In partnership with
How can AI power your income?
Ready to transform artificial intelligence from a buzzword into your personal revenue generator
HubSpot’s groundbreaking guide "200+ AI-Powered Income Ideas" is your gateway to financial innovation in the digital age.
Inside you'll discover:
A curated collection of 200+ profitable opportunities spanning content creation, e-commerce, gaming, and emerging digital markets—each vetted for real-world potential
Step-by-step implementation guides designed for beginners, making AI accessible regardless of your technical background
Cutting-edge strategies aligned with current market trends, ensuring your ventures stay ahead of the curve
Download your guide today and unlock a future where artificial intelligence powers your success. Your next income stream is waiting.
Welcome to the iPrompt Newsletter
Friday Edition
Sponsor:
iPrompt Signals
AI & robotics investing, explained so you can actually act on it.
ISSUE 34 · FRIDAY 23 OCTOBER 2026
FRONTWAVE MEDIA LTD
After the bell on Wednesday, Meta beat on revenue and its shares rose about 4%. By Thursday's close they were down 7%. The business hadn't changed overnight. Investors had read the cash flow statement, and the terms of a new share class Meta lists as a way to raise capital.
Last week the market sorted AI by who needs a lender. This week the question widened: when the biggest spenders run short of operating cash, who covers the gap?
Three things matter this week:
1 Alphabet and Meta each had a cash-negative quarter. Both hold large reserves; the question is what funds the next gap.
2 Meta's new shares don't dilute anyone in January. They give it a tool to raise equity later.
3 The ten-year fell below 5% and neocloud shares fell anyway. One session, not a verdict.
MARKET SNAPSHOT
One-day moves on the dates shown unless stated.
Session | Move | Driver |
Mon–Wed 19–21 Oct | S&P 500 +1.3%; Nasdaq +2.2% over three sessions | SpaceX's Q3 beat |
Tue 20 Oct | CoreWeave 2031 notes yield 15.58% | Up from about 11.3% in mid-August |
Thu 22 Oct | Chip index −6.7%; Lam −11.8%; Meta −7.0%; CoreWeave −10.1% | Meta's cash flow and share class; Lam's supply warning |
Thu 22 Oct | S&P 500 −1.2%; Nasdaq −2.6%; VIX 19.03; 10-year 4.96% | The 10-year's biggest one-day fall since March 2023 |
VIX, the 'fear gauge', prices expected S&P 500 swings over 30 days; above 25 is anxiety. For the week to Thursday, the S&P 500 was flat (+0.05%) and the Nasdaq fell 0.5%.
CHART OF THE WEEK
Neither is short of reserves. Both are short of cash from operations. Built for a phone screen, so share it.
TOP HEADLINES
1 Meta's record quarter came with a funding tool
Revenue rose 26% to $64.6bn. But operating cash of $32.79bn fell short of $34.91bn of capex, including finance-lease principal: free cash flow of −$2.12bn. No buybacks this year; $82.45bn of cash and securities at 30 September.
Then the share class. After 19 January 2027 every holder gets one non-voting Class C share per share owned, which works like a two-for-one split: same ownership, twice the shares. Meta lists capital raising among Class C's uses. Dilution would come only from later sales for cash, and none is announced.
NEW TO INVESTING A split isn't a share sale. Own 1 of 100 shares and you own 1%. After a two-for-one split you own 2 of 200: still 1%. If it sells 10 new shares for cash, you own 1 of 110: 0.91%. That needn't be worse. If the cash lifts profits by more than 10%, your 0.91% earns more than your 1% did. |
2 Huang, one week apart
Last Thursday Jensen Huang said CoreWeave's cost of capital was falling. This Thursday, in Berlin, he called neocloud demand 'off the charts' and said what's missing is working capital. Both can be true. Together they put the constraint on funding, not customers. CoreWeave fell 10.1% to $55.60, its lowest close of 2026.
3 Lam posted a record quarter and fell 11.8%
Record revenue of $8.22bn, a guide of about $8.6bn, and supply-chain and clean-room limits into 2027. With the chip index down 6.7%, Lam's own news is hard to separate from the sell-off. My read: a capacity limit caps how fast revenue grows. It isn't evidence customers are spending less.
AI INFRASTRUCTURE AND ROBOTICS
4 Tesla trimmed its 2026 capex plan by about $5bn
The 10-Q now says 'in excess of $20 billion', down from more than $25bn, though the CFO expects capex to rise again after this year. Robotaxis logged nearly 2m unsupervised miles; revenue is due from 2027. Tesla also signed $30bn of credit lines last month. Self-funded, mostly.
5 Nvidia is paying its suppliers up front
At GTC Berlin, Nvidia signed capacity deals with six suppliers. Two come with cash up front: $1.1bn to KYEC for US chip testing, and a multibillion-dollar advance to Japan's Ibiden to triple its Ono substrate plant. Samsung Electro-Mechanics joins with co-invested Korean lines, terms undisclosed. The richest buyer in the chain is funding its own supply, which moves build risk off suppliers at a price nobody has published.
TRIPWIRE TRACKER
'Happened' records the event. 'Thesis' records whether it moved the argument, and how far one data point carries.
Tripwire (Issue 33) | Happened? | Thesis |
Tesla Q3 shows robotaxi miles or revenue separately | Partly. Cumulative unsupervised miles, yes. Revenue, no | Unchanged. Fleet economics still not measurable from outside |
Bear case 1: ten-year back under 5% | Yes. 4.96% on 22 Oct | Not yet tested. CoreWeave fell 10.1% that day, leaning against the bear case. Recheck after the Fed |
CoreWeave issues unsecured debt at materially lower yields | No issuance | Unresolved. Related evidence only: 2031 notes at 15.58% on 20 Oct |
Anthropic prospectus public; SK Hynix and Fed, 28 Oct | Not yet. Anthropic reported, unconfirmed, for as early as today | Open |
Tesla · Q3 update · 21 Oct | Benzinga · CoreWeave 2031 notes · 20 Oct | Seeking Alpha · Anthropic S-1 timing (reported) · 22 Oct
OUR INVESTING ANGLE
Everyone's watching Meta's capex. I'd watch its share count, adjusted for January's split.
The thesis: over the next two or three quarters, I expect the largest AI spenders to hold their capex plans and fund any gap from reserves first, then debt, and at Meta eventually new shares. Nothing forces a share sale soon, but every route costs owners something: a thinner cushion, interest and refinancing risk, or a smaller slice. Split-adjusted free cash flow per share is the better scorecard.
Better placed if I'm right: suppliers paid out of that spending, such as ASML, TSMC and Lam, and those paid in advance, such as Ibiden. More pressured: CoreWeave, borrowing at junk rates, and IREN and Nebius, because the giants compete for the same lenders. Then Meta's per-share growth, if it sells Class C for cash. Alphabet less so, with $242.5bn of cash and securities.
Tesla's cut is a fair point against me. I'd weight it below a cut from a big buyer of chip tools, so here's the test, set in advance: the supplier half is challenged if Microsoft, Alphabet, Amazon or Meta lowers its 2026 or 2027 capex plan.
Where I could be wrong: if Meta borrows instead, owners keep their percentage but take on the interest, the refinancing risk and a heavier balance sheet. A different cost, not no cost.
The companion, The Four Wallets, separates four things that get blurred (a cash shortfall, a financing cost, dilution and value destroyed) and gives you a test to rerun on any holding. [COMPANION LINK]
WHAT COULD GO WRONG WITH OUR ANGLE 1 Reserves and bonds cover it. Alphabet raised more than $30bn of bonds in February on top of its reserves. If gaps stay small, a share sale may never come. 2 Revenue catches up. Alphabet's operating cash rose 41% year on year in Q2. If that outruns capex, no outside wallet is needed. 3 Suppliers aren't insulated. One big buyer trimming 2027 plans would shorten Lam's and ASML's queues, and their shares would likely fall first. Size positions for the possibility that the thesis takes longer to prove, or proves wrong. |
THREE IDEAS TO RESEARCH THIS WEEKEND
Not recommendations. Starting points for your own research.
1 Alphabet: the self-funder test
Why now: Q3 results on 28 October, hours after the Fed. The case: $242.5bn of cash and securities and $53.3bn of trailing free cash flow, even after a cash-negative Q2. The risk: no buybacks; capex rises again in 2027. The price: about $4.2tn, roughly 78 times trailing free cash flow, which assumes today's spending becomes much larger cash flow later. Tripwire: a second negative quarter plus a higher 2027 range means reserves and debt carry more of the build; positive free cash flow with buybacks restarted means operations still do. How to research: operating cash against capex, then the share count.
2 Ibiden: the customer pays up front
What beats a sold-out order book? One the customer pays for in advance. I keep coming back to this one. Why now: Nvidia's multibillion-dollar advance to triple Ibiden's Ono plant from early 2028. The case: building with the customer's cash keeps its own balance sheet off the cycle. The risk: concentration, and unseen terms such as reserved capacity or pricing. The price: about 64 times forward earnings in early October, which assumes the new capacity fills; Samsung Electro-Mechanics was near 35 times 2027 estimates. Tripwire: interim results, expected around the end of October. The advance on the balance sheet with the Ono plan intact makes the case measurable; a slipped timetable weakens it. How to research: 4062.T against 009150.KS.
Nvidia · Supply-chain release · 21 Oct | StockAnalysis · Ibiden valuation · 9 Oct | MarketScreener · SEMCO estimates · 8 Oct
3 Meta: bargain, or down payment?
Look, I'm not sold on this dip. Why now: a 7% fall on cash flow, not ad demand. The case: revenue up 26%; $7.40 a share before a $9.9bn legal charge. The price: about 24 times trailing adjusted earnings. The risk: Oppenheimer projects about $135bn of free-cash-flow losses through 2028, against $82.45bn of cash and securities. Tripwire: a first sale of Class C for cash before the Q4 report means owners start funding the build; a debt-funded plan with buybacks restarted means they don't. How to research: the 10-Q commitments note, and the split-adjusted share count.
Meta · Q3 results · 21 Oct | MarketBeat · Meta trailing EPS | investingLive · 10-Q commitments, Oppenheimer · 22 Oct
FRAMEWORK: CHANGES THIS WEEK
Cloud moves from 'credit led' to 'funding led': gaps can come from reserves, debt or equity. Meta moves to 'dilution watch'; no sale is announced. Added: Lam to Computing, Ibiden (4062.T) to Global. Not added: Anthropic, until its S-1 is public. Full seven-layer framework: [FRAMEWORK LINK]
WHAT WE ARE WATCHING
When | Checkpoint | Question |
23 Oct (reported, unconfirmed) | Anthropic public S-1 | How much of the compute bill can't be cancelled, and who is it owed to? |
28 Oct | Fed (14:00 ET); SK Hynix call (09:00 KST) | A hold, priced near 90% on Kalshi? Any cut to 2027 memory investment? |
28 Oct · after close | Alphabet Q3 | A second cash-negative quarter? Buybacks back? |
29 Oct; end Oct | Amazon, Intel; Ibiden interim (date unconfirmed) | Capex guides; how the Nvidia advance is booked |
YOUR MOVE
Pick one AI holding. Pull its diluted share count for eight quarters, adjusted for splits and stock dividends (Meta's doubles in January without diluting anyone), and its free cash flow for four. A count still rising after adjustment, with buybacks stopped, means owners are funding part of the build. Whether that's bad depends on what the money earns: the companion's test.
SHORT TAKE A broad chip fund such as SOXX or SMH owns this week's suppliers without picking which giant funds its gap how. Not safer: SOXX fell 6.8% on Thursday. Not a recommendation, a starting point. |
Stay curious, and stay qualified.
R. Lauritsen
Know someone who owns Meta or Alphabet? Forward this before Wednesday's results. They'll know which line to read first.
P.S. If Anthropic's S-1 is out by Monday, I'll score its commitments note in next week's Tripwire Tracker.
P.P.S. Reply with one word, spender or supplier, for the Q4 reports. I'll publish the split next Friday.
QUICK GLOSSARY
Dilution Your percentage shrinking because the company sold new shares. A split or stock dividend to every holder is not dilution.
Free cash flow Cash from operations minus capital spending. Negative means the gap came from reserves, borrowing or new shares.
SOURCE AND METHOD NOTE
Market data to the 22 October close. Capex and free cash flow follow each company's definitions; Meta's include finance-lease principal. Alphabet's multiple and the three-day index gains are our calculations. Reported and tentative dates are unconfirmed.
Disclaimer: For information and education only, not financial advice. iPrompt Signals is not a registered investment adviser. Conduct your own research and consult a qualified financial professional before making investment decisions.
iPrompt Signals · Issue 34 · 23 October 2026 ·
Sponsor:
A fast AI answer isn't always a coordinated one. When AI resolves B2B issues without looping in the teams responsible for them, confidence outpaces alignment. A new Harvard Business Review Analytic Services briefing paper, sponsored by Front, looks at closing that gap.