Breaking

George Noble Warns AI Build-Out Risks History's Largest Capital Misallocation

September 15, 2026 at 11:50 EDT

Finance
Infra & Chips

Veteran investor George Noble, the former Fidelity Overseas Fund manager who now runs Noble Capital Advisors, is warning that the artificial intelligence spending boom risks becoming the biggest misallocation of capital in history, as large technology companies pour hundreds of billions of dollars into data centers with still-unproven returns. He discussed the AI bear case on Bloomberg Surveillance on September 15 as markets weighed safety warnings from leading developers and a fresh pullback in chip shares.

September 15, 2026 · Noble Capital Advisors

AI spend without proven returns

George Noble says hyperscalers are pouring more than $600 billion into AI data centers this year — consuming about 90% of operating cash flow — while most CFOs still cannot show matching productivity.

$600B+

2026 AI data-center spend

17x

malinvestment vs. dot-com

~90%

cash flow going to capex

13%

priced chance of a 2026 bust

Malinvestment, drawn to scale

Garran's estimate: today's AI cycle is 17 times the capital misallocation of the dot-com era.

Dot-com era

17x

AI cycle now

Each block = 1 times the dot-com misallocation baseline.

How the warning is sequenced

1. CAPEX RACE

Hyperscalers spend to keep up

→

2. WEAK ROI

Returns still hard to measure

→

3. MACRO RISK

A reversal could hit GDP

→

4. ROTATION

Shift toward real assets

Still funding the build-out

Noble is not bearish on AI as a tool. Markets have kept financing capex, a year-end bust contract was recently near 13%, and Barclays has a higher year-end equity target on continued spending.

The return gap

Critics say LLM scaling has not matched revenue or productivity. A Fitch stress case maps a 35% U.S. equity drop. Noble's question remains: show the ROI.

Noble's distinction

Use the technology. Do not assume the capex pays.

He is not calling a top to the week and is not arguing AI is worthless — only that the easy money is behind, and that risk and reward have flipped until profits show up.

Continue reading

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