Technology can be useful. However, sometimes the people and companies behind it might take a few steps too far in directions that aren’t good for the economy or investors like you.
A look back almost 26 years ago might offer perspective — and a caution.
Tech was hot in the late 1990s and 2000s. Among other things, the Internet, World Wide Web, early mobile computing, and the Blackberry and iPhone sparked speculation of new forms of consumer marketing. There were many people who claimed that business was forever changed. No longer focused on profitability, the goal was “eyeballs.” It was the birth of the “attention economy.”
It was silly. Anyone who claims that business is completely different, that profits don’t matter, is either a fool or someone looking to lift your wallet. It’s probably impossible to know how many billions of dollars were flushed away. There were successes. Amazon spent beyond its revenue for years, its first nine as a public company, but it was a long game that worked, much as FedEx burned through cash for years, given the levels of necessary capital expenses.
Most of the companies didn’t succeed. In my deep and long experience covering tech and business at the time, they thrived on ego, presumption, and chaos. Service providers, like marketing and business consulting, regularly doubled their rates because the CEOs were demanding, inexperienced, and pushing everything into taking twice the amount of time that it should have needed. A lot of the companies went public, probably pushed by investors who could see what was happening and wanted fast exits, and then collapsed, destroying all shareholder value.
The business concepts were frivolous, but maybe there was something in the air, as the accounting scandals of Enron, Adelphia Cable, Worldcom, and others showed. There was fraud aplenty. I interviewed the replacement turnaround CEO at one of the companies (Adelphia, I think), who said that the accounting was so twisted that even with complete access to all records and forensic accountants, it took them six months to figure out what happened.
AI Brings Back Memories
To be clear at the start, I’m not claiming that the new developments in generative AI (other AI technologies have been around for decades) are all bad. Hardly. The use of complex statistical functions to mimic human communication is incredibly impressive, given the long history of natural language processing. I’ve spoken with multiple software companies that use large language models to parse human requests or to help translate among different systems, but always in conjunction with other technologies that can recognize patterns, do calculations, and more.
And there is no proof of the past’s business insanities. However, as the psychoanalyst Theodor Reik once wrote, according to the site Quote Investigator: “There are recurring cycles, ups and downs, but the course of events is essentially the same, with small variations. It has been said that history repeats itself. This is perhaps not quite correct; it merely rhymes.”
There are the beginnings of similar sounds. Pouring money into undertakings that have yet to prove their worth? There are some nascent businesses called “neolabs” that, as The Wall Street Journal reported, focus on long-term AI research and not immediate profits. Some are gaining billions in funding from venture capital firms, even with no products, let alone revenue.
On The Accounting Front
There are some … unusual questions about accounting. For example, not illegal but concerning is that U.S. GAAP (generally accepted accounting principles) rules give space for the big tech companies to hide tens of billions of dollars of liabilities of their AI data centers, according to major rating agency Moody’s, as Fortune reported.
The Moody’s analysis said that the top five U.S. large-scale cloud computing services — Amazon, Meta, Alphabet, Microsoft, and Oracle — so far have $662 billion in future data center lease commitments that aren’t current liabilities because the companies haven’t yet begun to receive the contracted services, and so do not appear on company balance sheets today. As the contractually obligated leases start, the billions shift from off-books to on.
By the end of 2025, these five companies had $959 billion in total future lease commitments for data centers that haven’t yet been built. The $662 billion figure is the subset for which the leases have yet to start but eventually will. That amount, according to the Moody’s analysis, is 113% of the five companies’ most recent adjusted debt. Put differently, the companies together have twice the debt than they might appear to.
Even if one of the companies cancels a lease or fails to renew it, they’re still on the hook to pay significant fees called “residual value guarantees,” or RVGs.
AI And Investors
Some investors are getting worried over heavy AI capital expense spending and the companies’ use of credit to fund it. “After Amazon, Meta, and Google-owner Alphabet all unveiled sizable increases in their full-year capex spending plans during earnings season, UBS data indicates that aggregated capex spend among AI hyperscalers could top $770 billion in 2026 — some 23% higher than previously expected,” UBS credit strategists noted, as CNBC reported.
UBS said that the increases imply a $40 billion to $50 billion in borrowing, resulting in a public market debt issuance between $230 billion to $240 billion this year.
“For years, we’ve been told this AI spend would be funded by generated cash flow — that it is equity risk, it is speculative, and not to worry about it from a credit point of view,” Al Cattermole, fixed income portfolio manager at Mirabaud Asset Management, told CNBC. “There now seems to be a change in the unspoken contract that while we would continue to lend to these businesses, really AI capex was still going to be equity or cash funded….By bringing capex spend into the debt markets, you now have the question of creditworthiness.”
TechCrunch reported that “the concept of investor ’loyalty’ [in AI] is only hanging on by a thread.” Venture capital firms are investing in different firms rather than concentrating on ones they “partner” with.
My own observation is that this type of shift is likely indicative of fear of not knowing who the winners might be. The investment space is one that is increasingly uncertain, subject to enormous amounts of debt and risk, some of which is being strategically kept off the balance sheet at the moment, and with great uncertainty as to how much market demand might exist. With the amount of money being poured into these companies and all the other AI wannabes, it’s a reasonable question to ask whether the structure will ultimately stand.
None of this is definitive or proof of a problem. It is, however, suggestive that something might be wrong.
Read the full article here