Google’s $85B Raise Opens the AI Infrastructure Capex Equity Dilution Cycle
The host tallies $360B of pending equity raises from Google, OpenAI, and SpaceX. Super Micro's 28% drop on a $7B raise is the preview of what dilution does while the Fed holds rates.
On June 2, 2026, Alphabet (GOOGL) priced an $84.75 billion equity offering, the largest single equity raise on record, Bloomberg reported. Eight days later Super Micro Computer (SMCI) announced a $7 billion raise and lost 28% in a session, per TIKR.
The host on The Weekly Wrap argued these are not two stories but one. The AI infrastructure capex equity dilution cycle has begun. The bill is being passed to public shareholders, and the Fed is not going to make it easier.
Key facts at a glance
- Alphabet priced an $84.75B equity offering on June 2, 2026 (record single raise), Bloomberg reported.
- Oracle (ORCL) booked Q4 FY2026 remaining performance obligations of $638B, up 363% year over year, per the Oracle Q4 release.
- Oracle FY2026 capex hit $55.7B versus $50B guided, and FY2027 capex is guided to $70B, per TradingKey.
- Super Micro disclosed a $7B raise in its free writing prospectus; the stock fell 27.98% on June 10, per TIKR.
- May CPI printed +4.2% year over year, the highest since April 2023, per the BLS release.
- The 10-year Treasury yield crossed 4.5% in early June; daily values are at Federal Reserve H.15 and FRED DGS10.
Google’s $85 billion sets the template
Alphabet’s offering is roughly the size of Petrobras’ 2010 record and dwarfs every prior tech precedent, Bloomberg reported. Proceeds fund 2025 capex now guided to $85B and a 2026 plan ranged at $175-190B, per CNBC.
The host’s read on the episode is that this is the opening of a sequence. “Add the 75 billion to be raised by SpaceX, and we’re talking about 275 billion of capital to be raised. Then add the 85 billion that Google raised and we are at 360 billion. That’s a lot of capital for equity markets to absorb,” he said.
The $360B figure is a synthesis of confirmed and inferred raises (OpenAI’s S-1 size is not public), not a market consensus. Treat it as a directional ceiling, not a forecast.
Oracle: a $638 billion backlog that still needs cash
Oracle surfaced a remarkable contradiction in its press release: a $638B backlog that now exceeds the comparable backlogs of larger hyperscalers, per Yahoo Finance.
Even so, FY2026 free cash flow was negative $23.7B. Management telegraphed a fresh roughly $40B debt and equity raise to fund AI infrastructure, with FY2027 capex guided to $70B plus $20-25B of component prepayments, per TradingKey and the SEC 8-K exhibit.
A historic demand backlog is not the same as cash-funded growth. Oracle is financing the next two years of build-out from outside the income statement.
Super Micro: $7 billion on a $20 billion cap
Super Micro disclosed $1.25B of common stock, $3.75B of convertible preferred depositary shares, and a $2B ATM program in its free writing prospectus; StockTitan carried the company release.
The two-day decline ran 26-27%, per MLQ, with June 10 alone closing at -27.98% per TechTimes. A pre-announcement market cap near $20B, charted by Macrotrends, implies roughly 35% dilution.
Markets priced the raise as the dilution it is, against a $39B AI server backlog that does not throw off cash today.
The cross-cutting theme: capex with no rate-cut escape
Hyperscaler 2025 capex landed near $400B per Visual Capitalist. 2027 is projected above $1 trillion by CNBC. The intermediate year, 2026, is tracking $600-725B, per DataCenterRichness. The host’s “close to $1 trillion this year” framing pulls the 2027 figure forward.
May headline CPI at +4.2% and core at +2.9%, summarized by CNBC and Kiplinger, keeps the Fed on hold. Polymarket prices about 57% odds on zero cuts in 2026, meaning markets still see ~43% odds on at least one cut. “Near zero” overstates the picture; the direction is right.
The 10-year Treasury closed at 4.55% on June 5, per Advisor Perspectives and earlier reporting from CNBC. Equities reacted the same session, with the S&P 500 down 2.64% and the Nasdaq down 4.18%, per CNBC and Fortune.
Index hugging meets private mega-caps
“active managers feel obligated to restructure their portfolios to reflect the eventual need to include SpaceX and soon entropic and open AI regardless of the underlying merits of the investment decisions,” the host said. If active managers front-run index inclusion of mega-cap privates, fundamental selection thins out at the same moment dilution accelerates.
Meta (META) is reportedly weighing a multi-tens-of-billions equity offering, per CryptoBriefing and Value The Markets. Microsoft (MSFT) is exploring similar moves in the same reporting, with nothing filed yet. Apple (AAPL) opted out of the LLM race years ago, a decision that looks better with every fresh capex slide.
Where the AI capex equity dilution cycle stands today
| Company | Raise structure | Period | Reaction or status |
|---|---|---|---|
| Alphabet (GOOGL) | $84.75B common equity (priced) | Jun 2 2026 | record offering; 2026 capex guided $175-190B |
| Oracle (ORCL) | ~$40B debt and equity (planned) | telegraphed FY27 | RPO $638B; FY26 FCF -$23.7B |
| Super Micro (SMCI) | $7.0B ($1.25B common + $3.75B pref + $2B ATM) | Jun 9-10 2026 | -27.98% on Jun 10; ~$12B market cap erased over two days |
| Meta (META) | reportedly multi-tens-of-billions | not filed | management discussion stage |
| Microsoft (MSFT) | exploring | not filed | per press reports |
Watchlist note
Picks-and-shovels exposure (semiconductors, electrical equipment, power) absorbs AI demand without carrying the capex liability. None of that is a buy instruction. It is a framing the host repeated on the episode: the model layer commoditizes, the physical layer monetizes. Apple’s deliberate non-participation is the cleanest example of the inverse trade.
What this does not tell you
The “$360B of raises” total is the host’s synthesis. OpenAI’s S-1 size is unconfirmed, SpaceX’s $75B is an estimate, and the “$1 trillion in 2026” capex framing pulls the 2027 number forward.
The “70% drop in PE deal value” cited on the show is the tech subset Q4 2025 to Q1 2026, not all private equity. Trailing-12-month PE deal value was $2.1T as of Q1, per KPMG and Bain, with a record unsold portfolio backlog rather than a strict liquidity crisis. The “4.5% Rubicon” is the host’s framing; analyst notes use a softer “warning zone.”
FAQ
How much equity capital is the AI capex cycle expected to absorb?
The host’s $360B estimate sums Google’s priced $84.75B, an inferred ~$75B for SpaceX, and ~$100B for OpenAI from an S-1 that is confidential. It does not include Oracle’s signalled ~$40B or any Meta or Microsoft action that has not been filed.
Is Super Micro’s 28% drop a one-off?
Disclosed structure plus a ~35% dilution math against a $20B cap argues the move priced economic reality, not panic. The wider read is that markets are willing to penalize large raises against thin equity bases even when the order book is real.
What changes if the Fed cuts?
The “near zero” framing overstates current pricing; markets still price about 43% odds on at least one cut this year on Polymarket. A cut compresses discount rates and softens the dilution hit. A hold sustains it. CPI prints and the labor data carry the most weight from here.
Disclaimer. This article is editorial commentary on public market events as of June 13, 2026. It is not investment advice, a solicitation, or a recommendation to buy or sell any security. Numeric figures carry the date and source shown; intraday quotes and balance figures can move materially before this reaches a reader.
Past capital raises, regulatory filings, and host commentary referenced here describe events that have already occurred. They do not predict that a similar dilution path will repeat for any specific issuer, nor that listed tickers will follow the same pattern observed elsewhere in 2026.
