TL;DR

· Goldman Sachs maintains a Buy rating on SanDisk with a $2200 price target, representing a 63.7% upside from the August 12th closing price.

· The valuation thesis is based on a target gross margin of around 80% for FY2028 to FY2030, as well as revenue and price visibility from a $94 billion long-term customer agreement.

· Low capital expenditure and a $15.5 billion remaining buyback authorization are expected to enhance EPS, but the long-term agreement is still not sufficient to completely mitigate the NAND cycle.

· HBF provides additional headroom for AI inference storage, although initial product samples are not expected until 2027.

Following SanDisk's 2026 Investor Day on August 13th, Goldman Sachs reiterated its Buy rating and $2200 12-month price target. This target price represents approximately a 63.7% upside from the closing price of $1344.29 on August 12th.

What supports the high price target is not just the fact that NAND prices are still elevated. Goldman is more concerned that SanDisk is attempting to reduce the cyclical nature of its legacy storage business through long-term customer agreements, low capital expenditure, and large-scale buybacks, transforming high profits into more stable free cash flow.

The long-term targets provided by the company for FY2028 to FY2030 are quite aggressive: revenue growth maintained at around 15% to 19%, non-GAAP gross margin around 80%, non-GAAP operating margin around 75%, adjusted free cash flow margin around 50%, and capital expenditure intensity kept in the mid-single-digit percentage range.

Whether these targets can be achieved is not about whether NAND prices can continue to rise indefinitely, but rather whether SanDisk can hold the price and demand floor through long-term agreements and then convert more profit into shareholder returns through manufacturing efficiency.

$2200 Price Target Betting on Sustained High Profitability

Goldman's $2200 price target is based on a 20x P/E ratio and $110 normalized EPS.

At the time of the report, Goldman stated that SanDisk's stock had risen approximately 15% post-Investor Day, but the firm believes that the company's better-than-expected long-term financial targets, capital return policy, and HBF technology potential have not been fully reflected in the valuation by the market.

According to Goldman Sachs' model, SanDisk's revenue is projected to increase from $20.248 billion in FY2026 to $71.041 billion in FY2028, while EPS is expected to rise from $70.16 to $284.98 during the same period. The EV/EBITDA multiple is forecasted to decrease from 6.3x to 2.5x, and the free cash flow yield is set to increase from 13.4% to 19.5%.

However, the FY2028 EPS target of nearly $285 is not the earnings base directly used by Goldman Sachs' price target. Goldman Sachs utilizes a normalized EPS of $110, indicating that its valuation has already factored in a certain discount to the current high NAND profit levels.

What is truly aggressive is SanDisk's own long-term operating targets. With approximately 80% gross margin and 75% operating margin, the company believes that its current profitability is not solely driven by short-term price increases, but may be sustained through customer agreements, product mix, and manufacturing efficiency.

This is also at the heart of the market debate: Is SanDisk merely riding the NAND upcycle, or has it established a new business model capable of reducing cyclicality?

$94 billion Long-Term Agreements to Set Revenue and Pricing Floor

The core of SanDisk's new business model is NBM, which stands for Non-Binding Memorandum.

According to the company, these agreements, with a total contract value calculated at the contract floor price, are approximately $94 billion, with around $91 billion in remaining performance obligations, covering 8 customers, and including about $16.5 billion in financial guarantees. The customers consist of 3 U.S. hyperscale cloud providers, with a weighted average contract term of about 4 years.

These agreements enhance visibility into future shipment volumes. SanDisk anticipates that around 50% of the planned shipments for FY2027 are already covered by NBM, with coverage extending to two-thirds in FY2028.

The pricing mechanism differs from traditional spot transactions. Recent contracts often have fixed prices, while prices are set with floors and ceilings in later years. Based on the company's Investor Day disclosures, even if executed at the contract floor price, the gross margin for related businesses can still reach around 80%.

This means that under normal contract execution, SanDisk is no longer fully exposed to NAND spot price fluctuations, and the revenue and profit floors are now clearer than in the past.

However, Goldman Sachs has not directly concluded that long-term agreements have eliminated industry cyclicality. The report clearly states that whether NBM can truly mitigate the NAND cycle remains to be seen, and its effects will take longer to be reflected in stock valuations.

The long-term agreement locks in a portion of customers, shipments, and price ranges, rather than the entire NAND market. Approximately 50% of the planned shipments for FY2027 and about a third for FY2028 are still uncovered. Supply expansion, decreasing demand, and variations in customer procurement pace could still impact overall pricing and profitability.

Therefore, the $94 billion contract seems more like a buffer for the cycle than a complete eradication of the cycle.

Low Capital Expenditure and Buybacks, Turning Profits into EPS

In addition to the long-term agreement, another advantage for SanDisk comes from manufacturing capital efficiency.

The joint manufacturing arrangement between SanDisk and Kioxia has been extended until 2034. The company stated that from 2021 to 2025, the SanDisk and Kioxia joint system contributed to 29% of the industry's bit growth but accounted for only 13% of the industry's capital expenditure during the same period.

By calculating the capital investment required to produce an additional 1EB of output by 2025, the industry's average level is approximately 2.7 times that of the SanDisk and Kioxia system. The company anticipates that by increasing single-wafer output, improving production efficiency, and maximizing the utilization of existing equipment and clean rooms, it can maintain double-digit bit growth while keeping the capital expenditure intensity in the mid-single digits of revenue.

This is particularly crucial for storage companies. Traditional NAND manufacturers often need to add significant capital expenditure during upcycles, which can then lead to oversupply from the subsequently added capacity. A lower capital expenditure intensity means that if SanDisk achieves its profit targets, more cash can be retained in the company rather than reinvested in the next round of expansion.

SanDisk has clearly outlined its capital allocation priorities: first, maintaining technological leadership and investing in the business; second, maintaining a debt-free and cash-sufficient balance sheet; third, returning 100% of excess cash flow after business reinvestment to shareholders, with a preference for share buybacks.

The company's board had previously authorized a $60 billion buyback, of which approximately $45 billion has been executed. With an additional $140 billion authorization, the remaining buyback capacity is about $155 billion.

If the adjusted free cash flow rate can indeed approach 50%, continued buybacks will reduce the outstanding shares and further amplify earnings per share. However, buybacks are merely a way to distribute cash flow and cannot replace the core operating objectives. Share buybacks contribute to valuation sustainability only if profits and free cash flow continue to materialize.

HBF Opens up the AI Inference Space, but Revenue Realization is Still Far Away

Beyond traditional NAND, SanDisk is also attempting to enter the higher-value storage segment of AI inference systems through HBF, or High Bandwidth Flash.

As model context lengthens and the inference chain grows, the KV Cache needs to store and quickly recall more intermediate calculation results. GPUs require not only computing power but also larger storage capacity and higher data read bandwidth. SanDisk refers to this challenge as the "Memory Wall" in AI inference.

The company anticipates that by 2032, the total addressable market for AI data center storage will reach 1.2ZB, with the KV Cache accounting for approximately 35%.

HBF aims to bridge the gap between HBM and SSD. According to SanDisk's technical vision, HBF can provide nearly HBM-level read bandwidth while increasing capacity to 8 to 16 times that of HBM.

In SanDisk's simulations, an HBF-only architecture can achieve the same token output with approximately half the GPU capacity of an HBM-only solution. If this result can be replicated in a real system, HBF could not only increase storage capacity but also potentially reduce the number of GPUs required for AI inference and overall capital expenditure.

However, this data is currently primarily based on company simulations and has not been validated through customer deployments. A Goldman Sachs report indicated that SanDisk's first HBF product has completed tapeout, and the initial samples are expected to be released in 2027. Subsequently, it will undergo customer validation, system architecture adaptation, cost evaluation, and mass production ramp-up.

Therefore, HBF is more suited to be viewed as a long-term upside option in SanDisk's valuation rather than a large-scale near-term revenue-generating business.

Goldman Sachs Bullish, But Long-Term Thesis Still to Be Tested by Cycles

Goldman Sachs remains bullish on SanDisk, firstly due to limited recent NAND supply additions, ongoing improvements in the company's product mix and customer structure; the longer-term thesis revolves around the NBM agreement, low capital expenditures, and repurchases potentially collectively enhancing earnings stability.

This thesis can be summarized as: long-term agreements increase shipment volume and price visibility, manufacturing efficiency converts profits into free cash flow, buybacks then convert cash flow into EPS, and HBF provides additional growth space for AI inference storage.

However, each aspect still requires validation.

While the NBM covers approximately 50% of the planned shipments for FY2027 and around two-thirds for FY2028, the remaining business will still be influenced by NAND prices and customer procurement pace. Competitors such as YMTC continue to advance their technology roadmap, which could also alter the industry's supply, pricing, and market share landscape.

Around 80% of the long-term gross margin is not solely dependent on contracts but also on product mix, manufacturing efficiency, and the overall NAND supply-demand balance. If industry prices experience a significant decline or if customer agreements are not executed as expected, the company's targets may still face pressure.

The HBF [Technology] needs to demonstrate that its technical model can translate into real deployment advantages. From 2027 sampling to revenue at scale, there are still steps involving customer validation, supply chain alignment, and system-level cost validation.

Therefore, this report does not provide a conclusion that the "NAND cycle has ended" but rather a conditional and high-resilience assessment: if SanDisk can execute its $94 billion long-term agreement as planned, maintain the gross margin at around 80%, and keep capital intensity in the mid-single digits, the current stock price may still undervalue its long-term cash flow; otherwise, if agreement execution, industry supply-demand, or HBF deployment fall short of expectations, the optimistic assumptions embedded in the $2200 target price will need to be re-evaluated.

Welcome to join the official BlockBeats community:

Telegram Subscription Group: https://t.me/theblockbeats

Telegram Discussion Group: https://t.me/BlockBeats_App

Official Twitter Account: https://twitter.com/BlockBeatsAsia