SanDisk Q4 Earnings: The NAND Recovery Is a Supply Story, Not a Demand Story

The Consensus Narrative

SanDisk reported record fourth-quarter results after its spinoff from Western Digital, and the conventional explanation is straightforward: AI inference workloads are creating unprecedented demand for NAND flash memory, driving the market higher. Revenue surged 372% year-over-year to $8.97 billion, non-GAAP gross margin hit 84.6%, and adjusted free cash flow topped $5 billion. The headline reads like a demand explosion.

That reading is backwards. The NAND recovery is not being driven by a surge in unit demand. It is being driven by constrained supply, pricing discipline, and a fundamental restructuring of how manufacturers sell to their customers.

The data from the earnings call makes this distinction explicit, and it changes how you think about the sustainability of the current cycle.

The Volume-to-Pricing Split That Matters More Than Revenue Growth

SanDisk disclosed that approximately two-thirds of its sequential revenue growth came from higher pricing and only one-third from higher volumes. The quarter saw roughly $3.0 billion of sequential revenue growth, of which approximately $2.0 billion was pricing and $1.0 billion was volume.

Revenue increased by 51% quarter-over-quarter while bit growth for the full fiscal year came in at mid-teens. That gap between revenue growth and volume growth is the entire story. SanDisk is not shipping dramatically more units. It is charging dramatically more for each bit, and the margin expansion to 84.6% from 78.4% a quarter prior confirms the pricing power is translating directly to profitability.

This is the supply discipline play. Post-2022, NAND manufacturers learned the lesson that every semiconductor supplier eventually learns: capex restraint controls the cycle more reliably than demand forecasting does. SanDisk's capital expenditure for the quarter was $562 million, or 6.3% of revenue. Management expects capex to increase in absolute terms in fiscal 2027 but decline to roughly 6% of revenue. That is not aggressive expansion. That is deliberate supply management.

Compare this to the traditional cycle where capex surges ahead of demand, unit shipments spike, and then ASPs - average selling prices - collapse because everyone is trying to fill new capacity. SanDisk is running the opposite playbook.

The Two-Market Bifurcation: Data Center Versus Everything Else

The NAND market is not recovering uniformly. It has split into two distinct sub-markets, and the split determines which revenues are durable and which are discretionary.

Data center revenue more than doubled sequentially to $2.98 billion in the quarter. Data center now accounts for 38% of SanDisk's bits exiting the fiscal year, up from 12% a year ago. Full-year data center revenue grew 437% to $5.15 billion. Management expects the data center share of the total addressable market to expand from roughly 30% in 2025 to 50% in 2026, with the overall NAND market expected to exceed $300 billion in calendar 2026 and approach $500 billion in calendar 2027.

Consumer revenue, by contrast, fell 32% sequentially to $556 million. Management attributed the decline to allocation choices - deliberately shifting bits away from lower-margin consumer applications toward higher-value data center and edge customers.

End Market	Q4 Revenue	QoQ Change
Data Center	$2.98B	+103%
Edge	$5.43B	+48%
Consumer	$556M	-32%

Edge includes PCs, smartphones, automotive, and IoT. Its 48% sequential growth is respectable, driven by higher storage content in premium devices and AI PCs. But the structural gravity is pulling toward data center, where AI inference workloads - KV cache, retrieval-augmented generation, and agentic AI - require high-capacity, low-latency NAND flash that commands premium pricing.

The implication is clear: SanDisk is increasingly a data center storage company masquerading as a general-purpose NAND business. The consumer segment is being managed down by design, not by weakness.

The NBM Contracts Are the Real Innovation

The most important disclosure from the call was not the revenue number. It was the status of SanDisk's New Business Models - long-term committed supply contracts that restructure the relationship between NAND manufacturers and their largest customers.

SanDisk has signed NBMs with eight customers representing a weighted average duration of over four years. These contracts now underpin more than 50% of fiscal 2027 bits and roughly two-thirds of fiscal 2028 bits. The total minimum committed revenue under NBMs is $93.9 billion at floor pricing, with $59.8 billion in remaining performance obligation. Financial guarantees totaling $16.5 billion back the counterparties.

NBM margins are running around 80%, with upside if market prices rise above floor levels. This is a fundamental shift from the spot-market model that has defined NAND for decades.

In the traditional NAND cycle, manufacturers and buyers negotiated quarterly volumes and prices, creating boom-and-bust dynamics when either side misjudged the other. The NBM model replaces that with multi-year visibility. SanDisk guarantees volume to hyperscalers and enterprises; those customers guarantee floor pricing. Both sides reduce risk. The manufacturer gets predictable revenue; the buyer gets guaranteed supply in a tight market.

The $93.9 billion in committed revenue, spread across multi-year contracts, is the structural anchor that transforms SanDisk from a cyclical commodity business into something closer to a contracted infrastructure play. That is the mechanism behind the 84.6% gross margin - not just pricing power, but pricing power with multi-year contractual backing.

Why the Stock Fell After a Record Quarter

SNDK closed 5.4% lower on the day and slipped another 7.29% in after hours trading despite beating on every major metric. The driver was guidance, not results.

SanDisk guided fiscal Q1 2027 revenue at $10.3 billion to $10.8 billion, with a midpoint of $10.55 billion. Wall Street consensus was approximately $10.82 billion. The midpoint came 2.5% below the Street estimate, and even the top of the range fell slightly short of consensus.

This is not weak guidance in absolute terms. The midpoint implies another $1.59 billion of sequential revenue growth and 15% sequential EPS growth to a $44-$46 range. But after several quarters in which SanDisk dramatically exceeded both its own guidance and analyst expectations, an in-line forecast can behave like a miss when the stock has been priced for repeated upward surprises.

The stock had run to a $200 billion market cap at a 17.5x trailing P/E. At that level, the market is no longer paying for the current quarter. It is paying for the trajectory. And the guidance suggests the steepest part of the curve may be flattening.

The Western Digital Contrast

The spinoff created two separate trading vehicles that illuminate the two-market split even further. Western Digital, now focused exclusively on HDDs post-spinoff, also reported Q4 results on the same day. Revenue came in at $3.75 billion, up 43.8% year-over-year, with trailing EPS of $16.75 and a P/E of 19.3x.

Western Digital operates in an entirely different part of the storage stack. Its 45.4% gross margin and $3.5 billion in trailing free cash flow reflect the economics of mechanical storage - lower margin, lower capex intensity, but growing on the AI infrastructure buildout that requires high-capacity, low-cost-per-bit archival storage alongside flash. The HDD market has its own supply discipline story, with factory closures and capacity consolidation supporting pricing.

SNDK and WDC now trade as independent entities. One captures the high-margin flash premium; the other captures the high-volume HDD base. Both benefit from constrained supply, but through different mechanisms.

Investor Takeaway

The key issue is not whether SanDisk's Q4 results were strong. They were exceptional by any measure. The more important question is whether the three pillars supporting those results - pricing discipline, supply constraint, and NBM contractual lock-in - can hold as the NAND market approaches the $500 billion TAM that management is forecasting for 2027.

The pricing pillar depends on continued capex restraint across the NAND industry. If competitors decide that 84.6% gross margins are an invitation to build more capacity, ASPs fall and the cycle reverts to its traditional mean-reverting pattern. SanDisk's own capex discipline at 6% of revenue is the right level, but the industry does not follow one company's rules.

The supply constraint pillar depends on bit growth staying in the mid-teens rather than accelerating. Management signaled that bits will remain on allocation beyond 2027, which supports this scenario.

The NBM pillar depends on counterparty execution. Eight customers controlling two-thirds of fiscal 2028 bits creates concentration risk alongside revenue visibility. If one or two of those customers pull back on AI infrastructure spending, the contracted floor pricing is protection, not a growth engine.

The stock's post-earnings sell-off reflects the market's recognition that the easiest growth has already been captured. The question for the next phase of the trade is whether SanDisk can maintain the supply discipline that created this cycle in the first place. If manufacturers stay restrained, the NBM model compounds. If they don't, the margins compress and the valuation at $200 billion becomes difficult to defend.