Jim Cramer Says Marvell Is Expensive Unless Everything Works, and Then It Is Cheap
Jim Cramer laid out a binary bet on Marvell that hinges on one executive's performance at a single October meeting, and the math behind it changes everything...
MRVL beat earnings yet dropped 10%, with a trailing PE of 83 that only makes sense if fiscal 2028's guided 50% revenue growth delivers.
NVDA trades at just 32x earnings on a 63% profit margin, while Marvell's 14.5% operating margin must reach 38-40% to justify its premium.
Marvell's October 6 analyst meeting is a genuine catalyst, but data center at 79% of revenue means one hyperscaler pullback unwinds the entire bull case.
Act now: the analyst who called NVIDIA in 2010 just named his top 10 AI stocks — and Marvell Technology didn't make the cut. Grab the names FREE today .
Jim Cramer's line about Marvell Technology ( NASDAQ:MRVL ) on CNBC's Squawk on the Street from Jackson Hole on Friday, August 28, 2026, did more work than the market gave it credit for. He said Marvell is a very expensive stock unless everything works, and then it is cheap, framing the setup as similar to NVIDIA a couple of years back: hit the number and the multiple takes care of itself.
That framing matters because Marvell reported an earnings beat on August 27, 2026, and still closed the next session at $216.62, down 10.28% on the day. Carl Quintanilla described the setup as a bridesmaid in the wake of the NVIDIA number, which captures the mood without explaining it. The stock is up 155.27% year to date and 181.09% over one year, so a beat that only nudges revenue past consensus by 1.2% will not reset the story. The question is whether the CEO can carry a credible fiscal 2028 and fiscal 2029 story into the October 6 analyst meeting, because that is what Cramer is asking investors to underwrite.
What Cramer Actually Said About the Valuation
Cramer's exact wording was that Marvell is a very expensive stock unless everything works, and then it is cheap, like NVIDIA. If they can do the number, you are at 10x in 2028, which frames the setup as a binary outcome.
The trailing PE for Marvell is 83x, with a forward PE near 61x and a price-to-sales ratio of 24.89x. Those multiples do not survive a stumble in fiscal 2028, when Marvell has guided total revenue to grow approximately 50% year over year and data center to grow more than 60%.
Cramer's math works only if that guide holds and the custom silicon ramp lands. Miss either and the multiple compresses fast, because there is no dividend cushion at a 0.1% yield and no cheap book value at 12.09 times.
Topics in this story
Gathered from external sources. Rights to this text belong to whoever originally published it.