In partnership with
Welcome to the iPrompt Newsletter
Friday Edition
What you get in the Friday Edition
Weekly Scoreboard
Top Headlines of the Week
Our Investing Angle
Three Ideas to Research This Weekend
AI Investment Framework
Your Move
…all in a FREE Weekly newsletter.
Sponsor:
2026 State of AEO Report
A year ago, most marketers weren't thinking about AI search. Now it's one of the fastest moving channels in the industry and nobody has a playbook yet.
So we built one. We surveyed hundreds of marketers to find out how they're approaching answer engine optimization, where they're investing, what's actually working, and what isn't.
The result is the 2026 State of AEO Report. Real data. Real strategies. A clear picture of where AI search is headed and how to get ahead of it.
iPrompt Signals
AI & ROBOTICS INVESTING — EXPLAINED SO YOU CAN ACTUALLY ACT ON IT
Issue 13 · Friday 5 June 2026 · R. Lauritsen
On Tuesday, Jensen Huang said five words on a stage in Taipei — “the next trillion-dollar company” — and Marvell added a third of its value in a day. On Wednesday, Broadcom grew AI revenue more than 200% and lost roughly an eighth of itself. Same week, same sector, opposite physics. The compliment soared; the blowout sank. Once you see why, you can’t unsee it — and it gives you a clean filter for which AI names are backed by numbers and which are running on noise.
Weekly Scoreboard
Ticker | Price | Week % | What happened |
NVDA | $216 | −0.6% | Barely moved — held up best through the selloff. |
AVGO | $285 | −13% | Grew AI 200%+ but didn’t raise guidance. Sold off hard. |
MRVL | $295 | +30% | Up a third on one Huang sentence at Computex. |
MU | $233 | −8% | Last week’s trillion-dollar memory star, caught in the chip rout. |
CRWD | $680 | −9% | Beat earnings, raised guidance, announced a split — still fell. |
S&P 500 | 7,560 | −0.7% | Barely moved. The damage stayed inside semis. |
VIX | 15.4 | — | The “fear gauge” (30-day S&P swings; >25 = real anxiety). Calm. |
Bottom line: a 13% drop in the biggest chip names while the VIX sat at 15. That gap — violent in semis, calm everywhere else — is the whole story, and we unpack it below.
Prices are Thursday 4 June closes, rounded; weekly moves vs the prior Friday close.
Top Headlines
🌱 New to investing? One word that explains the week: A “whisper number” is the unofficial result big investors quietly expect — usually higher than the published analyst forecast. A company can beat the official forecast and still fall if it misses the whisper. Think of a student who’s scored 95% all term: a 91% looks like a slip, even though it’s a great mark. After a long rally, the AI names are all 95%-students. |
1. Broadcom grew AI 200% — and that was the problem
Broadcom reported AI chip revenue up more than 200% year-on-year and guided next quarter to $16bn. Strong by any normal standard — but it sat below the whisper, and management reiterated rather than raised the full-year outlook. The stock fell 12–15% and took Micron, Marvell and Arm down with it. What changed: not demand — the bar. After a two-month melt-up, “great” only clears if it’s “better than the rumour.”
2. CrowdStrike beat, raised, split its stock — and dropped 9%
Revenue up 26%, net new ARR up 32%, full-year guidance lifted, a 4-for-1 split, record free cash flow. A clean win on paper — and the stock fell about 9% after hours, because a 60%+ YTD run had already priced in more than clean. Why beat-and-raise didn’t save it: when a stock is priced for perfection, perfection becomes the floor, not the ceiling.
3. One sentence added $80bn to Marvell
At Computex, Huang called Marvell — maker of the connectivity chips that stitch AI data centres together — “the next trillion-dollar company.” The stock jumped 32% in a session, its largest ever, then extended again the next day. Jim Cramer, no bear, called the move “concerning… not based on anything other than one person saying it.” What pure sentiment looks like: a single quote moving $80bn of market cap with no new numbers attached.
4. Nvidia put a humanoid robot on the research bench
Away from the share-price drama, Nvidia unveiled Isaac GR00T — an open humanoid reference design — and a research robot built on China’s Unitree H2 body, Singapore-made hands and Nvidia’s Jetson Thor brain, shipping to labs from Stanford to ETH Zürich. Computex’s theme was “AI Together,” and the floor was robots, not chatbots. Where the durable theme is forming: physical AI moving from demo to standardised platform — the unglamorous middle where real supply chains get built.
5. The quiet number nobody cheered: 4.69%
The 10-year Treasury yield ticked to a 16-month high mid-week, and oil firmed on renewed Iran tension. Neither made headlines next to Marvell. The pressure underneath it all: higher-for-longer rates are a quiet tax on every growth multiple — the reason this week’s wobble might not bounce straight back.
Our Investing Angle
Everyone’s watching the selloff. The smarter bet is watching what the selloff is sorting.
The thesis: the market has switched from pricing AI on direction to pricing it on surprise. When a sector is up 30–60% YTD, the consensus number is already in the price; only the gap to the whisper moves the stock. That single shift splits the AI complex into two buckets — and tells you which side to stand on:
1. Proven cash, resetting expectations — demand is banked and the only question is the multiple. NVDA (held up best this week), the hyperscalers, and a cheaper CRWD after a beat-and-raise selloff. The test: did this week’s drop come with worse numbers, or just a worse mood? Mood-only = a better entry into an unchanged business.
2. Expectation and sentiment — priced on a forecast or a quote, not a print. Marvell up a third on one sentence; Broadcom priced for a raise that didn’t come. The test: strip the narrative — is there a new number underneath? If not, the price is running ahead of the business.
The losers aren’t a sector — they’re a posture: anything whose price needs the next quarter to accelerate rather than merely deliver. Which bucket does each of your holdings sit in? This week’s deep dive is the classification tool — a five-step test plus the quadrant map that sorts any name into reset, crack, or hype: How to classify your AI holdings.
⚠️ What could go wrong? (The bear case) 1. It’s not a reset — it’s the top. Every late-cycle rally re-rates on sentiment before it rolls over. If hyperscaler capex guides start slipping, “expectations caught up” becomes “demand peaked,” and the proven-cash bucket gets repriced too. 2. Rates do the breaking. A 10-year at a 16-month high compresses every growth multiple at once. A sentiment reset can turn into a valuation reset without any change in the AI story itself. 3. The whisper game cuts both ways. If next quarter’s prints clear the whisper, this week looks like a dip and the chasers are vindicated. Buying the “reset” early means buying before the prints confirm it. Size your position for the possibility that this is the start of a longer repricing, not a one-week sale. |
Three Ideas to Research This Weekend
Not recommendations — starting points for your own research. One lead idea, two quick follows.
Lead idea — The beat-and-fell discount: CrowdStrike
Why now: CRWD beat, raised FY guidance and announced a 4-for-1 split — and still fell ~9% on a priced-for-perfection reaction. The clearest case this week of the price changing while the business didn’t.
The case: Net new ARR re-accelerated (+32%) and Falcon Flex keeps pulling customers into more modules. If you liked the business at the highs, you like it more 9% cheaper with better numbers.
The risk: It still trades near 39× forward sales; a market-wide multiple reset hits the priciest names first. Google Cloud’s new AI Threat Defense (launched 27 May) is a fresh competitive watch-item.
Tripwire: Next print (26 Aug) — net new ARR growth back below ~25% means the consolidation story is stalling and the discount isn’t a gift. Holding above ~30% = the reset was sentiment, not fundamentals.
How to research: CRWD direct; or a cyber basket via the CIBR / BUG ETFs to own the theme without single-name earnings risk.
Quick follow — The plumbing behind the hype: AI connectivity
I keep coming back to this one. Marvell’s real point survives the hype: as AI compute spreads across thousands of chips, value migrates to whatever connects them — a genuine, under-owned layer. Tripwire: Marvell’s custom-silicon revenue line at the Q2 report (late Aug) — deceleration while the multiple stays high = running on the quote; re-acceleration justifies the layer. How to research: MRVL and ANET (Arista); broader exposure via SMH.
Quick follow — The unglamorous one: Japanese factory robotics
Look, I’m not entirely sold on this one. Humanoids make great Computex theatre and lousy near-term P&Ls — which is why the boring incumbents already shipping industrial arms might be the cleaner physical-AI bet. FANUC and Yaskawa are the picks-and-shovels of automation, on industrial multiples, and a yen-denominated hedge on a US-tech wobble. Tripwire: FANUC’s next order-intake guidance — book-to-bill back above 1.0 means the cycle is turning up; below 1.0 it’s still a falling knife. How to research: FANUC (6954.T) and Yaskawa (6506.T) direct; or the BOTZ / ROBO ETFs.
AI Investment Framework
6 layers · Updated weekly · Not financial advice
Layer | This week’s signal |
Infrastructure | Demand banked; only the price moved. NVDA the anchor, AVGO the re-rate candidate once it clears the next whisper. |
Platforms | The buyers, not the bought. Holds while hyperscaler capex guides keep rising — the number that validates the whole trade. |
Applications | The YTD laggard. Watch PLTR/CRM/NOW for proof AI lifts software demand rather than eroding it. |
Physical AI | Demo → standardised platform (Isaac GR00T). FANUC over the humanoid bodies — see the lead-and-follows above. |
Cybersecurity | De-rated on CRWD’s beat-and-fell. Business intact, multiple priced for perfection — the lead-idea setup. |
Global | Non-US exposure + FX hedge. BABA (policy-risk asterisk), SoftBank (levered cycle proxy), SAP (steadier). |
That cliff on the right is the whole thesis in one line: everything ran, then the market started sorting.
WHAT WE’RE WATCHING
Date | Event | Question to track |
Wed 24 Jun | Micron fiscal Q3 | Does HBM contract pricing hold into 2027 guidance? (The Issue 12 fork.) |
Late Aug | CRWD / MRVL Q2 | Does net new ARR / custom-silicon re-accelerate, or was the run priced in? |
Each Qtr | Hyperscaler capex guides | Still rising = demand real. Any cut = the bear case wakes up. |
CHANGES THIS WEEK
Cybersecurity: conviction trimmed HIGH → MEDIUM, sizing to Selective. The business didn’t crack — we’d rather add on a confirming print than catch a falling multiple.
Infrastructure: signal ↓ (cooling) on the selloff; conviction unchanged at HIGH.
Everything else holds.
Disclaimer: This newsletter is for informational and educational purposes only and does not constitute financial advice. iPrompt Signals is not a registered investment advisor. Always conduct your own research and consult a qualified financial professional before making investment decisions. |
Your Move
1. Re-label your AI holdings into two buckets: proven-cash (demand banked) and sentiment (priced on a forecast or a quote). You’re holding more of the second kind than you think.
2. Treat this week as a sector reset, not a market one — S&P flat, VIX at 15. The buying opportunity, if there is one, is in the proven-cash names that fell on mood (NVDA, a post-drop CRWD).
3. For any sentiment name, don’t add until a print gives you a number. The next earnings date is your permission slip — not the share price.
Now research one. Pick a name that fell this week and ask the only question that matters: did the business change, or just the price? Write down the number that would prove which — a tripwire — and put its earnings date in your calendar before you do anything else.
🌱 Short Take One idea: good news stopped being good enough this week — the AI names had simply run too far, too fast. One action: if you want AI exposure without betting on any single earnings reaction, a broad semiconductor ETF like SMH owns the whole chain — the proven names and the sentiment ones — and averages out the whisper game. Not a recommendation — a starting point. |
Stay curious — and stay qualified.
— R. Lauritsen
Editor, iPrompt Signals
P.S. Last week (Issue 12) we mapped the memory re-rating and flagged Micron’s 24 June print as the fork. This week’s selloff is the first stress-test — MU fell 8% on sentiment, not news. The 24 June number still settles it.
P.P.S. One-word reply, and it genuinely shapes next week’s issue: was this a RESET (healthy sorting, buy the dip) or a TOP (first crack, trim risk)?
Quick Glossary
ARR (Annual Recurring Revenue) — The yearly value of subscription contracts. “Net new ARR” is the quarter’s addition — the cleanest read on sales momentum.
Book-to-bill — Orders received divided by products shipped. Above 1.0 means demand is outpacing deliveries — a turning-up signal for cyclical hardware.
Forward sales multiple — Share price relative to next year’s expected revenue. 39× means you’re paying 39 years of current sales — only justified by fast growth.
HBM (High-Bandwidth Memory) — The stacked memory bolted to AI chips. The supply bottleneck behind last week’s memory re-rating (Issue 12).
Whisper number — The unofficial result big investors quietly expect — usually above published consensus. Beating consensus but missing the whisper still sinks a stock.
VIX — The “fear gauge” — expected S&P 500 swings over 30 days. Below ~16 signals calm; above 25 signals real anxiety.
Sponsor:
The Last Time Stocks Were This Expensive Was December 1999.
"Right now, it's good. But it was in '72, '86, 2000, and 2007." - Jamie Dimon, May 2026.
The Shiller CAPE ratio just hit 42.3. The only time in 140 years it's been higher? December 1999.
Stocks can stay expensive for a long time...
It’s one metric to consider, but when your portfolio is built around the most expensive equities in modern history, what else you diversify with could really matter.
Blue-chip contemporary and post war art has shown near-zero correlation with the S&P since 1995.* Prices are largely driven by private collectors competing for a fixed supply of artwork by artists like Banksy, Basquiat, and Picasso.
Masterworks lets you invest in shares of that market.
$1.3B deployed across 500+ artworks
29 exits to date
Net annualized returns like 16.5%, 17.6%, and 17.8%, not including those unsold
*According to Masterworks data. Investing involves risk. Past performance is not indicative of future returns. See important Reg A disclosures at masterworks.com/cd.