SanDisk: The Market Is Buying a Pricing Boom and Calling It Demand

The Consensus Narrative

The prevailing story on SanDisk is straightforward: AI data centers need storage, SanDisk makes enterprise SSDs, therefore SanDisk is an AI infrastructure play with a multi-year runway. Investors who acted on this thesis - from the S&P 500 addition in November 2025 to the Nasdaq-100 inclusion in April 2026 - have seen the stock surge from its February 2025 spinoff price of $38 to a 52-week high of $2,354, a gain exceeding 3,000%. The stock has since pulled back roughly 41% to around $1,390, and that pullback has prompted headlines declaring "buy the dip" moments, with prominent investors like Michael Burry publicly noting SanDisk's unprecedented increase of 3960% from May 2025 to May 2026 as part of a broader commentary on tech valuations.

That narrative has the direction right but the driver wrong. This is not primarily a demand story. It is a supply-constraint story, and the distinction matters because supply constraints end when someone decides to expand capacity.

The ASP-Unit Divergence

The clearest evidence that pricing - not volume - is carrying SanDisk's current upcycle comes from its own financial disclosures. Through the first nine months of fiscal 2026, SanDisk's revenue growth was comprised of a 76% increase in average selling price per gigabyte. That figure dwarfs any unit shipment growth the company reported. When ASPs of this magnitude account for the majority of revenue expansion, the cycle is being driven by scarcity, not by a surge in demand.

The same pattern appears in margin expansion. SanDisk's Q3 fiscal 2026 gross margin reached 78.4%, up from 51.1% in the prior quarter and from 7% at the time of the spinoff. Operating margins are at 40.73%. Free cash flow margins sit at 33.83%. These are not the margins of a company processing volume. They are the margins of a company operating under severe supply discipline - its own and, critically, its competitors'.

Table 1: SanDisk Earnings Beats Show Consensus Lagged the ASP Inflection (Public Data, from Market Estimates)

Quarter	Revenue ($B)	Consensus Revenue ($B)	Actual EPS	Consensus EPS
FY26 Q1	2.31	2.15	1.22	0.88
FY26 Q2	3.03	2.69	6.20	3.62

Consensus estimates consistently underestimated both revenue and earnings, and the miss widened as ASPs accelerated. The market did not know how fast pricing would move because it was modeling demand growth rather than supply scarcity.

The Capex Question

Here is where the story gets more interesting, and more uncomfortable for buyers at current levels. SanDisk's trailing twelve-month capital expenditure is $179 million. The company's gross profit growth year-over-year is 214.4%. Free cash flow growth is 1,423%.

SanDisk is generating enormous cash while spending almost nothing on capacity expansion. That is the source of the current margins, but it is also what makes them dependent on an external condition remaining unchanged: competitors must also maintain supply discipline. SanDisk is not building the fabs that would lock in its position. It is extracting maximum value from a constrained market while competitors - primarily Samsung and Kioxia - face their own capacity limits and cost structures.

For comparison, Western Digital (the HDD-focused entity that spun SanDisk off) spent $381 million in capex over the same trailing twelve-month period, despite generating lower revenue growth of 32% year-over-year. SanDisk's revenue grew 82.76% year-over-year with less than half the capex. The company is riding the cycle, not building through it.

The Backlog and the Long-Term Contracts

SanDisk reported remaining performance obligations and contracted backlog came in between $41.6 billion and $42 billion. Its entire 2026 enterprise AI storage capacity is sold out. Bernstein has estimated that 60% of SanDisk's volumes are covered by long-term agreements with floor prices of roughly $0.29 per gigabyte, which Bernstein notes is broadly in line with Q2 2026 average selling prices.

The backlog is real, and it provides downside protection against a sudden pricing collapse - but it does not answer the structural question. The contracts run three to five years. The question for an investor is not whether SanDisk will be profitable in 2027. The question is whether the pricing environment that justifies a 45.6x trailing P/E ratio persists beyond 2028, when the supply constraint itself may have been engineered away by the very profits it is generating.

Table 2: Valuation Comparison - SanDisk vs. Peers (Current Market Data)

Company	Market Cap	Trailing P/E	P/S	EV/EBITDA
SanDisk (SNDK)	$205.5B	45.6x	15.6x	36.5x
Western Digital (WDC)	$183.1B	28.7x	15.6x	46.4x
Micron (MU)	$1.034T	20.5x	11.5x	14.8x

SanDisk trades at more than twice Micron's trailing P/E despite Micron having comparable revenue growth and a significantly larger scale. It trades at 1.6x Western Digital's P/E despite WD having lower growth. The market is paying a premium for SanDisk's pure-play NAND positioning and its long-term contract coverage - but that premium assumes the current pricing environment is a plateau, not a peak.

The Two-Market Split

The NAND market has bifurcated into two distinct sub-markets. On one side: AI-grade enterprise SSDs, where supply is constrained, pricing is elevated, and SanDisk's BiCS8 chips (which deliver 15% to 19% higher density and 13% lower power consumption than competing solutions) command premium positioning. On the other side: consumer NAND - memory cards, USB drives, mobile storage - where pricing normalization is ongoing and margins are thin.

SanDisk's financial results overwhelmingly reflect the first market. Data center revenue surged 645% year-over-year in Q3. But the existence of the second market means that a general NAND recovery - the kind that would occur if Samsung or Kioxia chose to expand capacity - would flow into consumer and mobile channels first, compressing overall ASPs even if enterprise demand remained strong. The market is currently pricing SanDisk as if the two sub-markets are permanently decouled. They may not be.

What the Pullback Signals

The 41% decline from SanDisk's June high to current levels is not a dislocation. It is the market beginning to price in the possibility that the supply constraint is cyclical rather than structural. The stock has moved from valuations that assumed perpetual ASP growth to a level that still prices in multi-year above-cycle margins but no longer assumes they are permanent.

SanDisk's Q3 forecast of $7.75 billion to $8.25 billion in Q4 revenue, with EPS of $30 to $33, would maintain its premium valuation if delivered. Goldman Sachs raised its target to $2,200 on July 5 while simultaneously cutting its valuation multiple from 22x to 20x and doubling its normalized EPS projection from $55 to $110 - a move that acknowledges earnings power is the real anchor, not the multiple.

Investor Takeaway

The key issue is not whether AI will continue demanding enterprise storage. That direction is clear. The more important question is whether NAND suppliers maintain the supply restraint that is currently supporting SanDisk's pricing power. SanDisk's $179 million in trailing capex means the company itself is not the variable - it is reacting to the market, not shaping it. The variable is whether Samsung, Kioxia, and SK Hynix choose to expand capacity at a pace that erodes ASPs before SanDisk's long-term contracts mature.

SanDisk is a company that has benefited enormously from supply discipline it did not create and may not be able to sustain. Buying at 45.6x trailing earnings is a bet that competitors will remain disciplined through 2028 and beyond. That is not a demand bet. It is a supply-discipline bet - and it is worth remembering that supply discipline is what ends every memory cycle.