Even as market clouds clear, AI investment anxiety still gnaws: Mike Dolan
LONDON, – Another blockbuster earnings season has swept Wall Street, powering corporate profits and driving stock indexes back to record highs. But even as the AI investment engine fires on all cylinders, underlying anxiety is still hard to shake. In terms of headline profits, earnings projections and corporate investment, the numbers keep ringing the bell. U.S. annual profit growth has reached 50% at midyear, while overall capital expenditure this year alone is heading north of $1 trillion and rising. A report in Monday’s Wall Street Journal claimed that spending among the so-called hyperscalers building data centers and wider AI infrastructure may be $3 trillion higher than reported if off-balance-sheet items such as purchase commitments and unstarted leases are included. Wherever the eventual number lies, hand-wringing among naysayers has focussed on the growing cash burn at firms at the vanguard of the AI race, and the debt and equity financing needed from corporate bond markets, equity markets and private funds. Latest estimatesshow that bond sales by the five main hyperscalers — Amazon, Alphabet, Microsoft , Meta Platforms and Oracle — are set to double this year to $250 billion and almost double again by 2027. Even if you still fret about the fate of the big spenders, chip and equipment makers that recently wobbled after nosebleed-paced stock booms earlier this year are back on the rise. Freed by last week’s soft U.S. inflation numbers from fears that the Federal Reserve may cool their jets by raising borrowing costs next month, stock investors are behaving as if all clouds have been banished once again and the only way is up. However, worries persist among many observers and finance officials — chiefly regarding the history of overconfidence, leverage and vertigo in such cycles. There is also a fear that patchy end-use cases mean the technology may never pay off for the megacaps spending such vast sums, as well as concern over how widely those companies’ stocks are held. Researchers at the European Central Bank on Monday posted their rather ominous conclusions on how all this may end, saying the timing and extent of a stock reversal were inherently “unknowable” in advance but that a correction was coming regardless.
The ECB blog noted that the financial-stability concerns were not confined to America, as U.S. megacaps are widely held by European households, insurers and pension funds. It said a correction was an inevitable feature of such technological revolutions and investment booms, while exposure was amplified by the concentration in giant market valuations and index tracking. All involved needed to be prepared.
“Historical experience suggests that technological revolutions carry risks of a boom-bust cycle in asset prices, and this risk does not depend on today’s valuations being rational or irrational,” they wrote, metaphorically fastening their safety belts.
HUBRIS AND HUMILITY The sort of over-concentration the ECB warned about was amply illustrated last week by the head of Norway’s $2.3 trillion sovereign fund, who bemoaned the risks attached to just 10 companies representing some 20% of its overall value.
“It’s chips, chips, chips, chips, chips … we’ve never seen such concentration before,” Nicolai Tangen said. That said, he hardly fits the bill of overconfidence and hubristic finance. Even after posting a record $184 billion profit for the first half of this year, Tangen warned that the fund’s entire value could be lost if the market crashes.
“I want to contribute to our mental emergency ​preparedness,” he said, by way of explaining his worst nightmare.
Others, perhaps with less to prove to their populations and governments, have also recently pared back the big AI trade. Big U.S. institutional investors pulled back slightly from semiconductors, AI infrastructure and the megacaps during the second quarter, according to Reuters analysis of some 6,371 13F filings to the Securities and Exchange Commission from pension funds, hedge funds and wealth managers. Maybe that suggests we’re some way from overconfidence too.
But the earnings season has also thrown up more fundamental questions about AI than whether the markets have priced it correctly. Reviewing S&P 500 company statements and calls with analysts, Goldman Sachs pointed out that there was a more detailed discussion of specific AI applications and associated costs. But only a small share of management teams had quantified AI’s impact — 11% on specific use cases and 2% on earnings.
Only 7% discussed AI-related expenses, Goldman added. Even though the overall spend is modest so far, they said, spending appears to have increased sharply in recent months. If end uses and costs remain so uncertain, the ultimate scale of investment, the financing needs and the eventual payoff remain finger-in-the-wind estimates.
That might be a story for another day, but it should keep markets on their toes despite the clear blue horizon.
Enjoying this column? Check out Reuters Open Interest , your essential new source for global financial commentary. Follow ROI on LinkedIn, and X. And listen to the Morning Bid daily podcast on Apple, Spotify, or the Reuters app. Subscribe to hear Reuters journalists discuss the biggest news in markets and finance seven days a week.
This article was generated from an automated news agency feed without modifications to text.
