
Michael Burry is again sounding the alarm over the record investment by Big Tech and cloud companies, arguing that hyperscalers are building up enormous financial commitments that could eventually result in massive write-offs.
In a new blog post published on Thursday, “The Big Short” investor compared the current AI spending boom with previous capital-cycle bubbles, including the dot-com era, pointing to net capital investment reaching 2.07% of GDP — the highest level in four decades outside the aftermath of the March 2000 Nasdaq peak.
“Today’s level is higher than any prior time save for the aftermath of the massive NASDAQ market peak in March 2000,” Burry wrote.
His concern is not the amount companies are spending, but how much of the spending is being financed through leases, purchase obligations and other commitments that do not immediately appear as conventional capital expenditure or debt.
Burry wrote that he expects the consequences to become particularly visible around 2028 or 2029, when he believes companies could face significant write-offs.
“The most profitable companies in the world, except Apple, are betting everything on this as their free cash flow turns negative and the borrowing pace increases,” he said. Here’s his view on the issue for each of the major hyperscalers.
Burry took a particular aim at Microsoft’s accounting treatment of its massive data-center buildout. He said the company nearly tripled uncommenced leases to more than $300 billion while extending the upper end of its data-center useful-life range from 15 years to 25 years.
The changes, he argued, allow Microsoft to shift spending toward operating leases and reduce reported capital expenditure in 2026 and 2027. “Too cute by half,” Burry wrote, adding that Microsoft appears increasingly willing to use depreciation assumptions to achieve desired accounting outcomes.
Amazon’s uncommenced leases and purchase commitments surged 81% in nine months to about $267 billion, according to Burry. Meanwhile, long-term debt doubled in six months to $128.9 billion.
While he credits Amazon for putting much of its borrowing directly on the balance sheet, he argued that the company’s financial position is becoming increasingly leveraged as free cash flow turns negative. “To Amazon’s credit, at least the borrowing sits on the balance sheet,” Burry wrote. “But sheesh, honest leverage is still leverage.”
Burry estimates Meta has roughly $700 billion of uncommenced leases and purchase obligations that remain off balance sheet, potentially approaching $1 trillion.
He challenged the argument that Meta’s investments are protected by long-lived physical assets, pointing to telecom infrastructure companies that suffered enormous write-offs during the early-2000s downturn.
“Whatever Meta’s physical assets are, they are far less long-lived than the telecommunications and data transmission grid,” he wrote.
Alphabet has nearly $900 billion in off-balance-sheet commitments and exposures, according to Burry, including purchase obligations and financing linked to AI partners.
Burry said he was particularly skeptical of arrangements involving Google investing in AI companies while also supplying them with computing infrastructure. He called such transactions “circular dealing” and argued that similar practices would attract greater scrutiny in other industries.
Burry questioned Oracle’s accounting around customer prepayments, highlighting an $11.4 billion operating-cash-flow item tied to customers with significant financing components.
By accepting customer prepayments, Oracle uses accounting rule ASC 606 to "magically" boost future cloud revenue by nearly 20% without delivering extra service or work product immediately. “In this way, the extra $1.9 billion becomes extra revenue created by the structure of the contract,” Burry wrote.
Burry argued that massive spending itself does not guarantee a durable competitive advantage. “Capital alone is no lasting competitive advantage,” he said, urging investors to wait for more attractive opportunities rather than chase the AI boom.
Big Tech and cloud hyperscalers are on track to spend roughly $800 billion on capital expenditure in 2026, with estimates pointing to $1 trillion or more in 2027 as companies race to build AI data centers and computing capacity.
The sheer scale has made investors nervous because spending is increasingly outpacing free cash flow, forcing companies to rely more on debt and other sources while raising questions about whether AI-related revenues will ultimately justify the investment.
Burry’s blog comes amid broader concerns that a slowdown in AI demand or weaker-than-expected returns could leave hyperscalers with excess capacity and declining returns on investment, potentially weighing on the broader market.
On Stocktwits, the retail view was ‘bullish’ for GOOGL, AMZN, MSFT, META and ‘neutral’ for ORCL as of early Friday.
For updates and corrections, email newsroom[at]stocktwits[dot]com.<
Stay updated with all the latest Business NewsShare Market NewsIPOsGold PriceDA Hike8th Pay CommissionAsianet News Official App