Michael Burry Raises Accounting Concerns Over Big Tech AI Infrastructure Depreciation Practices

This analysis is based on multiple reports from Business Insider and Yahoo Finance published in November 2025, which documented Michael Burry’s allegations regarding Big Tech accounting practices [1][2][3].

Michael Burry, renowned for his prescient 2008 housing market prediction, has launched a significant challenge to the earnings quality of major technology companies. In a November 11, 2025 social media post, Burry alleged that leading tech firms are artificially inflating earnings through aggressive depreciation schedule extensions for AI infrastructure, specifically servers and GPUs [1][3].

The allegations target the “hyperscalers” - Meta, Alphabet, Amazon, Microsoft, and Oracle - which collectively represent approximately $11.12 trillion in market capitalization [0][1]. These companies are projected to spend $349 billion on capital expenditures in 2025, with substantial portions directed toward AI infrastructure development [1].

Burry’s analysis focuses on a critical accounting discrepancy: while AI technology typically advances rapidly with 2-3 year product cycles, companies have been extending depreciation schedules from 2-3 years to 5-6 years for computing equipment [1][3]. This practice reduces non-cash depreciation charges, thereby boosting reported net income.

The financial implications are substantial, with Burry estimating a potential $176 billion earnings understatement across 2026-2028 [1][2]. Specific company impacts include:

• Oracle: 26.9% potential earnings overstatement by 2028 [1][2]
• Meta: 20.8% potential earnings overstatement by 2028 [1][2]

The technology sector has shown mixed performance following these revelations. While the broader Technology sector is up 0.49%, Communication Services has declined 1.88% [0]. Individual Big Tech stocks are experiencing pressure:

• Meta (META): $603.50 (-1.05%) - down 14.83% over the past month [0]
• Alphabet (GOOGL): $276.05 (-0.90%) [0]
• Amazon (AMZN): $235.72 (-0.78%) [0]
• Microsoft (MSFT): $504.86 (+0.31%) - showing relative resilience [0]

Evidence supporting Burry’s concerns includes Meta’s recent policy change extending estimated useful lives for “certain servers and network assets” from 4-5 years to 5.5 years effective January 2025, which the company estimates will reduce 2025 depreciation expense by $2.9 billion [3].

Notably, Amazon has moved in the opposite direction, shortening estimated useful lives for some servers and networking equipment from six years to five years in Q1 2025, citing accelerated AI technology development [3]. This divergence suggests varying levels of confidence in AI investment returns across companies.

Burry’s concerns emerge amid growing skepticism about AI investment returns. Jim Chanos, famed Enron short-seller, has similarly warned about massive AI spending outpacing revenue growth [1]. Wall Street analysts have been monitoring depreciation issues since at least November 2024, with some describing it as a “potential time bomb at the heart of the AI trade” [1].

The situation highlights significant information asymmetry in the market. While companies follow Generally Accepted Accounting Principles (GAAP), the subjective nature of depreciation assumptions allows for substantial earnings management opportunities. The fact that Burry has promised “more detail coming November 25th” suggests additional revelations may emerge [1].

The concentration of these accounting practices among market-leading companies creates potential systemic risk. With Mag7 earnings still projected to grow ~27% YoY despite massive AI capex and rising depreciation costs, the debate over whether profits are inflated by accounting or reflect genuine growth has significant implications for market valuations [1][2].

1. November 25th deadline
   : Burry’s promised detailed analysis could trigger significant market reaction [1]
2. Earnings season responses
   : How companies address these concerns in upcoming quarterly reports
3. SEC scrutiny
   : Potential regulatory investigation into depreciation practices
4. Competitive responses
   : Whether other companies follow Amazon’s lead in shortening depreciation periods [3]
5. AI ROI metrics
   : Evidence of actual returns on massive AI infrastructure investments

The divergence in company responses suggests varying levels of confidence in AI investment returns [3]. This could lead to earnings volatility as companies adjust depreciation assumptions, increased analyst scrutiny of AI-related capital efficiency, and potential valuation compression for companies with aggressive depreciation policies.

The controversy centers on legitimate accounting practices that may be masking the true economic costs of AI infrastructure investments. While companies operate within GAAP guidelines, the subjective nature of depreciation assumptions for rapidly evolving technology creates significant earnings management opportunities. The situation warrants careful monitoring of company disclosures, regulatory responses, and actual AI investment returns. The market’s ability to accurately price these risks will depend on transparency in depreciation methodologies and demonstrable returns on AI capital expenditures.