At Computex on June 2, Nvidia CEO Jensen Huang turned to Marvell CEO Matt Murphy on stage and declared: "The next trillion-dollar company, ladies and gentlemen." The stock climbed 33% in a single session, the kind of move that separates the believers from the people holding the bag when the music stops.
Marvell's market capitalization sat at roughly $275 billion as of early June. To reach a trillion dollars from here, the company must quadruple its valuation. The stock already trades at a trailing price-to-earnings ratio slightly over 100x, and data center revenue reached $6.1 billion in fiscal 2026, with custom silicon alone scaling to a $1.5 billion annual run rate. The arithmetic is simple: you're paying one hundred times earnings for a business that needs to grow into four times the current price without the multiple contracting. That is not a margin-of-safety trade; it's a bet on perfection.
The rally isn't baseless. Nvidia invested $2 billion in Marvell as part of an expanded collaboration centered on NVLink Fusion and custom AI silicon, and Marvell's full-year fiscal 2026 revenue grew to just under $8.2 billion, a 42% year-over-year increase, with first-quarter fiscal 2027 revenue jumping again to $2.4 billion. The company has carved out design wins with hyperscalers and locked in partnerships that matter. The problem is that the stock is priced as if those wins have already compounded for four years without a miss.
On June 8, Marvell saw its stock rise 7% to $281.88 following the announcement that it would join the S&P 500 index. Passive index funds that track the benchmark are required to hold its members, meaning they must buy both MRVL and FLEX before June 22. That's mechanical demand—trackers purchasing proportional to benchmark weight regardless of view, the kind of bid you can't trade against because it isn't driven by opinion. Profit-taking dragged Marvell shares down 16.7% during regular Friday trading hours after rival chipmaker Broadcom delivered earnings that failed to impress investors, a reminder that sector contagion moves faster than fundamentals when valuations stretch this far.
The recent volatility—up 33%, down 16.7%, then up another 7% on index inclusion—tells you more than any forecast. In the past three months, insider activity has shown no insider purchases, while insider sales amounted to $32 million, a quiet vote of no-confidence from the people who know the execution risk best. When insiders are sellers at one hundred times earnings and the CEO of your largest partner just called you a trillion-dollar company, you are the exit liquidity.
Jensen Huang is not wrong about the opportunity; he's early on the outcome and late on the valuation. The inference buildout is real, the custom-silicon demand is genuine, and Marvell is well-positioned to capture share. But the market just priced four years of flawless revenue growth, margin expansion, and multiple stability into a single week. The only way the call works from here is if Marvell executes perfectly, the macro cooperates, and the multiple holds. Miss any one of those and you've bought at the top of a move that front-ran what hasn't happened yet.