SNDK surged 11% in a single day; is the 24x P/E ratio cheap or a trap?

NEXT PICK · Market Insights

SNDK surged 11% in a single day; is the 24x P/E ratio cheap or a trap?

Friday, September 18, 2026

A stock that surged 10.99% in a single day, closing at $1,791.82, with a market cap surging to $262.36 billion, while its price-to-earnings ratio is only 24.29 times—this combination of "surge" and "affordability" is worth taking a look at.

What pushed SNDK to the top of trending searches today wasn't the earnings report or the new contract, but something extremely mechanical: S&P Dow Jones announced that Sandisk would officially enter the S&P 100 index before the market opened on September 21. All funds tracking this index must buy, regardless of how they view the NAND cycle.

Combined with the bullish research reports released by Mizuho and Citi that day, and the growth stocks taking a breather after the Fed's 25 basis point rate hike on September 16 and the decline in U.S. Treasury yields, plus the overall rebound in the storage sector—which had just pulled back about 10% the previous week—the question is, in this rally, how much is fundamental, and how much is just passive funds being driven by rules?

Latest price $1791.82 ▲ +10.99%

Price-to-earnings ratio 24.29 times ▼ Forward 8.37 times

FY26 revenue 20.25 billion USD ▲ +175%

A company that does only one thing turns it into a money printer

Sandisk spun off from Western Digital in February 2025 and is a company with an almost "single main business"—NAND flash-related products contribute over 99% of revenue, with no second leg to diversify risk.

The product line extends from flash memory wafers all the way to solid-state drives (SSDs), embedded storage, mobile storage cards, USB flash drives, and components, with customers covering computer OEMs, cloud service providers, data center operators, game host manufacturers, mobile phone manufacturers, and retail channels.

Its irreplaceability comes from its three-layer structure. The first layer is vertical integration: the entire chain from wafer to finished product is self-sustained, and few peers can achieve such depth.

The second tier is a joint venture with Kioxia's Flash Ventures. In January this year, both parties extended the joint venture agreement for the Yokkaichi plant until December 2034, and in August, announced plans to invest over 5 trillion yen (over 31.4 billion USD) in Japan over the next six years to expand production, including about 11.3 billion USD for the newly built Fab3 in Kitakami.

The third layer is the customer authentication barrier. Cloud service providers switching to storage providers have to redo the 6 to 18-month verification cycle and bear the risks of data migration and business disruption—once these costs sink, they become a moat that others can't enter.

If I had to give an analogy,

Sandisk doesn't seem like a chip seller; it's more like a water plant with water supply during the dry season—the water is still the same, but when the whole city is short of water, pricing power completely changes hands.

The "cheap" 24x P/E ratio and the "unreal" 84.6% gross margin

Let's get the numbers right first. Based on FY2026 realized earnings per share of $73.76, the current P/E ratio is 24.29 times, almost in the same range as competitor Micron's about 20.7 times; The forward P/E ratio is even more so, only 8.37 times.

Based on a market cap of $262.36 billion and FY2026 revenue of $20.25 billion, the price-to-sales ratio is less than 13 times; With free cash flow of $11.49 billion, the free cash flow yield (FCF yield) is about 4.4%. Looking at these numbers alone, it doesn't look like a stock that has risen more than ten times in a year.

So why be cautious? Because the denominator is problematic. Non-GAAP gross margin for the fourth quarter of FY2026 reached as high as 84.6%—not the usual gross margin for hardware companies, but rather "wartime pricing" when supply and demand are extremely imbalanced.

The quality of cash flow is indeed real. FY2026 operating cash flow was $11.67 billion, free cash flow $11.49 billion, almost one-to-one with net profit of $11.43 billion, indicating that profits were not heavily adjusted by accounting.

The board also cast a vote with real money

An additional $14 billion in buyback authorizations was added, bringing the total remaining licenses to $15.5 billion. Only a true cash cow would dare to spend money like this.

The question returns to the starting point

Is the 24x P/E ratio equivalent to EPS at the top of the cycle, or is it sustainable EPS? If NAND prices reverse in the second half of 2027 and the denominator shrinks by half, today's "cheap" will instantly become 50 times more expensive—that's why the same number can be read by bulls and bears to completely opposite conclusions.

As for whether today's gains were reflected in the current stock price,

Index inclusion was a one-time event; passive buying disappeared after September 21, raising prices rather than values.

Financially, it was a breakthrough, but technically, it was not far from its peak

Looking at the FY2026 curve, the word "breakthrough" is almost indisputable: revenue jumped from $7.36 billion in FY2025 to $20.25 billion, and net profit turned from a loss of $1.641 billion to a profit of $11.43 billion.

Revenue for the fourth quarter was $8.97 billion, a year-over-year surge of 372% and a quarter-on-quarter increase of 51%, with data center business up 437% year-over-year—this is no longer a linear growth slope, but a gap-like gap.

The company's guidance for the next quarter is even more extreme

Revenue ranges from $10.3 billion to $10.8 billion. Quarterly revenue alone exceeds the full FY2025 revenue of $7.36 billion.

But the market's attitude is not so uniform. Currently, the stock is about 19% above the 50-day moving average at $1505.53 and about 69% above the 200-day moving average at $1060.33, with an unmistakable upward trend; But the upper boundary of the 52-week range is at $2354.39—after today's big bullish candle, the stock price is still about 24% away from its all-time high.

In other words, while the fundamentals are still breaking through, the stock price has already broken through once and then given up another round of gains. The market is using prices to discuss in advance "when this round of price increases will end."

Five companies divided into different markets—who holds the moat?

NAND flash is a typical oligopoly market, with the world's top five brands taking up the vast majority of market share. In Q2 2026, the combined revenue of these five companies is about $68.87 billion, a 77% quarter-on-quarter increase.

Micron (MU) is the most direct competitor to the US stock market, with both DRAM and NAND deployments, making its business more dispersed and better buffering volatility. Its HBM products hold a leading position in the AI server market; Its valuation is also more "traditional," with a price-to-earnings ratio of about 20.7 times.

Samsung Electronics is the world's largest NAND manufacturer, leading in process, capacity, and brand; Sandisk's approach is to differentiate itself in niche markets like embedded storage and removable memory cards, rather than head-on competition for capacity.

SK Hynix is a global leader in HBM, and it competes more indirectly with Sandisk—the logic behind its benefit from rising server DRAM contract prices is almost the same as Sandisk's benefit from enterprise NAND price hikes.

After the split of its former parent company Western Digital, it focused on hard drives, forming a complementary rather than direct conflict with Sandisk's pure NAND route, but the compensation responsibilities and tax arrangements arising from the split still need to be tracked.

So why is Sandisk priced separately among the top five? The answer is most likely not in the product itself, but in Flash Ventures—a structure that shares capital expenditures and capacity with Kioxia—allowing it to achieve heavy asset output with a lighter balance sheet in an industry where single advanced wafer fab investments often exceed $10 billion.

The ceiling depends on how much money cloud providers are willing to spend

The driving force on the demand side is almost brutal

In Q2 2026, the top five enterprise SSD manufacturers combined revenue of $37.59 billion, a quarter-over-quarter increase of 103.6%—doubling in a quarter, which is almost unprecedented in mature hardware industries.

According to TrendForce's caliber, servers will consume 51.1% of NAND output in 2027, compared to 44.2% in 2026. Storage is shifting from consumer goods to infrastructure, with demand dominance shifting from phones to data centers.

Extrapolating the quarterly revenue of about $68.87 billion from the top five brands, the industry's annualized scale ranges from $280 billion to $350 billion; Sandisk, as one of the top five, assumes an 8% to 15% share, corresponding revenue potential of $22 billion to $55 billion.

The problem is, FY2026's $20.25 billion is already near the upper lower end of this range. Future growth can only come from two sources: grabbing market share, or continuing to rise prices—and the latter is precisely the most uncontrollable.

The supply shortage isn't over yet, and cash flow has already arrived

The logic of the bulls is not complicated

This is a structural shortage driven by AI infrastructure, and Sandisk happens to be standing at the narrowest bottleneck, having already turned its paper prosperity into cash.

▲ Bull Case

At the top of the cycle, growth and growth often resemble them most

Bears don't deny performance; what they deny is the reproducibility of these results—an 84.6% gross margin has never been a steady state for a hardware company; it's a product of shortages.

▼ Bear Case

Three time windows determine whether this narrative continues or shifts

In the coming year, SNDK's stock price will be dominated by three types of events: whoever cashes out first takes the pricing power.

•	Before the market opened on September 21, the S&P 100 was officially included: passive buying was concentrated and event-driven, followed by all the positive factors
•	FY2027 Q1 Financial Report on November 5: Validated revenue of $10.3 billion to $10.8 billion and guidance of earnings per share of $44 to $46, with the greatest directional impact
•	NAND supply-demand turning point in the second half of 2027: TrendForce expects supply growth to outpace demand, and if realized early, it will be substantially negative for prices and profit margins
•	The execution pace of Kioxia and Sandisk's over $31.4 billion expansion plan in Japan: short-term capacity assurance, medium-term supply increase across the industry

From being a "discarded pawn" during a spin-off to becoming the main figure who increased more than tenfold in a year

When it was spun off from Western Digital in February 2025, the market was generally skeptical about the profitability of this pure NAND company—FY2025 Q1 revenue was $1.7 billion, net loss $1.933 billion, and loss per share was $13.33, with valuation metrics rendered meaningless due to negative earnings.

The turning point will come in the second half of 2025. Enterprise NAND prices began to rise, with free cash flow of $488 million in the third quarter of FY2025 and earnings per share of $5.15 in the fourth quarter (compared to just $0.75 in the previous quarter), indicating marginal improvement in fundamentals, but the stock price gains were relatively restrained.

2026 will be the real double-click. Full-year revenue was $20.25 billion, net profit $11.43 billion, with a 52-week low of $93.54 and a high of $2,354.39, representing a yearly increase of over 540%—the market's judgment of the price increase was completely correct, but underestimating the extent of the increase was the biggest expectation gap of the year.

There's another detail that's easy to overlook

After hitting a high in August, the stock price experienced a pullback, but today's rally is a rebound from a pullback. The trend hasn't been broken, but the market disagreement over "how long the shortage can last" has already begun to show in volatility.

⚠️ Risk Notice • NAND supply is expected to shift to easing in the second half of 2027, with prices and gross margins facing systemic downward pressure • The current P/E ratio is based on profits at the top of the cycle; if profits are cut in half, valuations will passively double • The S&P 100 is driven by a one-off event, with short-term support disappearing after passive buying ends • NAND products account for over 99% of revenue, with no secondary business line hedging, resulting in high customer concentration • Less than two years after the spin-off, uncertainty regarding Kioxia's joint venture and tax compensation arrangements has not been fully resolved

🟡 Neutral Cyclical dividends are real, but valuations have already digested most of the shortage narrative in advance

💬 Discussion It's not a bubble, but it's already in the latter half of the price hike cycle. Real-time buy and sell positions, post SNDK to the official account.

Data source

•	Source: NextPick real-time snapshot + Sandisk official earnings and SEC 8-K/10-K announcements + S&P Dow Jones Index announcements + mainstream financial media and third-party research data