Michael Burry Is Betting 21% Of His Portfolio Against Big Tech - Here's The Catch WHY He's Wrong

Michael Burry became a Wall Street legend by being right once in 2008. But over the last 3 years, he has set tens of millions of dollars on fire trying to top tick the AI revolution. The quantitative receipts on his macro bear bets are brutal and undeniable. In 2023, Burry dropped $1.6 billion in notional put options against the S&P 500 and Nasdaq, only to be forced out of the position at an estimated 40% realized loss as tech momentum exploded. Shortly after, he bought put options on the SOXX semiconductor ETF at a $490 strike price right before the index ripped another 16% higher and completely vaporized his option premiums. Then came his high-profile $1.1 billion short bet against Nvidia and Palantir, where he stacked $110 puts on Nvidia, only to watch the semiconductor giant propel itself past a historic $5 trillion market valuation, leaving his derivative position deeply underwater. Yet despite a relentless string of crushed bear bets, the big short icon refuses to back down. In his latest filings and disclosures, short stock and put positions now consume over 21% of his total portfolio, aggressively targeting five core tickers across the tech and industrial value chain: Nvidia, Oracle, Palantir, Nubia's, and Caterpillar. But if you examine his execution mechanics closely, you can see the quantitative panic hidden beneath the surface. Burry just allocated nearly 4% of his portfolio into December Nvidia call options with mid-to-high $250 strike prices. Paying single-digit premiums per contract, he openly called this trade "upside insurance," explicitly admitting that post-earnings volatility spikes are a a coin flip that could instantly liquidate an unhedged short. This creates a bizarre valuation trap for the master contrarian. On paper, Burry admits Nvidia looks wildly undervalued at a forward PE near 23.5 times while generating over $96 billion in quarterly revenue at a 62% net profit margin. Yet, he insists that Big Tech's hardware hyper CapEx will inevitably crash into severe overcapacity, triggering a sudden devastating earning shock. So, is Burry's systematic call hedging a masterclass in tail risk management or absolute proof that hypergrowth technology is about to run him over once again? Let's dissect his five short targets one by one to see where the data actually points. First up is Nvidia, the ultimate anchor of Burry's bear thesis, where he claims the chipmaker is using dangerous circular financing to artificially inflate GPU demand. Burry argues that guaranteeing customer debt and offering residual value support up to $500 billion in AI infrastructure is classic vendor financing hiding an imminent collapse in compute demand. But, the raw quantitative data suggests Burry is fighting an economic powerhouse, not a house of cards. Nvidia isn't merely selling silicon chips. It is behaving like an AI nation-state strategically subsidizing its digital ecosystem to lock in decades of CUDA platform dominance. Wall Street concerns ignore the fact that Nvidia is partnering with Apollo, BlackRock, Blackstone, KKR, Brookfield, and Goldman Sachs to provide residual value support of up to 25% on $500 billion of infrastructure alongside a $105 billion backstop on OpenAI data center leases. This isn't artificial demand creation. It is rational ecosystem building with asymmetric informational advantages. Look at the numbers Burry is shorting. Nvidia boasts a breathtaking 62% net profit margin, while its implied PE multiple has compressed from 43.7 times down to 20.9 times and 23.5 times, making it cheaper than the Nasdaq average of 28.4 times despite projecting 61.96% revenue growth through fiscal 2028 to reach $665 billion. Furthermore, demand is broadening far beyond frontier labs. Nvidia's AI clouds industrial and enterprise segment exploded 138% year-over-year to $40.3 billion, generating nearly 59% of the sequential growth in total data center revenue. Downstream ROI is already proving real as seen in Meta's 14% increase in ad impressions and 12% rise in average ad pricing directly driven by AI recommendation engines. With $279 billion locked into supply chain purchase commitments to meet backlog, Burry's short ignores a fundamental market reality. Nvidia isn't manufacturing artificial demand. It is severely supply constrained. Next on Burry's hit list is Palantir, where he loaded up on March 2027 put options at low to mid $100 strike prices, betting on a massive collapse from its peak above $200 down toward $100. But betting against Palantir's operating leverage is proving to be a statistically lethal gamble. While Burry fixes his gaze on rich headline valuation multiples, the underlying fundamental engine is expanding at an unprecedented rate. Palantir's quarterly revenue surged from $1.18 billion to $1.94 billion over just three quarters, driven by triple-digit growth in its US commercial division, where customer conversion rates remain well over 100%. Even more staggering is its hyper profitability. GAAP net income margin exploded from under 10% in late 2024 to a massive 55% in Q2 2026, driving GAAP operating margins up to 47% and pushing its rule of 40 score to a world-class 155%. Because top and bottom line earnings are scaling faster than the share price, Palantir's trailing and forward valuation multiples are actually compressing even as the stock trades near highs, continually forcing Wall Street analysts to upgrade forward revenue models after every single quarterly beat. Burry is counting on retail price volatility to drag Palantir down to his $100 strike price, but shorting an enterprise software monopoly that converts enterprise AI demand into 55% GAAP net profits looks less like finding a valuation bubble and more like standing directly in front of an accelerating freight train. Third on Burry's hit list is Oracle, where he is betting that the company's aggressive AI data center expansion will collapse under a mountain of debt. Burry views Oracle's eye-popping $638 billion remaining performance obligation backlog as an accounting illusion, arguing that a massive $70 billion capital expenditure outlay for 2027 will wreck its balance sheet when customer demand evaporates. But the hard data contradicts his bear case at every level. Oracle finished fiscal 2026 with $67.4 billion in non-GAAP operating income and a 54% jump in operating cash flow to $32 billion. Inside the cloud engine, infrastructure revenue accelerated to 93% year-over-year, reaching $5.8 billion, while CPU and GPU infrastructure revenue surged 119% to $4.8 billion. Oracle added 1.2 gigawatts of data center capacity in 2026 and is adding another 1 gigawatt in Q1 2027 alone. Oracle isn't building empty data centers on hope. A staggering 98% of its available AI capacity is already contracted with actual infrastructure utilization sitting at 97.5%. Furthermore, the balance sheet risk Burry fears is heavily backstopped by customers themselves as the majority of Oracle's $85 billion in new quarterly backlog came through direct customer prepayments and bring your own hardware structures where clients supply their own expensive accelerators. Trading at a forward PE of just 18.2 times, a 21% discount to the broader technology sector despite projecting 34% revenue growth to nearly $90 billion in 2027, Oracle's fundamental metrics support a $185 price target based on an expected $8.05 EPS, making Burry's short bet look increasingly disconnected from operating reality. Fourth on Burry's list is Nebius Group, a fast-rising neo cloud provider where he is betting against the financial sustainability of high-growth infrastructure build-outs. Burry likely views Nebius as a speculative high-dilution trap pointing to its $5.75 billion convertible notes offering and active share dilution to fund massive capacity expansion. But once again, the hard numbers reveal an explosive fundamental engine. In Q2 2026, Nebius delivered a staggering 454% year-over-year revenue surge to $582.3 million while its annualized run rate skyrocketed nearly 600% to $3.0 billion. Far from an unviable money pit, Nebius generated $236.2 million in adjusted EBITDA delivering an elite 41% EBITDA margin and a 74% gross profit margin. Crucially, Nvidia holds a direct 9.3% equity stake in Nebius following a $2 billion direct equity investment securing its GPU hardware allocation and effectively removing the primary gating execution risk facing Neo clouds. Nebius secured debt at SOFR plus 2.50% while GPU rental spot prices hold firm at 52% of on-demand rates proving the market is structurally supply constrained. With new data center deployments generating over $20 million per megawatt in annual contract value and yielding rapid 22-month payback periods, Nebius is achieving a return on invested capital above 30%. Add in non-core equity holdings including a 28% stake in ClickHouse valued at $15 billion, an autonomous vehicle fleet in Avegant valued at $6 billion, and Taloka Arena, and Burry is shorting a company whose core infrastructure and equity assets could easily support a 300% valuation expansion over the next 5 years. Fifth on Burry's hit list is Caterpillar where his short bet signals a broader macroeconomic wager against the physical infrastructure powering the AI revolution. Burry treats Caterpillar as a bloated late cycle industrial whose premium 33.5 times forward PE multiple will crumble the moment hyperscaler data center build-out slow down. But the operational data reveals an industrial giant firing on all cylinders quietly becoming a massive direct beneficiary of AI power demands. In its latest quarter, Caterpillar delivered 24% revenue growth and a staggering 73% surge in diluted EPS to $8.17 while operating profits jumped 50%. The secret engine behind this acceleration is power generation where revenue exploded 72% year-over-year as big tech scrambles for heavy electrical generators to keep massive data centers running. Backed by a colossal 72 billion dollar order backlog and a 13% net profit margin, nearly double the industry average, Caterpillar is using its record cash flow to aggressively buy back 1.5 billion dollars in stock per quarter. While Burry shorts Caterpillar on the assumption that AI infrastructure spending is an unsustainable spike, Caterpillar's record back log and dominant grip on global power generation suggests the physical build out of artificial intelligence is only getting started.