📊 Ainvest Option Flow Digest - 2026-07-02 — "Looks Big, Isn't Directional" — Except AAPL: ≈$38M of Delta-Hedged Crosses, a $21M KRE Bank Structure (next-day OI overturned the bullish read), and a $14M Apple Call That's the Real Thing

✅ Updated 2026-07-06 — next-day OPRA OI (reflecting July 2 EOD) resolved every ⏳ flag from this session. Two reads INVERTED: EOSE — the "unprovable" deep-ITM put sale OPENED (+20,007), a confirmed fresh ≈$12M bullish income write (STO), not a close. KRE — the "moderately bullish capped call ladder" is OVERTURNED: only the $78 short-call leg opened (+56,071); the long $70 "bullish base" and the $80 short both saw OI fall, so the bullish base is unconfirmed — the one provable position is a short-call write (neutral-to-mildly-bearish / income). Also: BTDR split (Jul-31 opened; Jul-17 leg's OI fell = a net close/cover, not fresh bullish); MU put opened / call flat (conversion confirmed, non-directional); SOFI opened (BTO, still delta-hedged); XLI hedge confirmed ($165 long + $166 short opened, $155 short flat/transfer — no inversion); AAPL/DOCN/FLEX/GDX/SMH/VRDN all confirmed OPEN. Per-ticker details in each article's ✅ RESOLVED box.

Twelve names, ≈$87M in net options premium (≈$134M gross) — and for the third day running, most of the biggest single-stock tickets are not directional bets. We pulled the OPRA tape and the stock tape on every ticket, and ≈$38M of today's headline flow is delta-hedged: SMH's ≈$19M "put sale" (matched 165,000-share block, 97.5%); MU's ≈$15.6M is a full conversion (paired with a 50,000-share block — long stock + synthetic short = delta-neutral financing); SOFI's ≈$2M LEAP "call buy" (446,775-share block, 95.4%); and FLEX's ≈$1.2M deep-OTM put (26,000-share block, 98.7%). None is a view on the stock. The exception — and the day's cleanest conviction — is AAPL: ≈$14M of Dec $330 calls bought at the ask (lit, took liquidity), with NO stock hedge — a genuine bullish bet into Apple's July 30 earnings + iPhone-18/Siri-AI window. Around it: a ≈$21M net call structure on KRE (looked like a moderately-bullish capped ladder — but next-day OI overturned that; see the update banner above), a ≈$12M deep-ITM EOSE put sale (next-day OI confirms it OPENED — a fresh bullish income write), a downside XLI put-spread hedge on industrials, a post-FDA VRDN put-write, and cheap bullish calls on BTDR (the fresh piece ≈$2.8M), DOCN, GDX. The lesson, again: a big premium headline means nothing until you check for the stock leg — and until next-day OI confirms open vs. close. 🧭

🔁 OI Review — Last Session's Provisional Flags, Now Resolved (covers 2026-07-01)

The next-day OPRA open-interest snapshot for the 2026-07-01 session is in. We pulled it for all 22 legs across 14 names and resolved every ⏳ provisional flag. The headline: the big directional and hedge reads OPENED as written, but 2 legs INVERTED — and both trimmed a "fresh bullish/short" read down to something more cautious.

🔄 Inversions

• KEEL — the ≈$5.9M "dual OTM-call bullish bet" ($7 + $10 calls) split on resolution. The $10 call OPENED fresh (OI 32,385 → 61,691, Δ +29,306) — the genuine new bullish position — but the $7 call NET CLOSED (OI 63,903 → 60,031, Δ −3,872), despite printing as a BUY. Best read: the buyer covered/rolled an existing $7 position, not added fresh exposure. The fresh bullish premium at risk is the $10 leg only. KEEL
• SPY — the ≈$14M net put-spread hedge confirmed on its long legs but inverted on its near-dated short. Both LONG protection legs OPENED (Jul-31 $743 +25,155; Oct-16 $680 +9,955), so the downside-umbrella thesis holds. But the Jul-17 $740 short leg NET CLOSED (OI 60,358 → 54,207, Δ −6,151) — monetizing/closing prior longs, not a fresh bearish short-put write. Net: even more clearly portfolio hedging. SPY

✅ Confirmations

ASHR $33 put opened (+12,846). DASH $230 call opened while the $200 call closed as flagged. EOSE $10 put-write opened big (+30,017). IWM both put legs opened ($290 +47,823; $285 +5,535). META both call legs opened ($1,000 +5,484; $700 +1,689). MSFT $420 call-write opened (+4,055). NEM $99 call opened (+5,351). NKE $55 call opened (+9,001). PM $175 put opened (+2,000). PRCH short $17.5 straddle opened (put +4,998, call +5,003). QCOM $210 put-write opened (+1,489, still a delta-hedged financing package). XOP $165 call opened (+3,501).

What changed: next-day OI trimmed two of the day's reads — KEEL was a fresh bet on the $10 leg only (the $7 leg was a close/roll), and SPY's near-dated $740 short was a close, not a fresh bearish write. The standing lesson: a big BUY or SELL headline tells you nothing about open vs. close until next-day OI resolves it — this week alone, OI has inverted legs on KEEL, SPY, TSM and IGV.

⚡ The Quick Read

Today's tape spanned 12 names for ≈$87M net premium — but the headline sizes are deceiving. Five buckets:

1) The "looks big, isn't directional" trades (≈$38M, but delta-neutral). The biggest single-stock tickets each carry a matched stock leg:

• SMH — a ≈$19M SELL of Sep $550 puts, paired with a 165,000-share block (97.5% delta match). A financing/vol package, not "a desk getting long semis."
• MU — ≈$15.6M of "buy $1150 put + sell $1150 call" is a synthetic short, but it printed with a 50,000-share long stock block → long stock + synthetic short = a conversion, fully delta-neutral. A financing/arbitrage trade, not a bearish Micron bet.
• SOFI — a ≈$2M BUY of 2027 $25 calls, paired with a 446,775-share block (95.4% match). Delta-neutral, not a directional LEAP wager.
• FLEX — a ≈$1.2M BUY of deep-OTM Nov $90 puts (≈79% IV), paired with a 26,000-share block (98.7% match). A vol/skew trade, not a crash bet.

2) The day's REAL conviction — AAPL (≈$14M). AAPL stands apart from the hedged crowd: ≈$14M of Dec $330 calls bought at the ask — a lit trade that took liquidity, with no stock hedge. This is genuine, aggressive bullish demand (breakeven ≈$345), placed into Apple's July 30 fiscal-Q3 earnings, the Sept 9 iPhone 18 Pro launch, and the Gemini-powered Siri-AI reset. When a big print is lit, at the ask, and unhedged, that's the real thing — the opposite of today's delta-neutral crosses.

3) The one big "directional" bet that OI overturned — KRE (≈$21M net). KRE (regional banks) printed a three-leg $70/$78/$80 call structure that looked like a moderately-bullish capped ladder (BUY $70, SELL $78/$80). But next-day OPRA OI overturned that read: only the $78 short-call leg opened cleanly (+56,071); the long $70 "bullish base" and the $80 short both saw OI fall (−12,331 / −11,508), so the bullish base is NOT confirmed as a fresh open. The only OI-provable position is a large ≈$7.9M short-call write at $78 — a neutral-to-mildly-bearish / income posture, not the grind-higher bull bet the raw print suggested. Treat the directional lean as unproven.

4) Income put-writes + a downside hedge (≈$17M). EOSE — a ≈$12M SELL of 2028 $10 puts on the ≈$5.43 battery microcap. Size (19,778) sat below prior OI (31,000), so we could not prove open vs. close intraday — but next-day OI RESOLVED it as an OPEN (STO): OI rose +20,007, a confirmed fresh bullish income write at an effective ≈$3.90 entry (this inverts the classifier's provisional "close" lean). VRDN — a ≈$1.6M SELL of Oct $20 puts on the biotech that just got FDA approval (Lumvoa, Jun 26), opening — a "willing to own it at ≈$16.75" income bet. And XLI went the other way: a ≈$3.6M net Sep $165/$155 bear put spread (downside protection ≈11–16% below spot) plus an Aug $166 put sale — a defensive industrials hedge into the GE/RTX/Boeing/CAT earnings cluster.

5) The small, genuinely bullish calls (≈$9M). BTDR — ≈$4.9M of cheap ≈32%-OTM $19 calls (the Jul-31 leg a fresh open), a convex bet on Bitdeer's AI-datacenter pivot. DOCN — a ≈$2M near-dated $140 call on the AI-cloud name. GDX — a ≈$1.2M net $79/$82 bull call spread on gold miners, capped, ≈8 days to run.

Translation: ≈$38M of today's "whale" premium is delta-neutral financing and only ≈$14M of the biggest tickets (AAPL) is a genuine, unhedged directional bet. The rest is KRE's capped bank ladder, an industrials hedge (XLI), income put-writes, and small-cap bullish calls. When the biggest tickets carry a stock leg, respect the hedge — the headline size is not the signal. 🧭

👀 Today at a Glance

Premiums are per-structure. Day total ≈$87M net across 12 names (≈$134M gross; KRE/GDX/MU/XLI shown net-of-financing). ≈$38M (SMH/MU/SOFI/FLEX) is delta-hedged and non-directional; the genuine directional conviction is AAPL's ≈$14M unhedged call buy, plus an XLI industrials hedge, small bullish calls (BTDR ≈$2.8M fresh / DOCN / GDX) and income put-writes (EOSE, VRDN). ✅ All ⏳ flags are now RESOLVED via the July 6 pre-market OPRA OI (reflecting July 2 EOD) — see the update banner at the top. Two reads inverted: EOSE resolved as a fresh OPEN (bullish income write), and KRE's "bullish ladder" was overturned (only the $78 short opened). July 3 was the Independence Day market holiday, so the resolving snapshot posted Monday July 6.

🔬 The Standouts, In Plain English

🍏 AAPL — The One Big Trade That IS a Real Bet

Amid a tape full of hedged crosses, AAPL is the standout for the opposite reason: it's genuine. A desk bought ≈$14M of December $330 calls lifting the ask — a lit trade that actually consumed offers (unlike a negotiated cross), with no paired stock block anywhere in the equity tape. That combination — lit, at the ask, unhedged, and opening (size 9,306 vs prior OI 7,100) — is the signature of real, aggressive bullish conviction. The bet needs Apple above ≈$345 by December, and it's placed squarely across the busiest stretch of Apple's calendar: fiscal-Q3 earnings July 30, the iPhone 18 Pro launch September 9, and the Gemini-powered Siri-AI reset this fall. This is what conviction looks like when a desk isn't hiding behind a hedge.

⚖️ SMH, MU, SOFI & FLEX — "Whale Trades" With a Stock Leg on the Other Side

The now-familiar lesson: check for the stock leg. Four of today's big single-stock tickets are delta-hedged packages:

• SMH sold ≈$19M of $550 puts and a desk moved 165,000 shares (97.5% match) — net delta ≈ zero. Not a semis bet.
• MU is the cleanest example of the day: buying the $1150 put and selling the $1150 call is a synthetic short — but it printed with 50,000 long shares, which turns the whole thing into a conversion (long stock offset by synthetic short = delta-neutral). This is a financing/arbitrage trade — a way to lock in a rate or manage borrow — not a bet that Micron falls after its record quarter.
• SOFI bought ≈$2M of 2027 $25 calls against 446,775 short shares (95.4% match) — delta-neutral, not a directional LEAP.
• FLEX bought ≈$1.2M of deep-OTM Nov $90 puts (≈79% IV, all time value) against 26,000 shares (98.7% match) — a vol/skew trade, not a bet the AI-manufacturing name crashes ≈34%. Same headline ("$19M put sale!", "$15.6M options block!"), completely different meaning once the paired stock print is on the tape.

🏦 KRE — The "Bullish Ladder" That Next-Day OI Overturned

On the raw print, KRE's ≈$21.2M net three-leg call structure (BUY 54,267 Aug $70 calls, SELL the $78 and $80 calls) looked like a moderately-bullish capped ladder. Next-day OPRA OI told a different story. Only the $78 short-call leg is confirmed as a fresh open (OI 3,076 → 59,147, +56,071 ≈ the trade size). The long $70 "bullish base" leg's OI fell (57,288 → 44,957, −12,331) and the $80 short also fell (57,225 → 45,717, −11,508) — the opposite of what a 56,000-contract fresh long/short would produce. So we cannot confirm the bullish long-$70 base opened; those legs were heavily offset by existing holders closing/transferring, possibly a restructure of a pre-existing position. The one OI-provable fresh position is the large $78 short-call write — which caps/fades upside above $78, a neutral-to-mildly-bearish / income posture. Bottom line: the "regional banks grind higher" read is unproven; respect only the confirmed short-call write.

💰 EOSE & 💊 VRDN — Getting Paid, and Both Now Confirmed OPEN

Two desks sold puts to collect premium — and next-day OI confirms both opened. VRDN is the clean one: ≈$1.6M collected on the Oct $20 puts of a biotech that just won FDA approval for its thyroid-eye-disease drug (Jun 26) and trades near its lows — a "willing to own it at ≈$16.75" bet; OI rose 6 → 5,006 (+5,000 = trade size), a textbook STO open. EOSE was the intraday question mark: ≈$12M on the 2028 $10 puts with size (19,778) below prior OI (31,000), so open-vs-close was unprovable on the tape — and the classifier even leaned "close." Monday's OI resolved it decisively: OI rose +20,007 → a fresh OPENING income write (STO), not a reduction. So EOSE joins VRDN as a confirmed bullish put-write; the feared KEEL/SPY-style inversion to a close did not happen here.

📅 Upcoming Catalysts — Mind the Gap Between the Event and the Expiration

🧑‍🤝‍🧑 For Four Kinds of Traders

🎰 YOLO / high-risk: The cheap convex tickets are BTDR (≈32%-OTM $19 calls on the AI-datacenter pivot — needs the Tydal lease to land) and GDX's ≈8-day $79/$82 spread (a fast gold-miner bounce, capped). Both are small-dollar, high-variance, and likely expire worthless if the move doesn't come quickly. Do not chase SMH/MU/SOFI as "whale conviction" — they're hedged packages with no directional edge to follow.

📈 Swing trader: The posture read is clear — the mega-cap flow is delta-neutral financing, and the one big trade that looked directional, KRE, was overturned by next-day OI (only the $78 short-call write is confirmed — an upside-fade / income posture, not a grind-higher bull bet). So the cleanest genuinely-bullish single-name expressions with defined risk are AAPL's unhedged Dec $330 call (confirmed open, the day's real conviction), DOCN's near-dated call (respect the ≈2-week clock), or GDX's defined-risk spread.

💰 Premium collector: Today is a clinic in selling with a plan. VRDN sold puts on a just-de-risked name at a real discount (≈$16.75 effective); SMH/MU sold/structured premium inside delta-hedged packages (income without a naked directional bet); KRE used two short calls to finance a long call. The through-line: none sold naked into a catalyst. If you sell premium, cash-secure or hedge it — and remember EOSE shows that a deep-ITM option's headline price is mostly time value.

🌱 Beginner / new to flow: Today's lesson is the same one that keeps paying off: a giant "put sale" or "call buy" is often NOT a bet on the stock. SMH's "$19M put sale," MU's "$15.6M options block," and SOFI's "$2M call buy" each came with a matching stock trade that cancels the option's directional punch. Two habits beat the headline: ask whether there's a stock leg on the other side (a cross usually means yes), and check whether a "big SELL" is actually opening or closing (size below open interest, like EOSE today, means you can't tell until tomorrow's OI).

⚠️ Risk & Patience — Read This Before You Trade Any of It

• The biggest tickets are delta-hedged — do not read them as direction. SMH/MU/SOFI each carry a matched stock leg (95–98% delta match). The headline premium is financing/positioning, not conviction.
• KRE was NOT the bullish ladder it looked like. Next-day OI confirmed only the $78 short-call leg opened; the long $70 base and $80 short both saw OI fall. The one provable position is a short-call write (neutral-to-mildly-bearish / income) — do not treat KRE as confirmed bullish regional-bank conviction.
• Selling premium carries open-ended risk. EOSE's and VRDN's short puts can lose far more than the credit if those volatile names fall hard — both are speculative small/micro-caps.
• Open vs. close is now RESOLVED on every leg (July 6 OPRA OI). EOSE OPENED (+20,007, inverting the "close" lean); KRE's bullish base did NOT open (only the $78 short did); BTDR's Jul-17 leg's OI fell (net close/cover) while Jul-31 opened; MU's put opened / call flat; SOFI opened; XLI's hedge confirmed. This week alone, next-day OI inverted legs on KEEL, SPY, TSM, IGV, EOSE and KRE — proof that a trade-day headline is never settled until the OI prints.

Not investment advice. Options carry risk of total loss, and selling options can lose far more than the premium collected. Unusual flow is a starting point for research, not a recommendation. Always do your own work and size positions you can afford to lose.