THE WEEK AHEAD ECONOMIC DATA RELEASE 30TH NOV 2025 EX OIL COMMODITIES ARE SET FOR MORE UPSIDE IN CY26 CHINA IS IRREVERSABLY DECOUPLING FROM US: THINK 2027, THINK TAIWAN IS THIS DECEMBER DIFFERENT FOR DOLLAR THE WEEK AHEAD ECONOMIC DATA RELEASE 23RD NOV 2025 DUTCH PENSION REFORMS: THE NEXT LONG END WORRY NVIDIA: WINNER TAKES IT ALL UK AUTMN BUDGET: PREVIEW BUY 10YR UK GILTS AGAINST SELL 10YR GERMAN BUNDS BUY 10YR UK GILTS SELL 10YR UST BUY S&P 500

MICHAEL BURRY IS RIGHT ABOUT CAPEX DEPRECIATION ACCOUNTING ISSUES

ADMIN || 12th November 2025

The graphics-processing units (GPUs) fueling today’s AI mania are short-lived assets with a shelf life of perhaps five years. So, while it’s comforting that the companies spending most wildly today have mountains of cash to throw around (OpenAI aside), the brief useful life of the chips and the generous accounting assumptions underpinning all of this investment are less consoling. Michael Burry, who made his name betting against US housing and who’s recently turned to the AI boom, waded in this week, warning on X that hyperscalers, industry jargon for the giant companies building gargantuan data centers, are underestimating depreciation.

Michael’s point is extremely accurate. Hyperscalers are/will be artificially suppressing depreciation expenses by overstating the lifespan of their AI hardware investments. Obviously if you understate expenses then you overstate earnings. At the very least we can say that the capex of the country’s (and indeed the world’s) biggest stocks has exceeded depreciation and amortization expenses by a greater margin (relative to both the size of the economy and the total of those firms’ own sales) than anything observed over the past three decades.

Far from being a one-off outlay, there’s a danger of AI capex becoming a huge recurring expense. That’s great for Nvidia and co., but not necessarily for hyperscalers such as Google and Microsoft Corp. Some face a depreciation tsunami that’s forcing them to be extra vigilant about controlling other costs. Welcome to job cuts.

While Wall Street is used to financing fast-depreciating assets such as aircraft and autos, it’s worrying that private credit funds are increasingly using GPUs as collateral to finance loans. This includes lending to more speculative startups known as neoclouds, who offer GPUs for rent. Microsoft alone has signed more than $60 billion of neocloud deals.

Magnetar Capital, Blackstone Inc. and Macquarie Group Ltd. are among those providing such financing. Nvidia, meanwhile, is reportedly looking at using special-purpose vehicles to raise debt to buy and rent chips to customers such as OpenAI and Elon Musk’s xAI.

We’re about to go from tens of billions of debt that’s funding quickly depreciating GPUs, to hundreds of billions of dollars.

Michael points out in his tweet:

“Massively ramping capex through purchase of Nvidia chips/servers on a 2-3 yr product cycle should not result in the extension of useful lives of compute equipment. Yet this is exactly what all the hyperscalers have done. By my estimates they will understate depreciation by $176 billion 2026-2028. By 2028, ORCL will overstate earnings 26.9%, META by 20.8%, etc.”

We agree with his argument that most of the depreciation assumptions are stretched in above names and hence earnings overstated.

Legal Disclaimer:

Trading foreign exchange/commodities/equities/bonds on margin carries a high level of risk and may not be suitable for all investors. The high degree of leverage can work against you as well as for you. Before deciding to trade foreign exchange/commodities/equities you should carefully consider your investment objectives, level of experience and risk appetite. The possibility exists that you could sustain a loss of some or all of your initial investment and therefore you should not invest money that you cannot afford to lose. You should be aware of all the risks associated with foreign exchange/commodities/equities trading and seek advice from an independent financial advisor if you have any doubts.

Read More