Breaking

AI Chip Stocks Plunge; Authers Says It Isn't a Bubble Bursting

June 24, 2026 at 00:03 EDT

Infra & Chips
Finance

A global selloff in AI chip stocks led by Nvidia and major memory makers swept markets on June 23-24, 2026, and Bloomberg Opinion's John Authers argued the decline is not the start of a bubble bursting. In a June 24 piece titled Stocks Selloff: The Bubble Probably Isn't Bursting Just Yet, he wrote that "so far, it's not even a correction," while warning that "this probably isn't when it bursts, but there's room for plenty of over-leveraged damage."

February 20, 2026 · Opinion

The AI Roundabout Isn't a Replay of the Internet Bubble

Hyperscalers are funding most AI data-center buildout from their own cash flows — and demand is already here. But over-leveraged operators on the periphery still have plenty of room to feel the pain.

~$80B

Microsoft 2025 AI-related capex (planned)

~$75B

Alphabet / Google 2025 AI-related capex (planned)

~$3T

Private credit market in the line of fire if rates or earnings shift

Hyperscaler capex, drawn to scale (2025 planned)

Each block ≈ $10B. Funded largely from internal cash flow.

~$80B

Microsoft

~$75B

Alphabet / Google

$130.7B

Amazon — operating cash flow (TTM), capex surging against it

The pressure point: debt on the periphery

Borrowing is swelling in private credit and among smaller, newer data-center operators (CoreWeave, Fluidstack). Since 2025, buildout leans on project finance and asset-backed bonds — now one of the largest segments of the US investment-grade market.

Meta — $29B raised in private capital for US AI data centers

$26B debt

$3B

Roughly 90% of the raise is debt.

Why optimists say "not a bubble"

Demand for training & inference is already materializing — not "dark fiber."
Buildout funded mostly from hyperscaler cash flows.
Shortages of power (watts) and wafers brake overinvestment.
Persistent skepticism means the market hasn't reached euphoria yet.

Why skeptics warn of pain

Michael Burry: spending looks like upgrade investment without demand.
Concerns over OpenAI's valuation; most AI startups burning cash.
Marks & Perez flag rising debt dependence and opaque project finance.
Interlocking structures concentrate risk.

The bottom line

Even if AI demand itself doesn't collapse, the key scenario is the ~$3 trillion private-credit market and leveraged peripheral investors taking the hit from a shift in interest rates or earnings — the focus going forward.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.