michael burry's AI Bubble Thesis Crumbles Under 2026 Market Data

The contrarian reputation of michael burry remains legendary in investment circles, yet his current bearish stance on artificial intelligence stocks faces mounting empirical opposition in 2026. While his historical prescience during the 2008 financial crisis established him as a market sage, recent years reveal a pattern of premature calls that have systematically underestimated the sustained growth in equity markets. His latest provocative claim—that AI represents a replayed dot-com bubble destined for inevitable collapse—warrants closer examination against contemporary market realities.

From Financial Prescience to Persistent Bearishness: michael burry’s Shifting Track Record

michael burry earned his legendary status by correctly anticipating the 2008 subprime mortgage crisis while managing Scion Capital, where he generated approximately $100 million in personal profit and $700 million for his investors through credit default swaps. The narrative was so compelling it inspired Christian Bale’s Oscar-nominated performance in the acclaimed film The Big Short.

However, the subsequent fifteen years tell a different story. As U.S. equity markets have demonstrated sustained expansion, michael burry has repeatedly issued pessimistic predictions that failed to materialize. His track record of disappointment culminated in 2024 when he shuttered his hedge fund, citing fundamental misalignment between his investment thesis and actual market behavior. This operational admission represents a tacit acknowledgment that his recent market calls have not delivered the returns that cemented his reputation.

The Case Against michael burry’s Latest Contrarian AI Call

michael burry’s newest assertion positions artificial intelligence as a speculative mania mirroring the 1999 dot-com frenzy, with identical catastrophic consequences on the horizon. This thesis rests on three specific pillars of concern, each deserving systematic scrutiny against available evidence.

The Depreciation and Accounting Manipulation Argument

The first pillar of michael burry’s bear case alleges that technology titans including Meta Platforms, Microsoft, and Alphabet are artificially inflating earnings through accelerated depreciation schedules on computing infrastructure. He specifically highlights Alphabet’s depreciation window of four to six years for server equipment as evidence of manipulated accounting.

This argument overlooks crucial technical realities. While GPU hardware does depreciate faster than traditional infrastructure, the broader AI ecosystem demonstrates extended utility cycles. Most data center AI infrastructure maintains productive capacity for 15-20 years. Critically, older generation GPUs retain substantial residual value through inference applications—the operational phase where trained models serve users—rather than exclusively serving the computationally intensive training phase. Second-market GPU economics continue to support meaningful asset valuations beyond their initial deployment window.

The Cash Flow Strain From Capital Expenditure Boom

michael burry’s second concern warns that unprecedented capital spending on AI infrastructure will ultimately constrain operational cash flows and create unsustainable financial pressure.

Empirical evidence contradicts this projection. The hyperscaler ecosystem—dominated by companies like Alphabet, Microsoft, and Meta—demonstrates not cash flow deterioration but accelerating operational cash generation. Alphabet exemplifies this trajectory, with cash from operations surging from under $100 billion to $164 billion in 2026 (trailing twelve months basis). Simultaneously, profit margins have expanded dramatically across the technology sector.

Crucially, documented returns on AI investment now exceed $3 for every dollar deployed. The latest wave of autonomous agentic AI—systems that execute human-like tasks independently—is generating documented cost savings exceeding 25% for implementing organizations. This efficiency dynamic fundamentally contradicts michael burry’s assumption of deteriorating returns on incremental capital.

Valuation Comparison: NVIDIA Versus Cisco’s 2000 Peak

The third component of michael burry’s thesis draws a valuation parallel between NVIDIA and Cisco at its dot-com peak, suggesting equivalent overvaluation and comparable decline potential.

This comparison proves unconvincing upon closer examination. When Cisco reached its March 2000 climax, its price-to-earnings multiple exceeded 200—reflecting speculative excess detached from fundamental economics. NVIDIA’s current P/E valuation stands at 47, representing a differential of more than four times lower. Historical data demonstrates the magnitude separating bubble-era valuations from today’s AI infrastructure leaders. The comparison itself contains the seeds of its own refutation.

Deconstructing michael burry’s AI Valuation Concerns

The gap between michael burry’s reservations and operational market dynamics appears most acute in GPU infrastructure economics. NVIDIA’s H100 data center processor—the specialized chip accelerating AI model training and large language model development—has experienced notable pricing momentum. Since mid-December, H100 rental prices have climbed approximately 17%, signaling persistent chip scarcity and robust infrastructure demand.

This price trajectory reflects the explosive adoption of agentic AI systems, developments that generate bullish implications for the infrastructure supply ecosystem. Companies like Nebius Group, CoreWeave, and IREN operate at the intersection of this surging demand. The GPU shortage phenomena extends beyond chip manufacturers to complementary infrastructure providers. Bloom Energy, whose power solutions address the energy constraints limiting hyperscaler expansion, operates in an increasingly favorable supply-demand environment.

Institutional Conviction in AI Infrastructure Stocks Strengthens

Market participants with substantial capital deployment capabilities have made their positioning clear. Advanced options traders executed significant bullish bets across the AI infrastructure space ahead of critical earnings cycles.

During Monday trading, call options buyers established 400 max-strike contracts in Bloom Energy, representing a million-dollar commitment to upside exposure. The options market simultaneously reflected extraordinary institutional confidence when a major trader deployed approximately $9 million on March $205 calls in NVIDIA. Notably, Bloom Energy shares appreciated 8% despite broader Nasdaq weakness, approaching a decisive breakout from a textbook weekly bull flag formation, while NVIDIA benefited from continued institutional accumulation ahead of earnings announcements.

The Verdict: Why michael burry’s Bear Case Fails to Account for AI’s Real-World Momentum

michael burry’s historical achievement as a contrarian remains beyond legitimate dispute. However, his contemporary bearish narrative encounters formidable data-driven headwinds. From escalating GPU availability constraints reflected in rental prices to documented efficiency breakthroughs from autonomous agentic systems, the AI market infrastructure demonstrates sustained bullish momentum.

The distinction between michael burry’s legendary 2008 call and his current AI positioning may ultimately rest on a fundamental principle: prescient contrarian calls require not only identifying genuine market dysfunctions but also correctly timing their resolution. Current evidence suggests that while michael burry has identified legitimate questions worthy of consideration, his timeline for AI market correction has proven premature, and the structural economics supporting continued AI investment acceleration remain robust for investors positioned to capitalize on sustained infrastructure demand.