AI spending lifts borrowing costs worldwide as SoftBank’s next fund tests the doubters
A Bloomberg column says Masayoshi Son’s second hundred-billion-dollar vehicle strengthens Ray Dalio’s warning, and the BIS has put a revenue test to the companies building the data centres.
Bloomberg Opinion argued that SoftBank founder Masayoshi Son’s attempt to set up another $100 billion AI-focused venture capital fund will only give the doubters more to work with, and that it underscores Ray Dalio’s AI warning1. Son is asking Gulf investors for the money on top of his OpenAI bet, and last month the company tapped the bond market for $11.1 billion, the largest high-yield corporate bond sale ever recorded6.
The cost of that financing is now a macro variable. One account reported that AI spending is pushing up borrowing costs around the world, with Wall Street estimates putting this year’s AI-related investment-grade issuance at a figure the excerpt does not complete2. The same outlet wrote that AI borrowers now dominate half of America’s investment-grade market, in a sentence the excerpt cuts off3.
Get AiM Weekly, free.
Why it matters to investors
The official sector has started putting a number on what the buildout must earn. Bank for International Settlements president Pablo Hernández de Cos said in a recent speech that AI investments are being made on a basis the excerpt does not finish, in coverage framed as AI data centres facing a $6 trillion revenue challenge4. Against that, Microsoft, Meta and Google AI capital spending has reached $725 billion, and the hyperscalers are effectively betting that AI infrastructure is too big to fail, a framing the investor Michael Burry has used5.
The accounting is the part that reaches ordinary portfolios. Aswath Damodaran argues that AI capital spending is distorting earnings and cash flows, naming Meta, Alphabet, Amazon and Microsoft as the builders of the architecture whose spending will have an effect the excerpt does not state7. A separate commentary argued that the physics of corporate finance cannot sustain the current burn rate8. Both are arguments about reported profit rather than about chips.
What to watch
The equity signal came first. Memory chip stocks dragged the Nasdaq-100 into a correction this week3, which is the part of the complex closest to the capital spending itself.
Then the credit signal, because it is the one that prices everything else. If AI borrowers are half of the investment-grade market3 and the issuance is still rising2, the marginal buyer of a non-AI corporate bond is being asked to compete with a data centre. Third is whether the Gulf money arrives at all: the fund is an attempt, not a close1,6.
Sources
- SoftBank’s Next $100 Billion Underscores Ray Dalio’s AI Warning - Bloomberg.com, bloomberg.com (2026-10-11)
- AI Spending Is Now Pushing Up Borrowing Costs Around the World - Startup Fortune, startupfortune.com (2026-10-11)
- Memory chip stocks dragged the Nasdaq-100 into a correction this week - Startup Fortune, startupfortune.com (2026-10-11)
- AI Data Centers Face $6 Trillion Revenue Challenge, Big Tech Warned - Aju Press AMP, m.ajupress.com (2026-10-11)
- Microsoft, Meta, Google AI Capex Hits $725B [2026] - Tech Insider, tech-insider.org (2026-10-11)
- Masayoshi Son is asking the Gulf for $100 billion on top of his OpenAI bet - Startup Fortune, startupfortune.com (2026-10-11)
- Damodaran: AI Capex Is Distorting Earnings and Cash Flows - FourWeekMBA, fourweekmba.com (2026-10-10)
- When the Silicon Engine Stalls: The AI Bubble Reckoning and the Rise of Sovereign Compute, microwire.info (2026-10-11)
Assembled by Edwin, my AI assistant powered by Claude, from the public excerpts of the outlets numbered above. No human wrote or checked it before publication, so read the sources before you act on it.