Transcription
You may not have heard of them yet, but you will soon. The AI boom is in full swing. And right now, memory is the biggest bottleneck in the entire industry. Compute demands chips. Chips demand memory. And one name dominates as the most profitable in the stack, Skhinx.
Today, I'm breaking down my $40,000 investment. I'll explain my exact positioning and entries. and I'll explain why SKHEX, the company that supplies over half of all high bandwidth memory used by Nvidia, is my single favorite stock right now. As always, I give you good memory. So, let's get started.
AI data centers are the lynch pin of this century's economy. This is not the dot bubble. This is explosive growth driven by structural demand. During the industrial revolution, machines replaced manpower. During the intelligence revolution, AI is on the cusp of replacing brain power. This once in a millennium event has driven hundreds of billions of dollars of investment into AI data centers. The biggest hyperscalers, Microsoft, Amazon, Google, and Meta are expected to spend over $700 billion in capital expenditure in 2026 alone. The vast majority of that capex is going directly into AI data centers and the chips that power them.
But here's what most investors miss. AI data centers don't just need Nvidia's logic chips. They need an entire cooperating orchestra of hardware to function. High bandwidth memory, standard DRAM, NAND flash storage, advanced networking chips, high-speed interconnects, cooling systems and power infrastructure, photonics and optical networking, and a whole litany of other components. And memory sits right there at the heart of almost all of it. Right now, it would not be an exaggeration to say that high bandwidth memory is the single biggest bottleneck to the most important industrial buildout. perhaps in the history of mankind.
But when analysts say memory, what do they actually mean? Let me show you something real. This right here is the computer that I built the CryptoJet YouTube channel on. By the way, I built this computer with the first $2,000 I ever made online. How cool is that? This right here is DDR4 RAM, the kind you normally find in a regular PC or desktop computer. It's fast, but it's designed for general use. Now, in the same PC, the Nvidia GTX 1070 graphics card has its own high-speed memory called GDDR5, the ancient precursor to modern high bandwidth memory. That memory sits directly on the GPU, so it can feed the graphics chip data extremely quickly. Man, that thing's dirty.
HBM, the memory that we're talking about with SKH, is the next evolution of this idea. It's like an ultra advanced superefficient version of the GDDR that's in this graphics card, but the memory is stacked in multiple layers directly on top of or directly next to the GPU for efficiency. This gives the memory that is directly on the chip massive bandwidth, way faster than regular DDR like this or even GDDR like this, which is exactly why it's become such critical fuel for training AI models. And right now, one company dominates the production of this critical high bandwidth memory. SKH Highix. I feel like I could sell this stuff for a pretty penny, even though it's pretty old now. I mean, hasn't memory gone through the roof?
SKHENX isn't just a leader in AI. They're dominant. So dominant, in fact, that SKHEX is actively turning down business because their supply is quote essentially zero. Other reports have SKHEX having sold out all of 2026 in legally binding contracts with the revenue not yet received. They've even sold out most of 2027 and 2028 supply. Chips that won't be made for another 2 years are already being sold out. In fact, supply is so much lower than demand that for the first time in history, SKHENX's customers are offering to finance the creation of entirely new fabs just to create chips for them. This includes ASML's EUV lithography machines, which are used to pattern the circuits on silicon wafers and individually costs hundreds of millions of dollars. ASML, the company, and their machines are worth an entire video on just to themselves. But suffice it to say that ASML's lithography machines have come to be known as the most complicated machines ever built in the history of mankind. And SKHENX's customers are so desperate for memory that they're offering to finance entirely new fabs, including these machines, so that Skhinx can build memory for them. This has been a popular trend for logic chip manufacturers for a very long time. But this is new for memory companies and it is one of the final nails in the coffin that tells us that the cyclical nature of memory is most likely behind us.
This all comes on the heels of a quote in late April from SKH vice president Kim Kite head of HBM sales and marketing saying quote HBM demand required of the company over the next 3 years far exceeds our supply capacity. The memory super cycle is here. Their latest quarter proves it. In quarter 1 2026, SKH Highix reported $36.25 billion in revenue and approximately $25 billion in operating profit with an elite net profit margin of almost 77%. That is one of the highest margins in the entire semiconductor industry, let alone in the entire world, beating out Nvidia, Micron, TSMC, and most others. Revenue grew 198% year-over-year, and operating profit exploded by almost 400%.
This combination of soldout supply through 2028, desperate customers, massive pricing power, and explosive demand has driven the stock to rerate higher, making SKHEX the second largest company in South Korea. Furthermore, depending on the source, high bandwidth memory is slated to grow by between 10 and 15x over the next 10 years with a continual average growth rate reported between 28 and 40%. Making this one of the fastest growing industries in the world and from what we can tell that growth rate may be seriously understated. It's also important to note that SKHENX is the single largest producer of high bandwidth memory in the entire world. According to data pulled by Grock, SKH Highix ranks the number one largest high bandwidth memory producer with potentially over 60% market share in the high bandwidth memory space. As recently as a year ago, they took over Samsung as the largest DRAM producer after Samsung led the market for 33 years. And within a year, they're over 20 points ahead of Samsung. And it is not done yet.
But before I hammer home exactly why SKHENX is such an incredible buy, I want to explain a core principle that we use on this channel. Many of you may rightfully point out that SKHEX has already undergone an incredible rally, up almost 150% year to date. And as such, we should wait for a pullback. Believe me, I have a wife and three kids. I'm not in the business of blowing all my money betting on meme stocks. >> I have less money than I started with. I started with zero. which is why we follow a very simple formula when picking companies. We buy great companies at attractive prices. We've already addressed why SKHEX is such a great company. But how can it possibly be at an attractive valuation after such a massive rally? That's because we don't buy charts, we buy companies.
As a technical analyst with almost a decade of experience and tens of thousands of hours of charting under my belt, I can tell better than most when a chart is overextended. And the reason that SKHEX isn't massively overvalued even after this monster run is because of something that we call the forward PE. The PE ratio compares a stock's earnings, its profit, to its market capitalization. It effectively tells you how many years it would take for a company to recreate its market capitalization through its profits. When we take a forward-looking PE, we compare its expected earnings to its current market capitalization, which gives us a forward-looking view of how today's share prices compare to tomorrow's profits. PE ratios are extremely helpful because they help us understand far better than a chart ever could whether a company is overvalued or undervalued.
Most semiconductor companies have a forward PE of 15, 20, 30, 50, 100, even over 150x. We tend to stay away from those companies, not because they're not great companies. Many of them are, which is why they got to being so premium, but they're just not at good enough prices for us to consider an investment in. However, SKHEX is. If you use the financials that we referenced earlier, you will find that SKHX currently has a forward PE of five. Not 15, not 50, just five. For reference, here's a picture floating around on X right now of every single company in the S&P 500 and their forward PE ratios. You can see Tesla at 188, Palanteer at 91, Coinbase at 69, Broadcom at 31, and Nvidia at 24. Over here, you can see Micron trading at a forward seven. Way down here at the bottom, you can see the only two companies at five or lower, Global Payment and Charter Communication. This places this company once again in an elite class of incredible companies that are expected to absolutely explode over the next few years. This is why we are buying SKH Heinix through various methods right now because it's a world-class company at an incredible valuation. And I fully expect SKHEX's market capitalization currently sitting around $800 billion to completely explode over the next few months.
And here's why. You see, SKHENX is suffering from something called the Korean discount. respectively people don't invest quite as much money and you don't expect quite as high of a PE ratio from Korean companies because of their troublesome neighbor to the north Korea [music] >> and as such often times South Korean companies demand a slightly lower forward PE which is part of the reason why it's five and not 10 but the other part of the reason why SKHEX doesn't have a significantly higher forward PE is also because it's actually very difficult to get access to for US-based investors which means most of that $800 billion investment is coming from institutions and Korean nationals. Not many people in the retail space in America are actually able to buy it right now.
As far as I'm aware, there's only three real ways that you can buy SKH. So far, I've used two of them, and I'm working on the third. The first way is that you can buy the new DRM ETF, which we're going to talk about in a minute. The second way is you can buy ticker symbol EWY, which tracks the entire South Korean economy. We talked about that in the last video, and we're going to talk about it again here in a minute. And the third way is that you can buy it on Interactive Brokers. But SKHX has actually not had their ADR in the United States yet. So, you can't just go and buy it in your traditional brokerage account yet. That's about to change because an ADR is coming in the next few months. SKHEX is actively pursuing a US American depository receipt listing with reports indicating a target time frame of June to July of 2026. This is why I said you haven't heard of SKHENX yet because you're going to when it ADRs because you're probably going to see several hundred billion dollars worth of investment flow into it as soon as that listing happens. And you can with a great deal of confidence based on everything we've gone over in this video expect SKH's stock to massively rerate to significantly higher forward multiples. Probably 15 to 20 if not even 25. Probably not 50 or 100 because of the Korean discount, but probably significantly higher than it is.
As we said in our last video as it pertains to Micron, data centers don't work without chips. Chips demand high bandwidth memory and GPUs to work right alongside each other. You do not have a functional data center without both. So I don't personally understand why Nvidia is commanding a $5 trillion market capitalization and SKHEX is only 800 billion. It doesn't make any sense and that's why year to date Micron and Skhinx have both rallied almost identical percentages about 160% whereas Nvidia has traded sideways at 15%. The rally already happened on Nvidia, but we're right in the middle of it for these two memory stocks.
So let's go over my specific investments, the rationale behind them, and what I'm planning on doing next. As I mentioned at the beginning of the video, I have deployed over $40,000 into certain plays to take advantage of the memory boom that is taking place right now. And I want to run through each one of them individually. These numbers are different from last week's video because we've invested another $10 to $12,000 on top of what we did last week. So far in the month of May, we have added $5,500 to our VU position and $4,700 to our QQQ position. This acts as ballast so that we can have a place to park profits. Moving down to the single stocks, over the last week, we have added $10,000 to MU with an average cost basis of around $630, and this is up about 12% already. We also added $1,000 to SanDisk, AVgo, and Nvidia, which are our plays to take advantage of the AI data center buildout. We also added a $3,000 position to Microsoft, but as I mentioned, none of these investments actually capitalize on SKH.
On the other hand, these two do. DRAM is the hottest ETF on Wall Street right now and it just added a billion dollars in a day. In fact, DRAM is actually tied right now for the fastest growing ETF in the history of the US stock market tied with the IBIT launch a couple of years ago. And a big part of the reason that it's grown so fast is because its two largest holdings are SKH and Micron Technology. As a result of its major exposure to SKH over the last week, we have added over $6,000 to the position and we are going to be adding another $5,000 on Monday. Next, our $2,000 EWY position is unchanged, but it does have major SKHENX exposure because it is the South Korean ETF. We also have about $1,600 in Socks. This already ran like 45%, so I trimmed a little bit of profits and rotated it into our base right here in VU and QQQ.
This takes our May 2026 investment so far to $35,800. And we have $5,000 sitting in cash ready to deploy into DRAM come Monday morning. So, by the time you see this video, that deployment will have already taken place. That brings our total May investment to $40,800. These new deployments have allowed our brokerage account to grow over the last month from $29,000 to a current $78,000. And as you can see here, this has led to a 4% growth of about $3,000 on the day and a total growth of 13.37% of about $4,200. This is actually about $4,000 lower than it should be because we've just rotated some gains and it hasn't actually updated yet. So, all in all, I'd say $7,000 in capital gains since our video last week is a pretty great start. And most of that has been driven by our investment in this memory stock thesis.
So, if you want to learn more about why Micron and many other names are absolutely incredible names to invest in, if you believe in the memory stock thesis that we went over in today's video, then make sure to check out that video right there on why Micron might be the next Invidia. If you enjoy today's video, make sure to hit that like button and subscribe to the channel. And if you want to learn more about all of this handson, I am a financial consultant. So if you'd like to speak with me directly, feel free to reach out to my executive assistant, Malachi Pats, at malachiats@cryptojab.com, and I'd love to book a call with you. While I'm not going to give you specific investment advice, I would love to help you in any way that I possibly can. Before I go, I do just first want to thank each and every single last one of you for watching as always and I will see you guys in the next video. Peace.