When you want clean exposure to "everything semis going up" without picking which name leads, you buy SOXX. When you want it sized for the cycle and not the next print, you buy the calls long-dated. The most interesting sector-ETF print on our tape this year did exactly that.
On March 13, 2026 — four trading days before NVIDIA's Q1 GTC keynote and three weeks before the Apr 28–30 mega-cap earnings cluster — an institution bought 2,000 contracts of the SOXX 350-strike call expiring January 15, 2027 at $44.15 per contract. Total premium: $8.8 million. The strike was ~5% above SOXX's $331.32 spot at the time. The DTE was 10 months — through every 2026 quarterly earnings season for every major semis name.
By April 29, 2026, SOXX has rallied to $450 (intraday peak $463.87 on April 24, +40% from entry) and the 350-strike call is marked at $119.81 (peak $141.00) — a +171% return on the option, currently sitting on +$17.7 million unrealized.
First published: Daily Institutional Flow Digest, March 13, 2026 · SOXX flow on 2026-03-13.
The print
| Field | Value |
|---|---|
| Date | 2026-03-13 |
| Symbol | SOXX |
| Side | BUY |
| Type | CALL |
| Strike | 350 |
| Expiration | 2027-01-15 |
| Volume | ~2,000 contracts |
| Premium | $8.8M |
| Entry option price | $44.15 |
| SOXX spot at trade | $331.32 |
| Position confidence | VERY_HIGH |
| Outcome | open (~$119.81 mark) |
| Unrealized P&L | +$17.7M |
A few notes on the structure:
- Strike ~5% OTM. Tight enough that any sustained sector rally pushes the option in the money quickly; loose enough that the entry premium isn't dominated by intrinsic.
- 10-month DTE. Captures NVDA Q1, Q2, Q3 reports + AVGO + MU + AMD + earnings of every major semis name through year-end.
- $8.8M sized. Modest for an institution but meaningful for a focused single-leg ETF call. Sector-ETF UOA prints over $5M with VERY_HIGH classification are uncommon.
What SOXX did
| Date | SOXX close | Move from 3/13 |
|---|---|---|
| 2026-03-13 (entry) | $331.32 | — |
| 2026-03-25 | $355 | +7% |
| 2026-04-04 (post-NVDA pullback) | $328 | -1% |
| 2026-04-15 | $389 | +17% |
| 2026-04-24 (peak) | $463.87 | +40.0% |
| 2026-04-29 (today) | $449.99 | +35.8% |
SOXX did not rally in a straight line. The first three weeks were chop — NVDA's GTC keynote on March 17 was well-received but the broader semis tape sold off into the AMD March 5–11 cooling-off. Then VRT, Lattice, and Coherent were added to the S&P 500 on March 24 (validating the AI data-center thesis), Micron's HBM4 ramp commentary on April 8 reinforced the memory cycle, and Broadcom's April quarter pre-announce on April 14 pushed the ETF to its peak.
What the option did
| Date | 350C 2027-01-15 close | Move from entry |
|---|---|---|
| 2026-03-13 (entry) | $44.15 | — |
| 2026-03-25 | $58 | +31% |
| 2026-04-04 | ~$42 | -5% |
| 2026-04-15 | $97 | +120% |
| 2026-04-24 (peak) | $141.00 | +219% |
| 2026-04-29 (today) | $119.81 | +171% |
The option is up 171% on a +35.8% underlying move — that's ~4.8x the leverage you'd get holding SOXX directly. With 10 months still on the clock and a long-running AI capex cycle, the position has years of runway left.
Why SOXX rallied
The catalyst chain after the March 13 print:
- March 17 (NVDA GTC keynote) — Blackwell roadmap, sovereign-AI partnerships, networking architecture details.
- March 24 (S&P 500 rebalance) — VRT, LITE, COHR added; AI-cooling and AI-networking names get index buying.
- April 8 (Micron HBM4 ramp commentary) — confirms 2026 HBM supply is fully booked.
- April 14 (AVGO pre-announce) — Q3 AI-revenue beat raises full-year guidance.
- April 23 (INTC Q1) — beat with foundry losses narrowing; semis breadth widens.
None of those individual events were "the catalyst" — collectively they validated the thesis that 2026 AI-capex commitments translate into semis revenue across the sector. SOXX captured the basket effect; an OTM long-dated call on the basket captured the convex move.
Why this trade is a "thesis-based, no-single-catalyst" example
Most flashback-able UOA prints we've published anchor to a discrete event — earnings, FOMC, gap day. The SOXX 350C 3/13 is the opposite: an institution paid for a 10-month window and let the cycle be the catalyst.
When you see institutional sector-ETF longs with multi-quarter DTE, the smart money usually isn't picking a winner — they're betting the basket re-rates. With AI-capex now disclosed at $700B+ across the four hyperscalers for 2026 alone, the SOXX basket-re-rate trade had a long runway when the institution sized into it.
What this trade did NOT mean
The institution who bought the 350C 1/15/27 didn't have non-public information. They had a thesis on the AI-capex super-cycle, sized to a long-dated sector ETF call. Most UOA prints don't print like this — read the methodology piece for the honest aggregate stats.
What we're watching
The position is still open with 8+ months to expiration. Forward catalysts:
- NVDA Q1 FY27 earnings (May 2026) — first read on Blackwell ramp.
- AVGO Q2 earnings (June 2026) — AI revenue trajectory.
- Q3 hyperscaler capex updates (July 2026) — confirmation that 2026 spend is on track.
When the institution closes, the print will show up on the tape — and we'll surface it.
See the live tape
Upgrade to AIme Premium → to read sector-ETF positioning the morning it prints — institutional basket bets are often early signals for sector breadth moves.
Read more
Other flashback case studies — INTC · ARM · NVDA · AMD · AVGO · MU.
See This Analysis Live — Free
Sign up free to access the full options screener with 5-pillar scores for 5,000+ stocks, daily signals, strategy recommendations, and radar charts. No credit card required.
Free account includes: screener · 5-pillar scores · daily signals · strategy picks · radar charts
Or just get the weekly recap
Sundays. What moved this week, what catalysts and earnings drive next week, and the 5-pillar setups that stand out. No account needed.
Free. One email per week. Unsubscribe with one click.