AI Will Be Biggest 'Misallocation' of Capital, Says Noble

Summary

George Noble argues that AI investment risks becoming the “biggest misallocation of capital in history.” He distinguishes using a useful product from owning a good investment. The central question, he says, is whether the enormous spending on AI infrastructure can earn an adequate return on capital.

AI economics

Investors should compare the capital being committed with the gross profits and revenue required to justify it. Noble questions whether AI can support enough value to avoid becoming a low-margin commodity business. He says the technology may require vastly more computing power and capital than earlier “big data” applications.

Historical parallels

Noble says the situation is not simply another dot-com boom. He compares it to a combination of the dot-com era and housing-era Ponzi finance, while emphasizing that strong growth in internet usage did not prevent companies such as Nortel and Global Crossing from failing.

Market conditions and outlook

Noble describes a “narrative-dominant” or “post-truth” market in which liquidity and prevailing stories can matter more than fundamentals, making it easier to lose money.

His stated view is rotation, not recession. He cautions that the economy and the stock market are not the same thing. Heavy fiscal spending may keep economic activity going even as market risks rise. He is concerned about rising global costs of capital and a large US deficit.

Credit and bonds

Noble says rising credit-default-swap spreads and bond yields could signal stress in AI-related financing. He expects bond markets to remain under pressure and yields to rise, arguing that attempts to support prices not justified by fundamentals invite a market attack. He says he viewed the cited Treasury-market development as a signal to short bonds, not buy them.

Equities and earnings

The discussion cites S&P earnings growth figures of roughly 20% and 30%, but the exact periods and measures are unclear in the subtitles. The interviewer says strategists estimate that at least half of the growth was related to AI spending or margins. Noble agrees that this could be unfavorable for equities if the spending falters. He also notes that stocks had held up despite rising bond yields and oil prices.

Interest rates and the Fed

Noble sees the Fed chair as being in a difficult position: a less hawkish move could be interpreted as a signal that rate increases are finished, while further hikes could disappoint markets. He says he does not rely heavily on Fed-watching or short-term rate predictions.

Assets, companies, and instruments mentioned

• AI and technology infrastructure: Large-scale capital expenditure; Noble questions the returns, margins, and market size.
• Tesla: Likely the unnamed electric-car company. Noble calls it exceptionally unattractive and refers to a very large short position, but the amount is garbled in the subtitles.
• SpaceX: Described as a separate, highly unattractive company; it is not identified as a member of the “Magnificent Seven.”
• Google: Described as comparatively acceptable.
• Magnificent Seven: Discussed as a group, though Noble says the companies differ rather than treating them all alike.
• Oracle credit-default swaps (CDS): Cited as one possible indicator of credit stress.
• Bonds and Treasury yields: Noble expects continued pressure on bond prices and higher yields; he explicitly favors shorting rather than buying bonds in the situation discussed.
• Oil / crude: Rising oil prices are noted as a headwind for equities.
• S&P 500: Discussed in relation to earnings growth.
• Nortel and Global Crossing: Cited as examples of internet-era companies that failed despite enormous growth in internet traffic.
• Chinese AI: Noble mentions the possibility of much cheaper Chinese competition; the specific “95% off” reference is unclear.

Frameworks and cautions

• Assess AI investment economics: Start with capital expenditure, estimate the gross profits needed to justify it, and work backward to the implied revenue requirement and attainable margins.
• Separate product adoption from investment returns: A product can be useful without its providers’ stocks being attractive.
• Look beyond headline growth: Consider whether earnings growth depends on AI spending or margin effects, and whether those sources can persist.
• Watch financing and market signals: Noble highlights credit-default swaps, bond yields, and the market’s willingness to continue financing AI-related spending.

Noble does not make a broad recession call. His stated concern is market rotation, financing stress, and a potential collapse in the AI investment complex. The transcript contains garbled names and figures, so some references—especially the short-position amount and earnings-growth periods—are uncertain.

Presenters and sources

George Noble was interviewed by Bloomberg host(s) whose names appear in the subtitles as Tom and Paul, likely Tom Keene and Paul Sweeney. Other sources or references mentioned include Peter Lynch, a bank credit analyst rendered in the subtitles as “Peter Bear,” Wayne Gretzky, and Alan Keyes. No financial-advice disclaimer is stated in the subtitles.