AMD, Micron, Marvell — How Far Above Fair Value Are They Trading?
AMD, Micron, Marvell — How Far Above Fair Value Are They Trading?
TL;DR AMD (+61% YTD), Micron (+68%), and Marvell (+85%) — three names run through the same 10-year framework. All three trade at roughly 2–3x their fair-value midpoints. A reminder that "good business" and "good price" are separate questions.
Same framework, three stocks
Intel and Nvidia each got their own deep dives. The remaining three from the chip group — AMD, Micron, Marvell — went through the identical 10-year scenario analysis. Reading them side by side makes one thing obvious: "AI exposure = automatic buy" is a dangerous shortcut.
All three at a glance
| Stock | Current | Low | Mid | High | Premium to mid |
|---|---|---|---|---|---|
| AMD | $340 | $42 | $160 | $46* | ~2.1x |
| Micron | $579 | $168 | $300 | $540 | ~1.9x |
| Marvell | $163 | $30 | $60 | $112 | ~2.7x |
*AMD's high came out below mid because of tighter PE assumptions and stricter margin scenarios. Either way, none of the three scenarios reach the current price.
1. AMD — "Nvidia challenger" premium fully baked in
AMD is up 61% YTD. At $340 the model midpoint is $160 — roughly 2x.
The scenario where AMD takes real AI share and secures the major cloud customers is plausible. The recent traction is real. But the market has already priced in "AMD secures a meaningful #2 position." Earning the #2 position in this kind of market is hard — Intel collapsed from 93% market share. Pre-pricing AMD's #2 status feels early.
2. Micron — risk of pricing the top of a memory cycle
Micron +68%. At $579, the midpoint is $300 — roughly 2x.
Memory chips are the short-term storage for AI servers. HBM demand is driving prices up, that part is real. But memory is a cyclical business. Pricing the top of a cycle on the assumption "this time it's different" is where investors get hurt.
In 2018 the narrative was "AI and cloud mean perpetual memory shortage." In 2019 memory prices crashed. For this cycle to truly be different, there needs to be evidence that the structural dynamics have changed. Right now that's still a hypothesis.
3. Marvell — the most stretched of the three
Marvell +85%. At $163, midpoint $60. About 2.7x — the largest gap of the three.
Marvell makes data center networking chips and specialized AI infrastructure silicon. It benefits directly from hyperscaler capex. But the price assumes that capex doesn't even pause for several years.
If Amazon, Meta, Microsoft, or Google trims capex even modestly — and nobody can guarantee they won't — mid-tier names like Marvell move the most.
Stepping back across all five
Run the same lens across all five names and the picture simplifies.
- All five carry the same AI-supercycle narrative
- All five have real, improving fundamentals to varying degrees
- But all five offer little to no margin of safety at today's prices
Which means "which of these five should I buy?" is the wrong question. The right answer is often "none." The market has tens of thousands of companies. There's no rule that you have to own one of the five most-talked-about names in a sector.
My take
The sector's fundamental story is real. AI, data center buildouts, U.S.-onshored fabs — all genuine. But good business and good investment are different questions.
One thing to remember: price changes everything. The same company at $17 and at $98 produces completely different expected returns. Buy a great company at the wrong price, and you may wait years for the price to come to you.
Related: Intel's 151% rally — priced in?, Is Nvidia really a value play now?
More in this Category
Buy-and-Hold Is Dead: What -85% to -99% Crashes Tell Us About 2026
Buy-and-Hold Is Dead: What -85% to -99% Crashes Tell Us About 2026
Over the past 12 months PayPal fell 85%, Rivian 92%, Beyond Meat 99.7%. Here's why "buy a famous company and bury it" no longer works in 2026 — and what changed.
Wall Street Doesn't Hold Anymore: How to Follow the Money's Footprints
Wall Street Doesn't Hold Anymore: How to Follow the Money's Footprints
Institutions moving hundreds of billions of dollars can't hide. Here's how to read the footprints they leave in price action, volume, and sector ETF relative strength.
Single Stock vs Basket Investing: Which Survives 2026?
Single Stock vs Basket Investing: Which Survives 2026?
In the AI era, guessing which company wins is far harder than guessing which industry rises. Here's a side-by-side of basket investing vs single-name bets — and when each makes sense.
Next Posts
Buffett Indicator at 132% — The Most Overvalued Market in 100 Years
Buffett Indicator at 132% — The Most Overvalued Market in 100 Years
The market-cap-to-GDP ratio sits 132% above its long-run average. Historically, when this gauge is 50%+ overvalued, the next decade has averaged -2.4% per year — yet selling out is still usually the wrong call.
Dollar-Cost Averaging Beats Market Timing — Even From the 2000 NASDAQ Top
Dollar-Cost Averaging Beats Market Timing — Even From the 2000 NASDAQ Top
Investors who started buying NASDAQ at the March 2000 peak — and lived through an 82% drawdown — still earned 15% per year if they never stopped. Five reasons DCA quietly dominates timing.
Principal-Driven Investing — How I'm Handling Intel ($17 → $110) and AMD
Principal-Driven Investing — How I'm Handling Intel ($17 → $110) and AMD
The five tenets of principal-driven investing, applied to two real semiconductor names: Intel, which ran from $17 to $110 in 12 months, and AMD, where the story is real but the price is dangerous.
Previous Posts
Intuit Down 50%: Is the AI Fear Justified at 16x P/FCF?
Intuit Down 50%: Is the AI Fear Justified at 16x P/FCF?
Intuit, owner of TurboTax and QuickBooks, has been cut in half. The FCF multiple compressed from 40x to 16x, yet ~85% small-business accounting share and $6.84B in annual free cash flow are still intact.
Salesforce Down 28%: What a 'Low-Quality Beat' Hides at 12x P/FCF
Salesforce Down 28%: What a 'Low-Quality Beat' Hides at 12x P/FCF
Salesforce has dropped 28% over six months. Guidance of 10–11% growth missed the 5-year average of 15%, earning the 'low-quality beat' label — yet 12.3x P/FCF is rare among large-cap SaaS names.
ServiceNow: Can AI Really Replace It? The Invisible Moat of Switching Costs
ServiceNow: Can AI Really Replace It? The Invisible Moat of Switching Costs
ServiceNow is down 50% from its 2025 highs on AI-replacement fear. But the real question is whether you can rip a system embedded in hospital, bank, and government workflows out over a weekend.