Michael Burry Targets AI Boom: Cassandra Unchained & The New Big Short
- THE MAG POST

- 1 day ago
- 5 min read

In a move that has sent shockwaves through Wall Street and Silicon Valley alike, Michael Burry, the legendary investor immortalized in The Big Short, has officially altered his course. Known for his prescient bet against the subprime mortgage market in 2008, Burry has deregistered his hedge fund, Scion Asset Management, as a Registered Investment Adviser (RIA). But this is not a retirement. Instead, it is a pivot to a new, unconstrained battlefield: a paid Substack newsletter titled "Cassandra Unchained."
Burry’s transition from managing external capital to publishing unfiltered financial commentary marks a significant shift in how one of the market's most bearish voices intends to influence the narrative. His target? The unprecedented, and in his view, unsustainable, Artificial Intelligence (AI) boom.
With Scion now operating essentially as a family office, Burry claims he is finally free from the "regulatory muzzles" that limited his speech. His first order of business has been to dismantle the bullish thesis surrounding companies like Nvidia, Microsoft, and the constellation of AI startups that have driven the S&P 500 to record highs.
The New "Big Short": Anatomy of an AI Bubble
Michael Burry’s thesis on the AI sector is not merely skepticism; it is a structural indictment of the entire industry's current valuation. Writing in "Cassandra Unchained," Burry has drawn direct parallels between the current AI hype cycle and the Dot-com bubble of 2000.
His argument centers on what he calls the "Five Public Horsemen" of the AI apocalypse: Microsoft, Google, Meta, Amazon, and Oracle. According to Burry, these hyperscalers are engaging in a capital expenditure (Capex) war that has disconnected from reality. He specifically points to Nvidia as the "Cisco" of this era—a company providing the "plumbing" for a revolution that he believes will fail to monetize effectively in the near term.
The "Live Love Laugh" Warning
True to his eccentric style, Burry utilized a "Lord of the Rings" meme to illustrate investor complacency. The meme depicts Gandalf asking Frodo what he sees in the One Ring, but instead of the Black Speech of Mordor, it reads "Live Love Laugh." Burry’s implication is clear: investors are gazing into a dangerous, fiery asset class and seeing only benign positivity.
This cultural commentary is backed by significant financial positioning. Before deregistering Scion, 13F filings revealed that Burry held massive put options (bets against) on Nvidia and Palantir, totaling over $1 billion in notional value. He argues that the market is ignoring "supply-side gluttony," where infrastructure spending is outpacing actual end-user demand for AI services.
The Accounting "Gimmicks": A Mathematical Breakdown
One of the most explosive claims in Burry’s new blog is that major tech companies are artificially inflating their earnings through accounting tricks—specifically, by manipulating depreciation schedules. Burry alleges that companies like Oracle and Meta are extending the "useful life" of their servers and GPU clusters on their balance sheets to reduce depreciation expenses, thereby boosting net income.
In accounting, depreciation is a non-cash expense that reduces pretax income. By claiming a server lasts 5 years instead of 3, a company reduces the annual expense charged against its revenue.
The Depreciation Impact Formula
To understand the magnitude of this alleged manipulation, we can look at the formula for Straight-Line Depreciation:
Where:
## D_{annual} ## is the Annual Depreciation Expense.
## C ## is the Cost of the asset (e.g., H100 GPU clusters).
## S ## is the Salvage Value (assumed to be near zero for obsolete tech).
## L ## is the Useful Life in years.
The difference in reported pre-tax income is:
By simply changing an assumption in a spreadsheet, a company could theoretically generate an extra $1.33 billion in profit without selling a single additional product. Burry argues this creates a "mirage" of profitability that masks the rapid obsolescence of AI hardware.
Simulating the "Paper Profit" Effect
For investors attempting to audit these claims, it is crucial to model how sensitivity to depreciation schedules can alter the perceived valuation of a tech giant. The following Python script simulates the impact of extending asset life on Earnings Per Share (EPS), a key metric for stock prices.
This script demonstrates that for a company spending $50 billion in Capex (a realistic figure for top hyperscalers), shifting the depreciation schedule from 3 to 5 years could artificially boost EPS by roughly $0.53. At a 30x P/E ratio, that "fake" accounting profit adds nearly $16 to the stock price, purely based on an accounting estimate.
Circular Economics and the "Round-Tripping" Risk
Beyond accounting, Burry raises a structural concern regarding the revenue sources of AI companies. He describes a phenomenon of "circular deals" or round-tripping. This occurs when a major tech giant (e.g., Microsoft or Amazon) invests billions into an AI startup (e.g., OpenAI or Anthropic), and that startup immediately uses the investment capital to buy cloud computing credits back from the investor.
Burry argues that this is not organic revenue; it is simply moving capital from one pocket to another to inflate top-line growth numbers. He challenges the sustainability of this model, asking, "True end demand is ridiculously small. Almost all customers are funded by their dealers."
The Volatility of Valuation
Burry’s warnings come at a time of extreme market valuation. He notes that the Capex-to-GDP ratio is approaching levels seen before the 2000 crash and the 2008 financial crisis. To quantify the risk, sophisticated investors often look at the Sharpe Ratio to see if the returns justify the volatility risk, but Burry suggests the risk is currently unmeasurable because the revenue underlying the valuation is artificial.
From Hedge Fund to Substack: The democratization of Insight?
The closure of Scion Asset Management to outside investors is a symbolic moment. By moving to Substack, Burry is bypassing the traditional financial media and the regulatory constraints of the SEC. This allows him to be more combative, more frequent, and more specific in his warnings.
For the retail investor, this offers a rare window into the mind of a contrarian. However, it also comes with a caveat. Burry has been famously early—sometimes years early. He shorted the housing market in 2005 but didn't pay out until 2008. In the interim, he suffered massive bleeding. His current bet against the AI boom faces the same temporal risk: the "bubble" could continue to inflate for years before bursting, driven by momentum and the genuine transformative potential of AI technology.
Whether "Cassandra Unchained" becomes a roadmap for the next crash or a testament to being "early is the same as being wrong" remains to be seen. But one thing is certain: Michael Burry is no longer whispering his warnings; he is shouting them.






















































Comments