AI is running on borrowed money

Investment Analysis Clubs Macro Economic Trends and Risks
1

https://www.nytimes.com/2026/07/17/business/ai-spending-oracle-stocks-bonds.html

A.I. Is Running on Borrowed Money

Risk is rising as big tech companies like Oracle — the ultimate financial source of the Ellison media empire — need to turn to the bond market for staggering sums to finance data centers.

By Jeff Sommer, The New York Times, July 17, 2026

…
Data centers and the other infrastructure for A.I. involve staggering sums of money. These cascades of A.I.-driven cash have enriched diverse segments of the stock market, from semiconductor makers to engineering companies to utilities to energy producers. A.I. money is bolstering the entire U.S. economy, contributing perhaps 1.1 percent to the nation’s economic growth, JPMorgan Asset Management estimates.

But where’s that money coming from?..

A.I. data centers are increasingly running on borrowed money…

The gigantic A.I. infrastructure expenditures are outpacing growth in profits. According to Bank of America, total capital expenditures for Oracle, Alphabet, Microsoft, Amazon and Meta are exceeding their free cash flow. …

So the tech companies are going to the capital markets, mainly the bond market, which has begun to charge premiums for what it considers to be heightened risk. …

One problem is that the expected revenue for the data centers isn’t rock solid. Much of it is linked to A.I. start-ups like OpenAI and Anthropic, which themselves rely on borrowed funds and speculative investments by venture capitalists and private equity funds. …

If their returns from A.I. investments don’t pan out, or if their borrowing costs become onerous because of rising rates on debt, these companies may not be in an enviable position. There will be questions about whether their share pricing is “appropriate,” she said, given their “leverage and capital intensity.”… [end quote]

This is jargon meaning, “These stocks are too expensive because the companies are borrowing and spending too much and the profits won’t be there.”

Since the S&P500 Index is 40% AI companies, the whole index fund will be pulled down along with the AI companies. Many owners of the index through ETFs like SPY will sell as the NAV drops. That will force selling of all stocks in the funds, throwing out the babies with the bathwater.

Wendy

7 Likes
2

It has also been pointed out that they are issuing 10 yr bonds but service life of the chips is only five years. Of course the data center facility remains an asset even when the chips are obsolete and due for replacement.

Bond rating agencies should consider all aspects. Some of these bonds should be rated as junk.

3 Likes
3

Gemini says that the service life of the most expensive chips is 3 years because they run hot which physically damages them. But the companies depreciate them over 5 years.
Wendy

7 Likes
4

This is why you’re going to continue to see major advancements in cooling technologies, or increased thermal efficiencies in chips. Heat is basically the main factor for semiconductor degradation. If you can keep them cool you can double (if not more) their lifespan. With the amount of money being invested in GPUs now for AI specific datacenters, spending a ton of money on a top tier cooling system is worth it.

2 Likes
5

From NYTimes article:
Few people outside the markets have paid attention to what goes on behind the financial curtain for artificial intelligence. These big companies are able to categorize the money as an investment — a capital expenditure — and not as an expense. So under current accounting rules, the bulk of the spending has not yet counted against their gaudy earnings. That is helping to propel the stock market to new heights under rosy assumptions that A.I. will transform the world, and that the companies behind it will be making money.

Hm how many AI “investors” have skipped doing their due diligence before buying the stocks?

1 Like
6

@tjscott0 in a typical bubble, speculators follow the momentum. “Due diligence” is an old-fashioned concept which speculators don’t apply to “new era growth” assets (be they stocks, houses, internet cables, railroads, radio, tulips, etc.).

Wendy

2 Likes
7

The writer doesn’t quite understand. Capex is an expense; it is just expensed over several years. In addition, the OBBB excellerated the depreciation schedule so that it can be expensed in the first year. Thus the argument in not valid for 2026 capex (and in the future).

DB2

6 Likes
8

It’s not just speculators! When the Nvidia weighting in a particular index (let’s say the S&P500) becomes 7.316%, that means that the index funds based on that index must own approximately that much Nvidia. Even if they believe that some portion of the accounts receivable (or future orders, etc) will never see actual cash coming in to pay for them.

And if there is lots and lots of money in those index funds, then that means that lots and lots of money must be invested in it. Automatically.

2 Likes
9

“This time is different”. Where have we heard that before?

Regarding index funds, that is the pitfall to them - you own everything, good and bad. This particular problem is why $XMAG even exists, giving investors the ability to have the S&P 500 but have the Magnificient Seven names excluded.

10

XMAG has about $177 million in assets. The Vanguard 500 Index Fund has $1.675 TRILLION in assets. That’s almost 10,000 times bigger! And it’s only one of a few large S&P500 index funds.

My point wasn’t to say anything is different or not different this time, but rather than index funds BY THEIR VERY CHARTER don’t do any due diligence on their holdings.

1 Like
11

S&P Global Ratings cut Oracle’s credit rating to BBB- on July 9, one notch above junk status, citing extreme concentration risk from OpenAI, which represents roughly half of Oracle’s $638 billion backlog, alongside a projected free cash flow deficit of $42 billion in fiscal 2027.

2 Likes