SK Hynix Rebounds: 76% Margins and SanDisk's HBF Bet Revive the AI Storage Trade

SK hynix's quarter shifted the focus back to profitability

SK hynix's latest results made the memory trade harder to dismiss. The company reported 79.3187 trillion won in revenue, 60.5426 trillion won in operating profit, and a 76% operating margin. In a business known for cyclical swings, that level of profitability is hard to ignore.

More important, this was not just a spike in margins. SK hynix said high-performance products for AI servers led price increases, while management also emphasized mid-to-long-term growth alongside CapEx discipline. That gives bulls a case for a higher-quality cycle: strong pricing, disciplined spending, and profits that look more durable than a standard upcycle peak. Skeptics will still argue memory eventually resets, but the bar for treating SK hynix as more than a commodity trade has clearly risen.

SanDisk and OCP give the AI storage story a new node

The next catalyst is not just another AI headline. It is SanDisk's MOU with SK hynix on High Bandwidth Flash specification, plus the joint workstream under OCP. SK hynix and SanDisk say HBF is intended as a new memory layer between HBM and SSD, aimed at the AI inference era.

The partnership is still early, but the strategic angle matters. If SK hynix remains central as the standard takes shape, investors are not only backing today's HBM demand. They are also backing a possible next layer of AI storage monetization.

The July rebound looked sector-wide, not stock-specific

On July 30, the rebound hit several major names in the complex at once: SanDisk rose 26%, SK Hynix gained 17%, Western Digital climbed 15%, and Micron advanced 18%. That kind of broad move suggests renewed capital flow across the AI memory and storage stack rather than a one-stock pop. The move also came as confidence in AI infrastructure spending improved ahead of Big Tech earnings.

SK hynix was re-priced for resilience, not a perfect quarter

That context helps explain why the group bounced even after a mixed SK hynix quarter. In Q1, the company still delivered a record-high quarterly performance with a 72% operating margin, but revenue missed forecasts. Bulls can argue that pricing and product mix were carrying results before top-line consensus fully caught up. Bears can argue the opposite: expectations may have run ahead of shipment evidence.

The pullback made that split clearer. The sector had just gone through a rough stretch, with names down 18% from the peak; WDC fell 17% in the same period. Instead of breaking, the group rebounded quickly once sentiment improved. That suggests traders are still interested in the theme, but they are increasingly catalyst-driven.

SanDisk shows how far the optimism has already run

SanDisk is a useful gauge for how much of this trade has already been priced in. The stock had surged 574% year to date by late July, so this was hardly a sleepy name getting rediscovered. The fact that the group re-accumulated after such a move says conviction remains strong, but it also raises the bar for follow-through.

For SKHY, the next debate is pricing durability

After the reset and rebound, the key question is no longer whether AI storage matters. It is whether SK hynix can prove it is more than a one-quarter margin story.

Revenue guidance matters as much as margins

That debate was visible in Q1. SK hynix posted 52.58 trillion won in revenue against 53.55 trillion won expected, even while posting a 72% operating margin. Bears will say revenue should have won the argument. Bulls will say margin strength matters more because it points to favorable pricing and mix before the top line fully clears.

That interpretation is not purely speculative. Management said it has long-term agreements with around 10 key customers, and it confirmed that HBM4 achieves customer-required operating speeds. Those are useful indicators that demand is being secured ahead of the next print, even if investors still need top-line confirmation.

What to watch next

The next quarter needs to show that the Q1 revenue miss was an estimate issue, not a sign that customer pull is easing. For now, the setup looks constructive without being reckless:

• Positioning: constructive, but still catalyst-sensitive.
• Main proof point: revenue should clear expectations again while the high-margin product mix holds up.
• Watchpoint: another revenue miss, softer pricing, or weaker evidence around HBM4 adoption would shift the story back toward a traditional memory-cycle trade.