This Stock is the Next Micron, Nvidia, TSMC!

[00:00:00]two stocks, one's up 191% this year, the other 262%.

[00:00:07]Go back a year, 514% and,038%.

[00:00:12]And somehow the market calls them cheap.

[00:00:16]Meet Samsung Electronics and SKH Highix, the two companies at the center of the AI boom, yet still trading at a discount to Micron, Nvidia, and many other AI favorites. So what's the market missing?

[00:00:34]First, who are these two? The kings of memory. In DRAM, the chip inside every phone, PC, and AI server, Samsung and SKH highex control about twothirds of the global market. In nan flash storage, same story, number one and number two.

[00:00:51]And now here's the strange part. The two market leaders trade cheaper than Micron, the company in third place on a forward PE basis, meaning based on earnings expected over the next year.

[00:01:04]Samsung Electronics and SKH Highex trade roughly seven times the earning. Micron on the other hand trades more than 10.

[00:01:13]Expend that beyond the memory. Some forecasts now suggest Samsung Electronics could generate around 488 trillion1 in operating profits in 2027, slightly above Nvidia's projected 485 trillion1.

[00:01:31]SKH Highix more than 400 trillion1. Read that again. The leaders on sale. So why part of it is structural? Samsung and SKH Highix may dominate the memory market, but they still sit outside the core US equity system. Yes, their global listings and institutional access points, including GDR listings and ETFs, but compared to US-listed peers like Micron, access for global capital still feels more limited. The bigger reason is psychological because memory has always been cyclical. Boom, bust, repeat. And even now in the middle of an AI frenzy, a lot of investors still can't shake the feeling that this cycle could eventually turn. That hesitation may be exactly why these stocks still look cheap. But what if this cycle is different?

[00:02:30]AI demand itself may be evolving. As the industry shifts from AI training to inference, AI systems aren't just using more memory. They're using more kinds of memory at the same time. HBM, LPDDR, NAND, you name it. And that may be making memory demand far more persistent than in the past cycle. As a result, AI companies are no longer buying memory quarter by quarter. They're starting to lock it in years ahead through long-term agreements or LTA, multi-year contracts, upfront commitments, even penalties for backing out. Because in AI race, computing power is starting to feel less like a normal business expense and more like a strategic infrastructure. And this shift is already starting to show up in public disclosure. Micron recently revealed what it called its first ever 5-year strategic supply agreement.

[00:03:26]SanDisk went even further, saying just three long contracts locked in roughly $42 billion in future revenue. Samsung and SKH Highix are widely believed to be signing similar deals too. Unlike their peers, however, much of it still sits behind NDA's non-disclosure agreements, and some industry speculation goes even further. Although SKHix denied the offer, Reuters recently reported that big tech firms have discussed helping fund memory capacity expansion themselves, essentially telling the supplies, "Build more capacity. We'll pay for it. just make sure the memory is ours. And that may be changing how investors value these companies.

[00:04:08]Historically, Korean memory stocks were often valued on book PBR, not earnings because the profits were never trusted to last. But if AI demand becomes more predictable and supply gets locked in years ahead, then old cyclical framework starts to break. Even more, LTAs may also be creating what analysts calls a dual market. As more memory supply gets committed through long-term agreements, the remaining supply becomes more limited, making the price more resilient. That's why some analysts including Nomra argue Samsung and SKH could eventually deserve valuation closer to TSMC than traditional commodity chipmaker. And Goldman Sachs makes a similar point. The market is beginning to view memory as less cyclical than before, which creates a strange possibility. Even after massive rallies, the market may still be catching up to their AI story.

[00:05:15]However, every bullcase has a flip side because every layer of this trade ultimately depends on hyperscalers like Alphabet, Microsoft and Amazon continuing to spend aggressively on AI.

[00:05:29]Wall Street expects that spending to surpass a trillion dollars within just few years. But if that pace loses, the whole AI supply chains feel it. higher rate, slower cloud growth, investors beginning to question AI return. Those are the real risk.

[00:05:48]And yet, at least as things stand today, major global firms like NMRA, City and JP Morgan continue to raise their targets on Samsung and SKH because they believe the AI cycle has still further to run. And some Korean analysts now even see Samsung rising towards 610,0001.

[00:06:09]SK Hinix as high as 4 million1 roughly double from current levels. Massive calls. But maybe that's what happens when an industry starts shifting from cyclical to structural. And right now Korea's memory giants are sitting at the center of that shift. The question is no longer whether AI will need more memory.

[00:06:33]The question is just how much of an AI era these companies end up paring. So next time we'll break down on how to actually invest in Korean AI winners.

[00:06:43]More case stocks, more of what's happening inside Korean market before global money fully catches on. Hit like, subscribe, and stay tuned.