SNDK Beat Everything. Wall Street Sold It.

A 372% revenue surge and $39.25 adjusted EPS still could not stop the selloff. The question is whether the $93.9B NBM floor changes the calculus entirely.

Opening: The Numbers That Did Not Matter

SanDisk (SNDK) reported fiscal fourth-quarter results after the close on August 5 that, by any conventional measure, were extraordinary. Revenue hit $8.97 billion, up 372% year over year and 51% sequentially. Adjusted EPS came in at $39.25, beating the Street consensus of $34.51 by 14%. Gross margin reached 84.6%. The company also authorized an additional $14 billion buyback, lifting total remaining repurchase authority to $15.5 billion.

The stock fell anyway. Shares dropped more than 8% in after-hours trading on August 5, then extended losses to roughly 10% through August 6 pre-market. The selloff erased in two sessions much of the recovery rally that had pushed SNDK from near $1,000 to above $1,400 in the five trading days leading into the print.

This is not a case of a bad quarter. Every reported number finished above both the company’s own guidance range and Wall Street’s pre-earnings estimates. What the market sold was the forward picture, and a very specific part of it: the Q1 fiscal 2027 revenue guidance midpoint of $10.55 billion came in below where many analysts had been modeling. That gap, not the reported results, drove the reaction.

The deeper debate, the one that actually determines whether this selloff is a structural warning or a positioning flush, centers on a single disclosure that most post-earnings commentary has underweighted: $93.9 billion in minimum contracted New Business Model revenue at floor pricing, backed by $16.5 billion in customer financial guarantees. That number is 4.6 times SanDisk’s entire fiscal 2026 revenue base. Understanding what it means and what it does not mean is the only analysis that matters right now.

Data Section: What the Quarter Actually Showed

The fiscal fourth quarter ended July 3, 2026. Sequential revenue growth of $3.04 billion came approximately one-third from higher volumes and two-thirds from pricing, according to the company’s earnings release. The $8.97 billion result topped the consensus estimate by roughly $485 million, or 5.7%. Adjusted EPS of $39.25 beat by $4.29, or 12.3%.

Growth was driven by the Datacenter and Edge segments. Gross margin of 84.6% was a record for the company at this revenue scale. The consumer business, by contrast, fell 32% sequentially to $556 million, a deterioration that underscores how concentrated the earnings power has become in hyperscaler-facing products.

For Q1 fiscal 2027, management guided revenue of $10.3 billion to $10.8 billion, non-GAAP gross margin of 83% to 85%, and non-GAAP EPS of $44 to $46. The EPS guidance midpoint of $45 was broadly in line with the $45.58 consensus. The revenue midpoint of $10.55 billion is where the disappointment lives. Sequential growth of roughly 18% from Q4 to Q1 is a substantial absolute step-up, but many analysts had been modeling a higher number based on extrapolation of Q3’s blowout trajectory.

The NBM contract disclosures deserve their own line items. Since announcing five agreements on the April earnings call, SanDisk signed five additional NBM deals by August 5, including three with new customers and two expansions with existing partners, bringing the total customer base to eight datacenter and edge accounts. Total minimum contracted NBM revenue at floor pricing: $93.9 billion. Financial guarantees through cash deposits and instruments: $16.5 billion. Remaining performance obligations at quarter-end: $59.8 billion, rising to $91.1 billion including two deals signed after the quarter closed.

Strategic Interpretation: Beat vs. Guide, and Why the Market Chose the Guide

Memory stocks do not trade on what just happened. They trade on where the cycle is relative to mean reversion. That is the lens every professional portfolio manager applies, and it explains why a historic Q4 beat was met with a double-digit selloff.

The Q3 result, reported in early May, was so far above expectations that it reset the Street’s model for every quarter that followed. Fiscal Q3 revenue came in at $5.95 billion, up 97% sequentially, against guidance of $4.4 billion to $4.8 billion. Non-GAAP gross margin hit 78.4%, against guidance of 65% to 67%. Non-GAAP EPS of $23.41 beat consensus of $14.66 by nearly 60%. A beat of that magnitude does not recalibrate expectations. It scrambles them. Analysts who then modeled Q4 by applying a similar upside coefficient built in an error that became visible only when Q4’s guidance miss landed.

The Q1 FY2027 revenue guidance miss is real in dollar terms. But the size of the business it implies is not weak. A $10.55 billion quarterly revenue run rate annualizes to more than $42 billion. One year ago, SanDisk’s quarterly revenue was below $2 billion. The framing that matters is not beat-versus-miss on a single consensus line. It is whether the company’s contracted business model structurally alters how the forward earnings range should be calculated during a downcycle. That is the argument the $93.9 billion NBM floor attempts to make.

Susquehanna analyst Mehdi Hosseini told CNBC that the central investor debate for SanDisk is not market share. It is whether the company can maintain gross margins approaching 80% as AI-related memory demand transitions from model training toward inference workloads. That framing is more precise than the guidance miss debate, and it is the right one.

Sector Implications: The Selloff Did Not Stop at SNDK

The post-earnings reaction spread beyond SanDisk. Western Digital, the former parent that also reported results this week after beating its own estimates, weakened simultaneously. SanDisk, Western Digital, and Seagate have previously traded in tight correlation during broad semiconductor risk-off sessions, with SNDK functioning as the highest-beta expression of storage sector sentiment.

The simultaneous selloff in multiple storage names points toward a sector-level reset rather than a company-specific story. When two companies both beat reported estimates and both trade lower on guidance, the market is expressing a view about the entire sector’s forward earnings trajectory. The concern is not that SanDisk had a bad quarter. The concern is that the hyperscaler NAND procurement pace that drove the supercycle could moderate before the next contract cycle is negotiated.

The NAND market itself is expanding rapidly. SanDisk’s management projects total NAND market revenue will exceed $300 billion in 2026 and approach $500 billion in 2027, with the datacenter share of total addressable market growing from 30% in 2025 to 50% in 2026. If those projections hold, there is substantial room for incremental demand to absorb new supply from Samsung, SK Hynix, and eventually YMTC. If they do not, the math runs the other direction quickly.

Options Market Analysis: IV Rank 76, Expected Move 26.6%

The options market priced a significant move into earnings, and the realized move confirmed those expectations. As of August 5, SNDK’s 30-day at-the-money implied volatility stood at 169.5%, with an IV rank of 76 out of 100 over the trailing 52 weeks. That elevated rank reflects premium that has been bid aggressively into the event, and it creates a specific post-earnings dynamic.

The front-month expected move priced by the options market was approximately 26.6%, or roughly $331 in either direction from the pre-earnings close. The stock’s realized move, including the after-hours decline, fell within that range. That is mechanically important: when the realized move is contained within the implied expected move, the post-event IV crush typically compresses premium across all strikes, hitting long premium positions in both directions.

The put/call open interest ratio stood at 1.23 before the earnings release, reflecting a skew toward protective positioning rather than directional bearish speculation. Put/call volume, however, ran at 0.79, suggesting that while open interest leaned toward downside protection, intraday flow was more balanced. The dealer gamma posture was negative before the print, meaning hedging activity could amplify moves in either direction rather than dampen them.

Max pain at expiration was pinned near $1,310. The stock traded at $1,350.50 into the close on August 5, slipping to $1,287.67 in after-hours before further weakness on August 6. The confluence of negative dealer gamma and a stock price moving toward max pain in a high-IV environment creates a mechanical friction point that traders should monitor.

Structured Trade Framework

Given IV rank at 76 and the post-earnings IV crush dynamic already in motion, selling premium into elevated volatility is structurally advantaged over buying it. The frameworks below reflect the three primary market views on SNDK as of August 6, 2026. These are analytical constructs, not instructions. Each involves risk of total loss or loss exceeding initial outlay where noted.

Bull Case: The NBM Floor Is a Structural Moat

If you believe the $93.9 billion in minimum contracted revenue at floor pricing represents genuine earnings protection through at least fiscal 2028, with CFO Luis Visoso confirming NBM gross margins of approximately 80% with upside if market prices rise further, then the selloff is a pricing dislocation rather than a fundamental deterioration.

The bull view is reinforced by the buyback: $4.5 billion repurchased in Q4 alone, with $15.5 billion in remaining authorization representing roughly 7% of the company’s indicated equity value. A company with minimal long-term debt, multibillion-dollar quarterly free cash flow, and a contracted revenue base covering half of fiscal 2027 output at 80% gross margins does not require premium valuation to be a compelling position at current levels.

Bernstein has raised its price target on SNDK to $3,000 from $1,700, maintaining an Outperform rating, on the view that the long-term agreements provide more meaningful downside protection than conventional take-or-pay contracts. The firm estimates floor pricing at roughly $0.29 per gigabyte, broadly in line with Q2 2026 average selling prices, and argues that even in a worst-case price collapse scenario worse than the 2010 memory downcycle, the contracts should significantly mute earnings downside in 2029 and 2030.

A defined-risk bull structure at current IV levels would be a cash-secured put at a strike below current price targeting the August 13 Investor Day as a catalyst, or a bull call spread using the September expiration that keeps maximum loss limited to the net debit. For traders expecting continued appreciation into the Investor Day, a call spread that captures a move toward the $1,400 to $1,500 range while capping upside risk above the higher strike contains the premium outlay in an elevated-IV environment.

Bear Case: Peak Cycle, Locked-In Pricing, and Customer Concentration Risk

The bear argument does not require SanDisk to be a bad business. It requires only that the NBM contracts were signed at or near peak cycle pricing, and that floor pricing structured above long-run equilibrium eventually becomes a renegotiation liability rather than an asset. NAND is a commodity. Commodity margins mean-revert. The question is whether the NBM structure delays or eliminates that reversion.

Consumer revenue falling 32% sequentially to $556 million is a concrete data point the bears can point to. The non-NBM business will continue to fluctuate with spot market pricing. If NAND spot prices weaken faster than NBM volumes ramp, the blended margin picture could compress even if the contract floor holds. Goldman Sachs has flagged the risk that structural NAND pricing change fails to materialize, that YMTC continues advancing its technology roadmap, and that SanDisk fails to gain sufficient enterprise SSD traction to maintain its current positioning.

Customer concentration is the structural concern embedded in the model. Eight hyperscaler and edge customers cover half of fiscal 2027 bit volume and approximately two-thirds of fiscal 2028 output. Concentrated demand is a double-edged position: it provides revenue visibility, but it limits pricing flexibility if those customers push back when more NAND supply comes online from Samsung and SK Hynix.

For traders expecting continued downside, a defined-risk bear structure would be a put spread targeting the $1,100 to $1,200 range through the end of August, capturing the current momentum without uncapped risk. Given IV rank at 76, buying outright puts remains expensive. A put spread reduces the net debit substantially while maintaining directional exposure to further weakness.

Neutral Case: Wait for August 13

The Investor Day on August 13, 2026, is the single most important near-term catalyst for resolving the guidance miss debate. Every institutional analyst on the Street will be listening for three things: how many additional NBM agreements are in advanced discussion, what the fiscal 2027 contracted revenue mix looks like by year-end, and whether the HBF prototype delivery timeline for second-half 2026 remains on track.

A neutral structure in a high-IV environment would be a short iron condor or a short strangle around current price levels, positioned to benefit from the post-event IV crush that typically follows a major catalyst. The risk with this structure is that the Investor Day produces a surprise move in either direction, so defined-risk via an iron condor, which caps maximum loss, is preferable to an undefined-risk strangle at current implied volatility levels.

Risk Analysis

The primary risk for long positions is margin compression. Gross margin is guided at 83% to 85% for Q1 fiscal 2027. That is still historically extraordinary, but CFO Visoso has indicated that NBM contracts are being signed at approximately 80% gross margins, with variable upside if market prices continue to rise. If spot NAND prices decline before additional NBM volume ramps, the blended margin could drift toward the lower end of guidance or below it, which would reprice the stock downward regardless of the contract floor protection.

Inventory build is a second risk. Management has flagged that higher inventory days will continue as the company builds stock to service NBM commitments. Elevated inventory in a commodity sector is a signal that experienced investors treat with caution, even when the underlying demand picture remains robust.

Capital spending will increase in absolute terms in fiscal 2027, guided at approximately 6% of revenue. BiCS8 and BiCS10 capacity investments, alongside the HBF prototype line, will consume cash that could otherwise support buybacks. Management expects bit growth in the mid-teens for fiscal 2027, slightly below the long-term mid-to-high teens target, partly because of the inventory build for NBM commitments.

On the geopolitical side, YMTC is the competitor risk that has been dismissed too easily. As the company advances its technology roadmap and Chinese domestic NAND demand deepens, incremental global supply could arrive faster than the market expects. Bernstein’s floor pricing estimate of $0.29 per gigabyte holds at Q2 2026 levels. A structural oversupply event of the kind that hit the NAND market in 2019 would stress-test that floor more aggressively than the current bull case contemplates.

Forward Outlook: HBF, the $500B NAND Market, and What Bears Are Underweighting

The two variables that the post-earnings selloff commentary has most consistently underweighted are High-Bandwidth Flash and the NAND total addressable market projection for 2027.

SanDisk is targeting first HBF samples in the second half of calendar 2026, with first AI-inference devices using HBF expected to sample in early 2027. The technology stacks NAND dies to deliver bandwidth comparable to high-bandwidth memory at a fraction of the cost, with the first-generation product targeting 1.6TB/s read bandwidth and 512GB capacity per 16-die stack. SanDisk and SK Hynix published the first open technical specification for HBF through the Open Compute Project on August 3, with Google and Tenstorrent already part of the consortium. A pilot production line is expected to begin operation around year-end, with commercialization targeted for 2027.

HBF is not a drop-in replacement for any existing product. It requires a different interface, and standardization through OCP is the mechanism by which the technology achieves broader adoption. The consortium’s structure, with Google directly involved, is a concrete signal of demand-side interest rather than a supply-side technology push in search of a market. That distinction matters for valuation.

The NAND total addressable market projection is the other underweighted variable. Management and TechInsights data both point to the NAND market exceeding $300 billion in 2026 and approaching $500 billion in 2027, with datacenter share expanding from 30% to 50% of total addressable market. SanDisk’s forward PEG ratio of 0.11 sits at a 90% discount to the sector median of 1.31. That gap either reflects an extraordinary opportunity or a market conclusion that current earnings growth rates are not sustainable past fiscal 2027. The Investor Day on August 13 is the next chance for management to close that gap with specifics rather than projections.

Stocks to Watch

SanDisk (SNDK). The central name. Revenue beat by 5.7%, EPS beat by 12.3%, gross margin at 84.6%, $93.9 billion in contracted NBM revenue at floor pricing backed by $16.5 billion in financial guarantees, $15.5 billion remaining in buyback authorization, and an Investor Day seven days out. The selloff is a positioning event driven by a guidance miss on a single revenue line. Whether it resolves higher or lower depends almost entirely on what management discloses on August 13 about the contracted revenue build for fiscal 2027 and the HBF commercialization path.

Micron Technology (MU). The cleanest comparable with overlapping enterprise NAND and AI infrastructure exposure. Micron’s own long-term agreement disclosures have supported bullish analyst views on the storage complex. Bernstein has noted that Micron’s pricing floors are considerably lower than SanDisk’s, meaning Micron carries more spot price exposure in a downcycle. If SanDisk’s Q1 guidance miss signals broader hyperscaler NAND procurement moderation, Micron faces similar pressure without the same floor protection depth.

Western Digital (WDC). The former parent weakened after its own earnings beat, confirming a sector-wide reset rather than a SanDisk-specific reaction. Western Digital has its own enterprise SSD ambitions and long-term agreement discussions. Traders watching the storage complex should treat WDC as a directional confirmation signal for SNDK rather than an independent thesis.

Kioxia. SanDisk’s joint venture partner at the Yokkaichi and Kitakami fabrication plants is the least visible but most operationally tied name in this analysis. SanDisk extended the Yokkaichi joint venture through December 2034. The new BiCS10 QLC die, announced jointly with Kioxia at the Future of Memory and Storage show, delivers a 60% bit density increase over eighth-generation QLC at more than 37 gigabits per square millimeter, the industry’s highest QLC density. If the NAND supercycle enters a more mature phase, fab utilization economics at these shared facilities will surface as a material factor.

NVIDIA (NVDA). The indirect signal. The AI inference workloads driving enterprise NAND demand run alongside NVIDIA processors in the datacenter stack, and both depend on the same hyperscaler capital expenditure cycle. A sustained moderation in NAND procurement pace and a moderation in GPU procurement pace would reflect the same capex rationalization pressure. The two demand signals are not independent. Any meaningful deceleration in NVIDIA’s data center order rate would stress-test the NAND market demand projections on which SanDisk’s fiscal 2027 guidance is built.

Action Checklist

• Verify the NBM math. $93.9B in minimum contracted revenue at floor pricing covers 4.6x fiscal 2026 total revenue. NBMs represent 50%+ of fiscal 2027 bit volume and roughly two-thirds of fiscal 2028 output. This is the single most important structural disclosure in the earnings package. Model it explicitly before forming a directional view.
• Separate the guidance miss from the business trajectory. The Q1 FY2027 revenue guidance midpoint of $10.55B missed consensus. The EPS guidance midpoint of $45 was roughly in line. The business is guiding to more than $42B in annualized revenue. The miss is real. The business is not deteriorating.
• Watch IV compression post-earnings. IV rank at 76 before the event means premium sellers have a structural edge in the days following the catalyst. Monitor whether IV collapses back toward the 52-week mean or holds elevated into the August 13 Investor Day.
• Track the put/call volume ratio and max pain. Pre-earnings put/call open interest was 1.23. Max pain was $1,310. With negative dealer gamma, moves away from max pain are mechanically amplified. The stock’s behavior relative to that pin level is actionable information for short-term positioning.
• Mark the August 13 Investor Day as the next hard catalyst. CEO David Goeckeler, CFO Luis Visoso, and the leadership team will address business state and forward outlook. The market needs specific fiscal 2027 contracted revenue mix data and HBF timeline confirmation. That event, not the earnings print, may determine the stock’s next 20% move in either direction.
• Monitor YMTC and Samsung supply signals. The bear case hinges on incremental NAND supply arriving faster than the market absorbs it. Any acceleration in YMTC’s technology roadmap or Samsung capacity ramp announcements is a direct stress test of the NBM floor assumption.
• Size for volatility. SNDK has a 52-week range of $40.53 to $2,354.39. It is a high-beta instrument in the semiconductor sector. The stock fell roughly 13% in a single broad sector selloff in July. For any defined-risk structure, confirm the maximum loss before entry. This is not a name where position sizing errors are recoverable quickly.

Analysis current as of August 6, 2026, pre-market. All figures sourced from SanDisk’s August 5 earnings release, conference call transcript, and options data as reported by OptiView and Unusual Whales. Forward-looking statements reflect analyst consensus and management guidance only. This is analytical commentary, not a solicitation or recommendation to buy or sell any security.