NVIDIA

NVIDIA delivered the most-anticipated print of the year on May 20, 2026 — and the stock barely moved, slipping roughly 1.5% after hours despite a clean beat. Q1 FY2027 revenue hit a record $81.6 billion, up 85% year over year and 20% sequentially, the fourteenth straight quarter of sequential growth and the largest dollar jump in company history. GAAP EPS of $1.87 cleared the $1.77 consensus, gross margin held at ~75%, and the board raised the quarterly dividend 25x (to $0.25) and added $80 billion to the buyback. The lesson of the quarter: the AI thesis is no longer the debate — the valuation is.

Data Center revenue was $75.2 billion, up 92% year over year, split between $60B of compute (Blackwell systems Jensen Huang called off the charts, with cloud GPUs sold out) and $15B of networking, which nearly tripled. Critically, NVIDIA shipped zero Data Center compute into China and its guidance assumes none going forward — the H20 export ban remains in force and CFO Colette Kress reiterated the China AI-accelerator market, on track toward ~$50B, is effectively walled off. Management guided Q2 FY2027 to $91 billion ±2% with margins near 75%, more than $4B above the Street, even with China zeroed out. The next proof point is the Vera Rubin ramp.

Michael Burry remains the most-watched skeptic in this name. His Scion-disclosed put complex against NVIDIA — roughly $186.6 million notional at peak — rests on a single structural analogy: NVIDIA is Cisco at the dot-com infrastructure top, a must-own stock that fell ~90% and needed 25 years to reclaim its high. His specific worry is hyperscaler depreciation accounting, which he argues capitalizes GPU purchases over useful lives that overstate AI ROI. After Scion deregistered with the SEC in November 2025, he has continued airing the thesis publicly. It is a structural call, not a near-term timing signal — and the record Q1 print did nothing to vindicate it yet.

On the institutional side, Goldman Sachs flagged that hedge funds cut NVDA exposure at the fastest pace in 13 years heading into the print. Insider behavior stays one-directional: CEO Jensen Huang's Form 4 history shows dozens of pre-set Rule 10b5-1 sales and zero open-market purchases over five years, and aggregate insider selling since January 2024 runs into the billions with no offsetting buys. None of that is necessarily bearish — 10b5-1 plans are scheduled and diversification is routine at this market cap — but the absence of a single insider buy alongside Burry's escalating put book is the contrarian flag bulls have to weigh against a Strong Buy sell-side consensus.

At roughly $222 per share in early June 2026, NVIDIA carries a market cap near $5.45 trillion — the largest chip company on earth — on a trailing P/E around 34× ($6.56 TTM EPS). On forward FY2027 estimates the multiple compresses toward the low-30s, still a premium to the broad market but modest against 70%+ revenue growth and ~75% gross margins. AMD trades on materially lower growth, Broadcom carries AI-networking exposure without a full GPU stack, and Intel sits on a turnaround with negligible accelerator revenue. The sell-side is overwhelmingly constructive: roughly 60 covering analysts skew Strong Buy, with an average 12-month target in the ~$296-305 band, implying mid-30s percent upside. The tension is simple — a record quarter that the tape ignored suggests the market is no longer paying up for beats and is instead waiting on Vera Rubin execution and any thaw in China.

1. August 26, 2026: Q2 FY2027 earnings call — the first test of the $91B guide and the quarter where the Vera Rubin ramp and Vera CPU revenue should begin to register; the highest near-term volatility event for the stock
2. H2 2026 (Q3 FY2027): Vera Rubin production shipments begin across AWS, Google Cloud, Azure, OCI, CoreWeave and Lambda — the key swing factor for whether ASP step-up extends pricing power into FY2028
3. Ongoing 2026: U.S.-China export policy — any approval of a China-compliant accelerator would reopen a ~$50B market currently zeroed out of guidance, a meaningful upside optionality the model does not assume

The most telling thing about NVIDIA right now is the non-reaction. A record $81.6B quarter, a 25x dividend hike, an $80B buyback and a $91B guide above consensus — and the stock slipped. That tells me the easy money on beats is gone; the market is now grading NVIDIA on execution, not surprises. I am watching two things into August. First, does the Vera Rubin ramp show up in the numbers on schedule, or does it slip? A slip is the one thing that breaks the growth narrative. Second, China — guidance assumes zero, so any policy thaw is pure upside optionality the model isn't pricing. At ~$222 and a low-30s forward multiple against 70%+ growth, this is no longer the screaming discount it was a month ago, but it is far from the bubble Burry describes. I treat the August print as the next real catalyst and would not chase the stock into it. This is analysis, not a recommendation — size any position to your own risk tolerance.