Sandisk Up 650%: The Right Advice Pointed at the Wrong Number

There's a piece of advice floating around Sandisk, and it's worth taking seriously precisely because it's unusual. A NAND investor told the market to "stay focused" on the flash-memory maker, arguing the AI boom can compound without the supply glut that has wrecked every memory cycle before it. The business genuinely looks better this cycle than it ever has. The problem is that after a roughly 650% run this year, the price already assumes the story is true — and in a memory stock, staying focused on the wrong number is how people get hurt near the top.

The boom that made the stock a rocket

Start with the math that made Sandisk, the NAND flash company spun out of Western Digital, one of the year's biggest winners. It just closed a fiscal year with revenue up 175%, to $20.25 billion, and its latest quarter grew another 51% on top of that. The engine is a pricing explosion: NAND contract prices have been rising roughly 70% to 75% quarter over quarter, pushing gross margins to about 71% and free cash flow to roughly $11.5 billion over the trailing year. The shares climbed from a 52-week low near $85 to around $1,764.

All of that surge is cyclical. NAND is the textbook boom-and-bust business: demand spikes, prices spike, every maker pours in capacity, and a year or two later supply floods the market and prices — and share prices — collapse. That history is exactly why this bull case is interesting. It claims this cycle is different.

What "stay focused" actually means

The stay-focused argument is not empty. Sandisk has signed long-term customer agreements covering roughly half of its expected 2027 bits and about two-thirds of its 2028 bits, with structured pricing and what management calls minimum financial guarantees. It also says it will not adopt every new manufacturing node just because it is available — adding supply only when the market actually needs it, a discipline the NAND industry has rarely shown. Management lays out targets of mid-to-high-teens revenue growth for 2028 to 2030 and promises to hand back all excess cash.

That is the most concrete answer this industry has ever produced to its own self-destruction problem. In past cycles, NAND makers built on hope and contracted almost nothing, so the downturn crushed everyone. Sandisk has actual commitments on the books. Credit where credit is due.

The price already banked it

Here is where the advice runs off the rails. Morningstar rates the stock two stars, gives it no economic moat and very high uncertainty, and values Sandisk at about $1,000 — roughly 40% below where it trades today — while expecting NAND prices to peak in early 2028 and then fall hard in 2029 and 2030. The same firm finds that less than 20% of Sandisk's contract value is actually financially guaranteed, and that customers can renegotiate terms if spot prices fall. The "minimum financial guarantees" in the bull pitch are a real but modest cushion, not a floor.

Now put a multiple on it. The stock trades at roughly 22 times trailing earnings — and those are peak-cycle earnings, not average ones. Morningstar's $1,000 estimate implies the through-cycle reality is worth only about half of today's price. And the company's own long-range plan admits the deceleration: after growing 175% last year, it models only mid-to-high-teens growth in 2028 to 2030. The market is paying a top-of-cycle multiple for a business whose own guidance is a sharp slowdown — in the one industry where a supply glut is already the standing forecast. TrendForce sees NAND undersupplied through this year but supply growth outpacing demand in 2027.

That is the whole tension in two lines: the business is more durable than any prior NAND cycle, and the market has already paid full price for that durability. "Stay focused" is fine advice — just focus on the right number. The right number is not the AI-demand story, which is real and not in dispute. It's the roughly 22 times peak earnings on a business that decelerates by design, sitting against an independent estimate that the structural case is worth about half of today's price. The structural improvement is the reason to keep Sandisk on a watchlist; the price is the reason it is not an obvious buy here. If industry supply discipline holds through the 2027 glut, this begins to look like a durable compounder on sale. If it doesn't, a seven-times-in-a-year run is exactly the exit ramp memory stocks reserve for investors who stayed focused on the narrative.