Michael Burry's AI Warning: A Value Investor's Checklist for the Bubble's End

Michael Burry's stark warning about an AI bubble is not a rejection of technology itself. It is a classic value investor's caution against capital allocation that destroys value. His central thesis is that much of today's AI spending is a zero-sum game, where companies chase competitors in a costly arms race with no durable margin improvement. The core question for investors is whether today's AI leaders possess wide, sustainable moats or are merely beneficiaries of a speculative capital cycle.

Burry frames this with a parable from Warren Buffett's early career. In the 1960s, Buffett's department store, Hochschild-Kohn, was in a competitive retail neighborhood. When a rival installed an escalator, Buffett felt compelled to follow suit. The result? Both stores spent heavily on a project that provided no lasting competitive advantage. As Buffett and Charlie Munger later described it, it was like "standing on tiptoe at a parade." Everyone had to do it, but no one gained a lasting edge. Burry applies this directly to the AI buildout: When the department store across the street put an escalator in, he had to, too. In the end, neither benefited from that expensive project. No durable margin improvement or cost improvement, and both were in the same exact spot. That is how most AI implementation will play out.

This escalator story is the essence of the value investor's lens. It asks: Does this spending create a wide moat that protects future profits, or does it merely raise the bar for everyone, leaving the industry as a whole no better off? The evidence suggests we are deep in the latter phase. The consensus estimate for 2026 capital spending by AI hyperscalers is now $527 billion, up sharply from earlier projections. Yet, as Burry notes, trillions of dollars of spending with no clear path to utilization by the real economy is so concerning. The market is already showing signs of rotation, with investors moving away from infrastructure companies where capex is debt-funded and earnings growth is under pressure, and toward those with a clearer link between spending and revenue.

The bottom line is that Burry's warning is a call to distinguish between capital expenditure that builds intrinsic value and that which simply maintains parity. For a value investor, the goal is to identify companies whose AI investments are truly like a new escalator that draws customers to the store, not just a costly upgrade that everyone else must copy. The current divergence in stock performance among hyperscalers is the market's first test of that distinction.

Assessing the Targets: Nvidia and Palantir Through a Value Lens

The value investor's checklist demands a clear-eyed look at two of the AI era's most prominent names. Both Nvidia and Palantir are capital-intensive, high-growth companies that have captured immense investor enthusiasm. Yet their current valuations and market dynamics present a classic tension between growth potential and price paid.

Nvidia's story is one of near-perfect execution and market dominance, but it now commands a premium that leaves little room for error. The company's price to earnings ratio is 48.37, a multiple that reflects not just its current profitability but the immense growth expectations baked into its stock. This is the price of being the indispensable supplier in a capital-intensive buildout. For a value investor, the question is whether this multiple is justified by a wide and durable moat. Nvidia's position in AI chips is formidable, but the escalator parable reminds us that even a dominant supplier can see its margins pressured if the industry-wide spending cycle becomes a zero-sum arms race. The high P/E ratio is a direct measure of how much the market is paying for future perfection.

Palantir presents a different kind of challenge, one rooted in sentiment and scale. The company has become a star of the retail investing world, with individual investors on track to buy nearly $8 billion in its stock in 2025. This enthusiasm has fueled a 150% surge in 2025 and a nearly 3,000% gain over three years. Yet this retail fervor often exists alongside institutional caution about valuation. The disconnect is stark: while retail investors see a story of relentless growth, institutional analysts are more focused on the path to sustained profitability. Citigroup's recent upgrade and $235 price target are based on a view of accelerating commercial revenue, but that growth must now be translated into robust and predictable earnings to justify the stock's run.

<p>

Both companies are in the midst of a capital-intensive expansion. Nvidia is building the engines; Palantir is building the software that runs on them. The risk for both is that their massive spending, while necessary for growth, does not immediately convert into durable margin improvement. As Burry's warning suggests, the market is beginning to separate the wheat from the chaff. For a value investor, the lesson is to scrutinize not just the revenue growth, but the quality of that growth and the financial discipline with which it is funded. The path to profitability is not guaranteed, and the current valuations leave little margin for a stumble.

The Capital Allocation Reality: Are Companies Building Value or Obsolescence?

The financial mechanics of the AI boom reveal a dangerous disconnect between spending and proven value creation. Analyst estimates for AI hyperscaler capital expenditure have consistently been too low, with actual spending exceeding 50% growth in recent years. This pattern of underestimation shows how difficult it is to forecast the scale of this buildout. The consensus estimate for 2026 spending is now $527 billion, a figure that itself may be conservative given the track record of upward revisions.

This massive capital allocation is the core of the value investor's concern. The market is beginning to rotate away from companies where this spending is debt-funded and operating earnings growth is under pressure. The divergence in stock performance among hyperscalers is the clearest signal. As Goldman Sachs Research notes, investors have rotated away from AI infrastructure companies where operating earnings growth is under pressure and capex spending is debt-funded. This is a classic market correction: it punishes capital expenditure that does not immediately translate into durable profit.

Michael Burry's specific warning about this phenomenon is a direct critique of this dynamic. He points to "supply-side gluttony," such as massive data-center build-outs and multibillion-dollar commitments without real end-user demand as a key risk factor for value destruction. In other words, the industry is building a vast, expensive infrastructure that may not be fully utilized. This is the modern version of the escalator parable-everyone is spending heavily to keep pace, but the collective return on that investment is uncertain.

The bottom line is that for intrinsic value to compound, capital must be allocated to projects that generate a return exceeding the cost of capital. When spending is driven by competitive fear rather than a clear path to revenue, it risks becoming obsolescence. The market's rotation is a vote of no confidence in that model. For a value investor, the question is not whether AI will be transformative, but whether today's capital-intensive bets are being made with the discipline of a true moat-builder or the recklessness of a zero-sum arms race.

Practical Takeaways for Patient Investors

For the patient investor, Michael Burry's warning is not a call to sell everything, but a framework for navigating a dangerous cycle. The key is to monitor the financial mechanics that separate durable value from speculative noise. The most telling signal will be a divergence between AI capital expenditure growth and operating earnings growth. When capex continues to climb while the profits it generates fail to keep pace, it validates the escalator parable. This is the precise dynamic investors are already rotating away from, as seen in the divergence in the performance of AI-related stocks. A widening gap between spending and earnings is the canary in the coal mine.

The next phase of the AI trade, as Goldman Sachs Research expects, will involve a shift in focus. Attention is starting to move from the infrastructure layer to two other phases: AI platform stocks and productivity beneficiaries. Platform companies, like providers of database and development tools, have recently outperformed, suggesting investors are seeking the next link in the value chain. Meanwhile, the group of potential productivity beneficiaries-companies that could see AI-driven efficiency gains-has underwhelmed, but analysts see an attractive risk-reward for investors seeking to expand their exposure beyond the infrastructure layer. This is where the long-term value of true moats may become clearer, as companies demonstrate a tangible link between AI adoption and improved economics.

The central risk, as Burry himself admits, is that the bubble's burst is unpredictable. He has stated it is impossible to time or predict the burst of the AI bubble. This makes a tactical exit difficult. Instead, the disciplined approach is to use the current volatility as a filter. Watch which companies are rewarded for their spending and which are punished. Favor those where capital allocation appears to be building a wide moat, not just maintaining parity. The market's rotation is already doing the work of separating the wheat from the chaff. For the value investor, the goal is not to catch a falling knife, but to identify the companies that will still be relevant when the speculative frenzy ends.