MILLIONAIRE MASTERCLASS INVITE: Altucher

Analyst price target moves following Tuesday’s Investor Day:

• Jefferies (Blayne Curtis): Buy, target raised to $450 from $325
• Evercore ISI (Mark Lipacis): Outperform, target raised to $433 from $275
• TD Cowen (Sean O’Loughlin): Upgraded to Buy from Hold, target raised to $350 from $245
• Raymond James (Simon Leopold): Described FY31 outlook as “extraordinary”

The Numbers Management Put on the Board

Start with where Marvell actually sits. Q2 FY2027 revenue was $2.739 billion, a record, up 37% year-over-year. Data center contributed $2.17 billion of that, or 79% of total revenue, up 46% year-over-year. That is the run rate management is scaling from.

Now look at what they guided. Marvell raised its FY2028 revenue target to $20 billion, up from its prior $18 billion target disclosed in August. Then the bigger number: the company expects annual revenue to reach $70 billion to $90 billion by FY31, driven by strong demand for AI infrastructure, custom chips, connectivity, and optical products.

The $80 billion midpoint and the implied EPS framework above $30 are far above what most models were carrying into the event. That gap is not a rounding error. It is the entire debate.

Where the $20bn Comes From

Data center revenue alone is expected to be roughly $18 billion in FY28. The incremental roughly $2 billion added to Marvell’s FY28 outlook is being driven by continued strength in scale-out optics, scale-up optics, and switching, rather than an immediate acceleration in custom silicon. That framing matters: the near-term beat is connectivity-led, which carries less execution risk than a custom-chip ramp.

Management expects FY2028 revenue to grow roughly 67% year-over-year, with data center revenue projected to increase around 80%. From a $2.74 billion quarterly base today, hitting $20 billion by FY28 implies quarterly revenue averaging $5 billion within six quarters. Steep, but the trajectory from the last two quarters supports the direction even if not the precise destination.

The FY31 Framework, Broken Down

At the $80 billion midpoint of the FY31 revenue framework, Marvell projects interconnect at roughly $37.5 billion, custom compute at roughly $30 billion, switching and storage at roughly $10 billion, and communications and other at roughly $2.4 billion.

Connectivity is doing most of the work. That is the signal behind Jefferies moving its target 38% higher in one session. Jefferies sees interconnect revenue potentially reaching $37.5 billion at the midpoint, with custom revenue approaching $30 billion by FY31, supported by opportunities in memory connectivity and AI inference.

Why the Sell-Side Moved So Hard

Three things drove the target increases simultaneously. First, the FY28 raise was concrete enough to model immediately. Second, the FY31 framework gave analysts a valuation anchor further out than they previously had. Third, the mix shift toward connectivity de-risked the story. TD Cowen said the fundamental growth drivers have shifted to Marvell’s connectivity franchise and that concentration and margin risk associated with custom XPU programs has been largely de-risked.

Evercore ISI noted that at an estimated $12 billion in calendar year 2026 revenues, Marvell has visibility into $70 billion to $90 billion by 2030, positioning it to deliver seven times revenue growth over four years, making it one of the fastest-growing companies in Evercore’s coverage universe.

Bull / Base / Bear

Bull: Hyperscaler capex sustains above $300 billion industry-wide through 2028. Marvell’s optical and custom-silicon wins compound, data center revenue hits $18 billion in FY28, and the street converges toward the $450 target.

Base: FY28 lands near $18-19 billion rather than the full $20 billion as one product ramp slips a quarter. FY31 targets stay intact conceptually but are discounted. Stock trades in a range anchored by the $350 TD Cowen target.

Bear: Custom chip programs require long lead times and substantial capital commitments from customers. If demand softens or competitors gain share, the company could face pressure to lower targets. A capex slowdown from any major hyperscaler collapses the FY31 model entirely.

Technical Overlay

MRVL closed at $287.01 on October 6, with a 52-week range of $70.69 to $329.88. The session high was $301.27 before the stock faded to close up 5.81%. That intraday rejection of $300 is the first level to watch. The prior 52-week high at $329.88 is resistance; the $271 area, which was the prior close, becomes the first floor. MRVL is trading near the top of its 52-week range and above its 200-day simple moving average.

Bottom Line

The FY28 target is aggressive but modelable. The FY31 framework is a capital allocation argument as much as a revenue forecast: it tells hyperscalers that Marvell intends to be a dominant supplier for the next infrastructure cycle, and it tells analysts they need a longer discount window to value the stock fairly.

At Investor Day, management emphasized that the global AI infrastructure buildout continues at massive scale and speed, supporting long-term demand for Marvell’s custom compute and optical networking components. The question is not whether that buildout is real. The question is how much of it flows to Marvell specifically, and at what margin. That answer will not come from Investor Day slides. It will come from Q3 results, currently expected around November 26, 2026.