After Four Earnings Reports, SNDK’s Real Test Is How Big the Beat Can Be

These four earnings reports did not prove that the storage cycle is ending. If anything, they confirmed the opposite: AI data-center orders remain strong, storage prices are still rising, and supply remains tight.

What has changed is the market’s scoring system.

Previously, revenue growth and record profits were enough to push a stock higher. Now, even the strongest results in a company’s history can trigger a selloff if they fail to beat already-extreme consensus expectations.

That is also how I view the upcoming earnings report from $SanDisk Corp.(SNDK)$

The results will probably be very strong. The problem is that “very strong” may no longer be enough.

Seagate: The Only Company to Beat on Revenue, Earnings, and Guidance[Grin][Grin]

Seagate delivered the cleanest and strongest report of the four.

Quarterly revenue reached $3.629 billion, up approximately 49% year over year. Adjusted EPS came in at $5.71, non-GAAP gross margin increased to 52.7%, and free cash flow reached $1.1 billion.

Its next-quarter guidance was even stronger, with revenue expected to reach approximately $4.1 billion and adjusted EPS projected at $7.30.

Seagate sells HDDs rather than NAND, so its results cannot be used to predict SNDK’s profit directly.

However, the report proved that AI data centers are buying much more than GPUs. Hyperscalers have not slowed their purchases of high-capacity storage.

Among these four companies, Seagate left the least room for debate. Revenue beat expectations, EPS delivered a clear beat, and the next quarter is expected to accelerate again.

That is why STX delivered a much higher-quality report than the others.

Samsung: The Storage Supercycle Is Still Alive, but Record Profits Are No Longer Enough[Duh]

Samsung reported quarterly revenue of KRW 171.5 trillion and operating profit of KRW 89.5 trillion.

Its semiconductor division alone generated KRW 127.5 trillion in revenue and KRW 89.2 trillion in operating profit, implying an operating margin of nearly 70%.

More importantly, Samsung said demand for server DRAM, enterprise SSDs, and HBM continues to accelerate. Even as the company attempts to increase production, supply constraints are expected to persist.

Samsung’s operating profit beat the LSEG SmartEstimate by approximately 2.5%.

That is still a beat, but it was not a major one.

The message from Samsung is clear:

The storage price increases and supply shortages are real, but the market has already priced in much of the good news. Simply reporting the highest profit in company history is no longer enough to guarantee a higher share price.

SK Hynix: A 76% Operating Margin Was Still Considered a Miss[Cry][Facepalm]

SK Hynix reported revenue of KRW 79.32 trillion, up 257% year over year. Operating profit reached KRW 60.54 trillion, up 557%, while its operating margin climbed to an extraordinary 76%.

Yet the market had expected approximately KRW 84 trillion in revenue and KRW 64 trillion in operating profit.

That means revenue missed by roughly 5.6%, while operating profit missed by approximately 5.4%. Part of the gap came from HBM4 shipments and revenue recognition arriving later than expected.

This does not mean HBM demand has collapsed.

SK Hynix has signed long-term supply agreements with approximately ten customers, while demand for AI servers, HBM, and enterprise SSDs remains strong.

The problem is simply that the market wanted more.

A company can deliver a 76% operating margin and 557% profit growth, but if the numbers miss expectations by 5%, the stock can still get punished.

Kioxia: Extraordinary Numbers That Still Failed to Meet Expectations[Spurting][Cry]

Kioxia reported quarterly revenue of JPY 1.767 trillion, up 76.2% sequentially and 415.5% year over year.

Non-GAAP operating profit reached JPY 1.326 trillion, representing an operating margin of approximately 75%.

Revenue from its SSD and storage business reached JPY 1.175 trillion, up 95.7% quarter over quarter. Average selling prices increased by approximately 70%, while shipment volume also recorded low-single-digit growth.

But the market had expected even more.

Kioxia’s revenue missed Bloomberg consensus by approximately 4%, while operating profit came in roughly 7.3% below expectations.

That is why a report containing seemingly incredible numbers was still treated as disappointing.

However, the most important part was its next-quarter guidance.

Kioxia expects next-quarter revenue to reach JPY 2.39 trillion, representing another 35.2% sequential increase. Operating profit is expected to reach JPY 1.89 trillion, up 48.8% quarter over quarter.

This suggests NAND prices and profitability should continue rising for at least another quarter.

Kioxia’s problem was not weakening fundamentals. The problem was that market expectations were rising even faster than its profits.

SNDK and Kioxia Really Do Use the Same Factories

Many investors do not realize that SNDK and Kioxia are not simply suppliers, customers, or ordinary competitors.

The two companies have worked together for more than 25 years. They jointly develop and manufacture BiCS NAND and share production capacity at the Yokkaichi and Kitakami facilities in Japan through joint ventures.

Their Yokkaichi agreement has been extended through 2034, while the Kitakami partnership has also been extended to 2034.

More precisely, the factories are managed and operated by Kioxia, but approximately 80% of total capacity is shared equally between Kioxia and SNDK through their joint ventures. The remaining 20% belongs to Kioxia.

That means approximately 60% of the flash memory produced at these facilities goes to Kioxia, while around 40% goes to SNDK.

So saying that SNDK and Kioxia use the “same factories” is basically correct.

They rely on the same core manufacturing facilities, jointly developed NAND technology, and shared equipment investments. This makes Kioxia’s earnings one of the most direct leading indicators for SNDK’s production costs, capacity utilization, and NAND pricing.

However, the same factories do not produce identical financial results.

SNDK has its own customers, product mix, enterprise SSD business, pricing structure, and long-term contracts. The two companies do not necessarily receive the same product mix or realize the same selling prices.

Kioxia’s growth therefore cannot simply be copied directly onto SNDK.

My SNDK Earnings Forecast

SNDK is scheduled to report earnings on August 5.

The company previously guided for:

• Revenue of $7.75 billion to $8.25 billion

• Non-GAAP gross margin of 79% to 81%

• Adjusted EPS of $30 to $33

However, consensus estimates across different platforms have already risen to approximately $8.4 billion in revenue and $33 to $34 in EPS.

In other words, the market’s revenue expectation is already above the top end of SNDK’s official guidance.

Based on Kioxia’s ASP growth, its accelerating SSD revenue, and SNDK’s greater exposure to higher-value customers, my base-case forecast is:

This means SNDK will probably beat its original guidance, but it may not deliver a meaningful beat against the market’s latest expectations.

Three Numbers Will Decide SNDK’s Post-Earnings Reaction

First, can revenue reach at least $8.4 billion?

If revenue comes in near $8.2 billion, the result would still be close to the top of management’s guidance. However, the market could treat it as a miss.

Second, can EPS exceed $35?

An EPS result near $33 is already close to the market’s minimum requirement. SNDK may need to deliver at least $35 to create a genuine surprise.

Third, and most importantly, can next-quarter revenue guidance approach $10 billion?

Kioxia expects its next-quarter revenue to grow by 35% sequentially. If the same factory output, NAND price increases, and enterprise SSD demand flow through to SNDK’s results, SNDK should also guide for another quarter of substantial growth.

If SNDK reports only $8.2 billion in revenue, $33 in EPS, and then issues conservative guidance, the stock could repeat the post-earnings reactions of SK Hynix and Kioxia—even if the results are objectively strong.

My final view has not changed:

The storage fundamentals have not collapsed, and AI orders have not disappeared.

Seagate proved that data centers are still buying enormous amounts of storage. Samsung proved that supply remains tight. SK Hynix and Kioxia proved that market expectations have become extremely demanding.

SNDK is unlikely to report bad earnings.

The real question is whether a good earnings report can still be considered a beat in this market.