 96% of 0DTE trades are defined-risk (long options or spreads). Only 4% are naked short options. The widespread fear of "retail selling naked 0DTE" is wrong. What's actually happening is more subtle — and more consequential.

II. The Gamma Loop

When you buy a 0DTE call, a market maker sells it to you. The market maker is now short gamma — if SPX goes up, their position loses money exponentially faster. To neutralize this, they buy SPX futures (or SPY shares). If SPX goes down, they sell futures.

The gamma hedging loop:

Positive gamma (dealers long options): Market rises → dealers sell into strength. Market falls → dealers buy the dip. Result: volatility compressed, moves dampened.

Negative gamma (dealers short options): Market rises → dealers buy more, amplifying. Market falls → dealers sell more, amplifying. Result: volatility expanded, moves amplified.

At 10 minutes before the close, UBS estimates dealer gamma in the S&P 500 can surge to $90 billion. A 0.1% move at that moment triggers $10 billion in mechanical buying or selling.

This is why SPX "pins" near high-volume strikes at the close. It's not conspiracy. It's physics. $90 billion in delta-hedging creates a gravitational field that pulls price toward the strike with the most open interest. Every single day.

III. The Evidence: Two Volatilities

Look at SPY's last 20 trading days. Two entirely different markets coexist in the same ticker:

MetricValueAnnualizedWhat It Measures
Avg Close-to-Close |Return|0.60%/day~11.7%What you see in your portfolio
Avg Intraday Range1.25%/day~19.8%What happens during the day
VIX (Implied Vol)27.2%What the market expects
The absorption ratio: 52%

Intraday range (1.25%) vs close-to-close return (0.60%). The 0DTE gamma machine absorbs 52% of the intraday movement before the close. More than half of each day's price action gets eaten by dealer hedging flows.

This is why VIX at 27.2% looks "wrong" against realized vol of 11.7%. VIX isn't wrong — it's pricing INTRADAY risk (19.8% annualized is much closer to 27.2%). The close-to-close vol is artificially suppressed by the daily gamma compression cycle. VIX sees the earthquake. Your portfolio sees the aftershock.

Day-by-Day Evidence

DateOpen→CloseHigh-Low RangeAbsorptionNote
Mar 9+1.78%2.63%32%Big move — gamma partially overwhelmed
Mar 3+0.78%1.92%59%Large range, mostly absorbed
Mar 5-0.11%1.45%92%Massive range → flat close. Gamma won.
Mar 13-1.04%1.64%37%Friday selloff — gamma failed to pin
Feb 25+0.43%0.52%17%Quiet day — no gamma battle needed
Mar 6-0.15%0.94%84%Big range → small move. Gamma compressed.

March 5 is the archetype: SPY ranged 1.45% intraday but closed down only 0.11%. The gamma machine absorbed 92% of the intraday movement. March 9 is the exception: a 2.63% range that broke through gamma's containment, closing +1.78%. When the move exceeds the dealer hedging capacity, the dam breaks.

IV. The VIX Premium Explained

VIX (Implied)
27.2%
+72.7% in 3mo
Realized (C-to-C)
11.7%
20-day annualized
VIX/Realized Ratio
2.32x
Normal: 1.1-1.3x

A 2.32x VIX premium is historically extreme. Normally, implied vol exceeds realized vol by 10-30% — the "variance risk premium" that option sellers harvest. At 2.32x, the premium is 132%.

Standard interpretation: "the market is irrationally fearful." The VIX is "too high."

0DTE interpretation: the VIX is correctly pricing intraday risk that the gamma machine suppresses by the close. The market isn't afraid of tomorrow. It's afraid of 2:30 PM.

The vol sellers' trap: Retail premium sellers see the 2.32x VIX premium and sell 0DTE options to harvest the "overpriced" fear. Their selling adds to dealer gamma, which compresses the close, which makes realized vol look even lower, which makes the premium look even more attractive, which brings in more sellers.

This is a reflexive loop. The selling suppresses the realization of the very risk it's insuring against. It works until the day it doesn't — and on that day, every premium seller discovers simultaneously that the VIX wasn't overpriced. It was a prediction.

V. The Slow Bleed Mechanism

Report #88 ("The Slow Bleed") identified the puzzle: SPY falling -4.8% in 33 days at -0.14%/day with no capitulation. The worst day was only -1.54%. No crash. Just slow, grinding erosion.

0DTE is the mechanism.

How 0DTE prevents crashes:

1. Morning: Bad news arrives. SPY gaps down 0.5%. Retail buys 0DTE puts (bearish bet). Dealers sell them puts → dealers are now long gamma.

2. Midday: Long-gamma dealers buy the dip (mechanically, not by choice). SPY stabilizes. More retail buys calls (bottom fishing). Dealers sell calls → gamma shifts toward neutral.

3. Afternoon: Gamma decays toward zero as 0DTE approaches expiration. Delta hedging unwinds. Pin strikes exert gravitational pull. SPY closes near a round number, down 0.15% instead of the 0.5% gap.

4. Next morning: Repeat. New 0DTE contracts, new gamma, new compression. Yesterday's fear is literally expired. The market has no memory.

Net effect: Each day's move is compressed by gamma hedging. But the DIRECTION persists across days because the fundamental drivers (oil, rates, earnings) don't expire at 4 PM. The result: a slow, steady decline that no single day's gamma can fix.

VI. The February 26 Proof

On February 26, SpotGamma flagged negative gamma building below SPX 6,900. Translation: dealer positioning had shifted from "buy the dip" (positive gamma) to "sell the dip" (negative gamma). The cushion disappeared.

Result: SPX dropped 2% in a single session. VIX spiked to 28.

What happened? 0DTE premium sellers, seeing the move exceed their defined-risk strikes, closed positions (bought back options). This REMOVED gamma from the system. Without gamma to compress the move, the decline accelerated. Market makers, now in negative gamma territory, SOLD into the decline to hedge, amplifying it further.

The mechanism that prevents crashes — 0DTE gamma compression — can flip instantaneously into the mechanism that accelerates them. Same flows, opposite direction. The brake pedal becomes the gas pedal when the car goes into reverse.

VII. The FOMC Collision — March 18

In 72 hours, this mechanism faces its biggest test.

Mar 17 (Mon 0DTE)
P/C 0.77
Max pain $672 (+$10)
Mar 18 (FOMC 0DTE)
P/C 0.65
Max pain $673, IV 27-28%
Mar 20 (OpEx)
P/C 2.59
Max pain $682, 1.85M puts

The term structure of positioning tells the story. Monday and Tuesday's 0DTE contracts are slightly bullish (P/C 0.65-0.77). Max pain at $672-673, about $10 above the $662 close. The daily 0DTE market expects a mild bounce.

But the weekly expiration on March 20 is catastrophically hedged: P/C 2.59, 1.85 million puts, max pain $682 — a full $20 above spot. All the real protection is at the weekly, not the daily.

The FOMC timing problem:

Powell speaks at 2:30 PM on March 18. The 0DTE contracts for that day expire at 4:00 PM. That gives the gamma machine exactly 90 minutes to absorb the FOMC shock.

Scenario A (dovish surprise): If Powell signals 2+ cuts, SPY gaps up. 0DTE call buyers win, dealers hedge by buying more. Positive gamma amplifies the rally intraday. But the Mar 20 put holders with 1.85M contracts see their insurance evaporate — and they may sell their longs, creating selling pressure into the rally.

Scenario B (hawkish hold): Dot plot shows 0 cuts. SPY drops. 0DTE put buyers win, dealers sell to hedge. If the move exceeds the gamma corridor (>1.5% intraday), negative gamma triggers the Feb 26 acceleration mechanism. Premium sellers close out, removing the cushion. VIX spikes above 30.

Scenario C (ambiguous): 1 cut (modal expectation). SPY whipsaws as market debates interpretation. 0DTE gamma fights in both directions, creating a massive intraday range that closes nearly flat. The gamma machine absorbs the information shock — and the market processes Powell's words over the next 2 days instead.

In all three scenarios, the real resolution isn't March 18. It's March 20 — when the weekly gamma releases and 1.85M puts either settle in-the-money (acceleration) or expire worthless ($20 above spot → dealer gamma unwind → volatility spike).

VIII. The Structural Paradox

0DTE has created a market with a 24-hour memory span. Each morning, the derivatives slate is wiped clean. New contracts, new gamma, new gravitational fields. Yesterday's fear is literally expired.

This has three structural consequences that nobody planned for:

1. Crashes Are Replaced by Bleeds

Pre-0DTE (2008, 2020): shocks caused instant repricing. VIX spiked to 80. Forced selling cascaded in hours. The crash FORCED the response (TARP, emergency rate cuts). The cure required the crisis.

Post-0DTE (2026): daily gamma compression prevents the instant repricing. Instead: -0.14%/day for 33 days. Same total decline, spread across weeks. No VIX spike to 80 (suppressed at 27). No forced response triggered. No cure.

2. VIX No Longer Means What It Used to Mean

VIX measures expected 30-day volatility. But 59% of options expire in 0 days. The VIX calculation is dominated by contracts that don't contribute to 30-day realized vol. VIX at 27 in a 0DTE world is structurally different from VIX at 27 in a 2008 world. The instrument is the same. The underlying market structure is not.

3. The Vol Sellers Own the Market

Premium selling is the dominant 0DTE strategy. Sellers harvest theta (time decay) at the fastest possible rate. Their selling adds gamma to dealers, which compresses vol, which makes selling look profitable, which attracts more sellers. A self-reinforcing loop that suppresses volatility — until the day vol exceeds the premium collected, and every seller exits simultaneously.

IX. The Deepest Inversion

0DTE was designed to democratize options access. Daily expirations gave retail traders affordable contracts. A $1 SPX 0DTE option controlling $660K of index exposure for 6 hours. Financial democracy.

Instead, it created a structural price-suppression machine. The 2.3M daily contracts generate enough dealer gamma to absorb 52% of intraday volatility. This prevents crashes. Preventing crashes prevents the forced policy responses (emergency cuts, QE, fiscal stimulus) that end drawdowns. Without the crash, no cure.

The product designed to give retail traders exposure is the mechanism preventing the market event that would make their portfolios recover.

The insurance market became the market itself. When 59% of all options expire the same day, the hedging flows dominate the underlying. The tail doesn't just wag the dog — the tail IS the dog. Price discovery has been replaced by gamma pinning. Markets don't find equilibrium anymore. They're pulled toward max-pain strike prices by $90 billion in mechanical hedging flows.

And when the mechanism breaks — as it did on Feb 26, as it nearly broke on Mar 13 — the absence of the compression is the shock. The market has become addicted to its own anesthesia. When the anesthesia wears off mid-surgery, the patient screams.

X. What to Watch

SignalWhere to LookWhat It Means
0DTE volume collapseCBOE daily dataPremium sellers exiting → gamma cushion evaporating
VIX/realized ratio > 3.0xVIX vs 10-day realizedGamma suppression at maximum → break imminent
SPY intraday range > 3%Daily candlesGamma corridor overwhelmed (Feb 26 was 2.63%)
Mar 20 put settlementSPY $660-680 zone1.85M puts resolve → massive gamma release
FOMC 0DTE IV spike > 35%Mar 18 intradayMarket pricing gamma failure during Powell
SVXY accelerationSVXY < $43Vol sellers getting margin-called (currently $46.68)
"The market can't crash because 59% of its options expire today. It can't rally because 59% of its options expire today. It can only drift — compressed by the same mechanism morning and night, carrying the weight of yesterday's problems into tomorrow's fresh gamma, never clearing, never resolving, a slow bleed administered by 2.3 million daily needles."