Ainvest Option Flow Digest - 2026-07-07: 🎯 $92M and Finally Some Real Structures — an NVDA Earnings Roll, Two Bull Spreads, a Bear Bet on a Hot AI IPO, and Four LEAP Crosses

📅 July 7, 2026 | 🔎 THE READ: NVDA's $25M bullish call roll into Aug-26 earnings + SOXX's $21.5M bull-call spread into the chip-earnings run + a rare bearish SHAZ put spread | ⚠️ Half the tape is readable directional structure; the other half is direction-unprovable LEAP crosses

✅ Updated July 8, 2026 (pre-market) — this session's ⏳ flags are now resolved by next-day OI. GFL (both legs), SOXX (both legs), SHAZ (both legs), IBIT, PDD and QQQ all confirmed fresh OPENS. NVDA's Sep $195 long leg is confirmed open; its Jul $200 short-leg close wasn't independently confirmed (strike OI rose on broad noise) and rests on the archive match. The one inversion: 🍽️ TOST — the $2.18M $35-call cross resolved as a net CLOSE / unwind (OI fell 21,148 → 17,750, −3,398), NOT a fresh bullish bet. Full detail feeds tomorrow's OI Review.

---

🎯 Quick Read: $92.0M Net Across 8 Names — Cleaner, More Directional Than Yesterday

Today gave us ≈$92.0M in net institutional premium across 8 names with actual readable structure. Four are provable directional bets (an earnings roll, two bull spreads, a bear spread); the other four are direction-unprovable LEAP call crosses (QQQ, IBIT, PDD, TOST) — read the mechanism, not the dollar sign.

• Biggest structure: 🤝 QQQ $29.9M — a sold Sep-2027 $720 LEAP call, block cross → direction unprovable (overwrite/financing/short-vol).
• Smartest timing: 🐋 NVDA $25.1M — a bullish call roll: closing stale July $200 calls (bought back in May, now deep underwater after NVDA's ≈18% pullback) and opening fresh Sep $195 calls that capture the Aug-26 earnings print. The short July leg expires before earnings; the long Sep leg owns it.
• Cleanest bull bet: ⚙️ SOXX $21.5M — a $670/$820 Sep bull-call spread into the semis-earnings gauntlet (ASML Jul 15, TSMC Jul 16, NVDA Aug 26, AVGO Sep 3).
• The contrarian: 🐻 SHAZ $0.54M — a defined-risk bear put spread on SharonAI, a pre-revenue AI-cloud IPO up ≈127% YTD that just did a dilutive $1.6B raise. The rare downside bet of the day.
• The four LEAP crosses (direction unprovable): 🟠 IBIT $6.7M sold 2028 Bitcoin calls · 🛒 PDD $3.05M sold 2027 $90 calls (Temu/Pinduoduo) · 🍽️ TOST $2.18M 2027 $35 calls (was weak buy-lean, size<OI — now resolved: net CLOSE, OI fell −3,398). Plus 🗑️ GFL $3.05M bullish call ladder into Q2 earnings (late-reported tape).

The theme: AI/semis (NVDA, SOXX, QQQ, SHAZ), Bitcoin (IBIT), China e-commerce (PDD), restaurant fintech (TOST), waste (GFL). The discipline: even the readable structures are bets, not certainties — and the four crosses aren't signals at all.

---

🔁 OI Review — Last Session's Provisional Flags, Now Resolved (covers 2026-07-06)

The next-day OPRA open-interest snapshot for the 2026-07-06 session is in (posted 2026-07-07 pre-market). We pulled it for all 11 legs across 10 names and resolved every ⏳ provisional flag. The headline: the two loudest "someone sold a big block" prints both INVERTED to net CLOSES — neither SMH's short put nor TSLA's ITM call sale opened a fresh position; both net-reduced open interest.

🔄 Inversions

• SMH — "bullish premium-collection short put (STO)" → net CLOSE. The 40,000-lot $540 put block printed with size below prior OI (54,000), so open-vs-close was unprovable intraday. Next-day OI FELL 53,998 → 41,114 (−12,884). A fresh short-put write would have driven OI up toward ≈94,000; instead it dropped nearly 13k — so on net this SELL reduced open interest (most likely long puts sold to close ahead of ASML Jul 15 / TSMC Jul 16), NOT a new "get paid to buy SMH at $540" bet. The bullish framing is superseded; being a cross, the specific side that closed is unprovable.
• TSLA — "ITM call write (STO)" → net CLOSE / transfer. The ≈9,681-lot $395 call sale (multi-leg auction) printed below prior OI (12,000). Next-day OI FELL 12,105 → 11,232 (−873). It did not open a fresh short call; net OI declined, and the tiny net change versus the printed size means the volume overwhelmingly passed between existing holders (transfer). No large new short-call position was created.

✅ Confirmations

• INTC — the one "provisional (size < OI)" name that resolved to a net OPEN: OI rose 3,086 → 5,037 (+1,951) despite the 2,700 block being smaller than prior OI, so the cross largely created new contracts. Direction still unprovable (cross).
• SNDK $1,750 put OPENED (889 → 2,751, +1,862 ≈ block, BTO) — the day's one clean lit long-put/hedge, confirmed fresh.
• ROST $185 put OPENED (76 → 5,818, +5,742 = block, STO) — premium-collection income write confirmed.
• NBIS Dec-2028 $340 LEAP call OPENED (60 → 1,510, +1,450 ≈ block) — direction unprovable (cross). MRNA Jan-2027 $80 put OPENED (429 → 3,927, +3,498) — defensive-lean. ECHO Sep-2026 $100 call OPENED (824 → 4,891, +4,067) — deal-close bullish-lean.
• STX bull-call ratio built fresh (both legs opened). BIIB deep-ITM $160 financing combo OPENED (+200).

What changed: the two biggest "smart-money sold a block" headlines — SMH and TSLA — both turned out to be closes, not bullish premium-collection writes. The standing lesson, again: a big SELL headline is never settled — and never "bullish income" — until next-day OI proves the position actually OPENED.

---

---

📋 Today's Flow at a Glance

Net premium collapses multi-leg spreads to their net debit/credit (e.g. SOXX gross ≈$35.5M across both legs; ≈$21.5M net).

---

🔬 The Eight Positions — Decoded

1. 🤝 QQQ — The $29.9M LEAP Call Block Cross (direction unprovable)

SEE THE $30M NASDAQ-100 LEAP CALL SALE NOBODY CAN PROVE IS BEARISH →

• Flow: ≈$29.9M — sold 3,250 Sep 17 2027 $720 calls (LEAP) at $92.05 as a block cross. OI 615, size 3,250 → fresh open. A cross has no aggressor → direction unprovable (covered-call/overwrite, financing, or short-vol).
• Context: $720 is the exact gamma-resistance-wall strike, ≈2% above spot. Sits ≈14 months past every scheduled catalyst — this is a vol/positioning trade, not an event bet.

2. 🐋 NVDA — The $25.1M Bullish Call Roll Into Earnings (the smart-timing trade)

SEE HOW A DESK ROLLED A LOSING NVDA CALL INTO A FRESH EARNINGS BET →

• Flow: ≈$25.1M net debit — a multi-leg auction that closed ≈21,005 July $200 calls (bought back in May, now worth $2.38 after NVDA's ≈18% pullback) and opened ≈21,005 Sep 18 $195 calls at $14.35.
• Why it's smart: the short July leg expires before NVDA's Aug 26 earnings; the fresh Sep $195 calls own that print. It's a bullish reposition — rolling down-and-out to a nearer strike and a longer tenor to re-arm for the catalyst.

3. ⚙️ SOXX — The $21.5M Bull Call Spread Into the Chip-Earnings Run

SEE THE $21M SEMIS CALL SPREAD BETTING ON THE EARNINGS GAUNTLET →

• Flow: ≈$21.5M net debit — long ≈14,476 Sep 18 $670 calls ($19.68) / short ≈14,476 Sep 18 $820 calls ($4.85). Both fresh opens. Spot $539.78.
• Read: clean bullish bet targeting the $670–$820 zone (≈24–52% above spot — aggressive) into ASML (Jul 15), TSMC (Jul 16), NVDA (Aug 26) and AVGO (Sep 3). SOXX is ≈+73% YTD.

4. 🟠 IBIT — The $6.7M Bitcoin 2028 LEAP Call Cross (SELL, unprovable)

SEE THE $6.7M BITCOIN LEAP CALL SALE — DIRECTION UNPROVABLE →

• Flow: ≈$6.7M — sold 5,000 Dec 15 2028 $35 calls (≈2.4yr LEAP) at $13.33 (mid) as a block cross. OI 3,000, size 5,000 → fresh open. Direction unprovable.
• Context: IBIT is ≈−29% YTD; BTC fell below $60K on a hawkish Fed, and June saw record spot-BTC-ETF outflows. Selling long-dated upside fits an overwrite/short-vol posture — but a cross can't confirm the side.

5. 🗑️ GFL — The $3.05M Bullish Call Ladder (late-reported tape)

SEE THE $3M BULLISH BET ON GFL BEFORE Q2 EARNINGS →

• Flow: ≈$3.05M — bought 10,000 Aug 21 $45 calls ($1.80) + 10,000 Aug 21 $50 calls ($1.25). Both fresh opens. Spot $40.81.
• Caveat: these printed late/out-of-sequence on a thinly-optioned name with a degraded quote — the bullish direction is clear, but mechanism confidence is lower than a clean lit trade. Positioned for Q2 earnings (Jul 29/30) and a possible SECURE $6.4B deal close.

6. 🐻 SHAZ — The $0.54M Bear Put Spread (the day's contrarian)

SEE THE DEFINED-RISK BEARISH BET ON A HOT AI-CLOUD IPO →

• Flow: ≈$0.54M net debit — long 982 Aug 21 $70 puts ($11.37) / short 982 Aug 21 $60 puts ($5.84). Breakeven ≈$64.47. Spot $74.77.
• Read: a bearish, defined-risk bet that SharonAI (an AI/GPU-cloud IPO up ≈127% YTD, pre-revenue, fresh off a dilutive $1.6B raise) unwinds toward $60–$70. Note the big scheduled events (Aug 27 meeting, Sept 1 revenue start) land after expiry — this leans on drift, not a dated catalyst.

7. 🛒 PDD — The $3.05M LEAP Call Cross (SELL, unprovable)

SEE THE $3M LEAP CALL SALE ON THE TEMU/PINDUODUO STORY →

• Flow: ≈$3.05M — sold 2,500 Jun 17 2027 $90 calls (LEAP) at $12.20 as a block cross. OI 575, size 2,500 → fresh open. Direction unprovable (cross).
• Context: PDD is ≈−28.7% YTD amid EU de-minimis repeal (Jul 1), a €200M EU fine, and a China take-rate war. Selling upside calls fits an overwrite or an upside-cap/fade — but a cross can't confirm the side.

8. 🍽️ TOST — The $2.18M LEAP Call Cross (❗ RESOLVED: net CLOSE / unwind)

SEE THE $2.2M RESTAURANT-FINTECH CALL BLOCK — RESOLVED AS A CLOSE →

• Flow: ≈$2.18M — 7,500 Jan 15 2027 $35 calls (LEAP) crossed at the offer ($2.90). The at-offer print looked like a weak bullish lean, and size (7,500) < OI (21,000) left open vs. close unprovable intraday.
• ✅ Resolved (Jul 8 OI): open interest FELL 21,148 → 17,750 (−3,398) — the opposite of a fresh open. This was a net CLOSE / partial unwind of an existing $35 call position (profit-take/de-risk after TOST's ≈34% rally off the trough), not new bullish accumulation. The at-offer print carried no directional signal.
• Context: Toast (restaurant POS/fintech) reports Q2 ≈Aug 4 and Q3 ≈Nov; ToastIQ AI monetization is the growth story — but this specific flow was someone trimming, not building.

---

👥 How Four Different Traders Should Read Today

🎲 YOLO Trader

The cleanest convex setups are SOXX (a ≈24%-OTM bull call spread — cheap, defined-risk lottery on the chip-earnings run) and SHAZ (a bear put spread on a parabolic IPO — ≈0.8:1 reward:risk, so size it as a hedge, not a jackpot). QQQ/IBIT/PDD/TOST are sold-/crossed LEAP-call blocks you can't chase (direction unprovable). Size tiny; these are bets, not signals.

📈 Swing Trader

NVDA is the template — a desk rolled down and out to re-arm for Aug 26 earnings; wait for a pullback toward the $185–$190 gamma support rather than chasing. SOXX gives you defined-risk exposure to the same semis run. Let ASML (Jul 15) and TSMC (Jul 16) be your first read on whether the chip-earnings thesis is intact before adding.

💵 Premium Collector

QQQ, IBIT, and PDD are your playbook at institutional scale — selling long-dated calls to harvest premium. If you own the shares, a small covered call mirrors it; never sell naked calls on a melt-up. Respect that all four of today's LEAP-call blocks (QQQ, IBIT, PDD, TOST) are crosses — you don't know if they're bullish overwrites or bearish fades.

🌱 Beginner

Today splits evenly: four are provable directional structures (NVDA bullish roll, SOXX bullish spread, GFL bullish calls, SHAZ bearish spread) and four (QQQ, IBIT, PDD, TOST) are unprovable LEAP-call crosses. But "readable" still means "someone's bet, not a sure thing." Notice how yesterday's two loudest sales (SMH, TSLA) both turned out to be closes once next-day OI printed — the number is never the whole story. Watch, learn the structures, and don't copy size you can't afford to lose.

---

⚠️ Risk Control & Patience — Read This Before You Copy Anything

• A directional structure is still a bet. NVDA's roll, SOXX's spread, and GFL's calls are readable — but they can all lose. SOXX needs a ≈24% move just to reach its long strike's payoff zone.
• Crosses aren't signals. QQQ, IBIT, PDD and TOST are LEAP-call crosses with a known counterparty — you can't tell bullish overwrite from bearish fade off the headline dollar figure. TOST is the cautionary example: next-day OI showed its "big call" cross was actually a net close/unwind, not a buy.
• Open vs. close is everything. Yesterday's SMH and TSLA "sales" both resolved as closes, not fresh bullish writes. We resolve every ⏳ flag in the next OI Review — that's the discipline.
• Tenor matters. GFL/SHAZ expire in ≈6 weeks (Aug), NVDA/SOXX are Sep, and QQQ (2027) / IBIT (2028) are LEAP-horizon. Match the trade to your patience.

Tenor tags: 📅 Monthly (Aug): GFL, SHAZ · 🍂 Quarterly (Sep): NVDA (Sep leg), SOXX · 🚀 LEAP: QQQ (2027), PDD (2027), TOST (2027), IBIT (2028)

---

Every classification above is read from the real OPRA tape (mechanism, size vs. open interest, aggressor), not from a screenshot label. Provisional (⏳) flags are resolved in the next session's OI Review. Educational flow analysis, not investment advice — trade your own plan and manage risk.