Asian Memory Stocks Explode – Samsung, SK hynix & Kioxia

QQQ3, the 3x leveraged, daily rebalanced, Nasdaq 100 tracker, is hitting new highs; that is a good omen for future progress. As I have noted on numerous occasions, the secular bull market in US shares is one of the wonders of the known universe. Anyone, Zack Polanski, who thinks wild anti-wealth creation strategies will make for greater general happiness, should look at the chart of the Nasdaq Composite index below.

Nor should they imagine that all the money belongs to Elon Musk. There is a reason why 39m Mexicans are living in the USA and why in the 19th century Italians and Irish poured into America.

As of early 2026, the average annual salary in the U.S. is approximately $64,505 ($31.01 per hour), according to Forbes Advisor. Data from the Social Security Administration (SSA) puts the 2024 national average wage index at $69,846, while the median annual earnings for full-time workers is generally lower, estimated around $63,360, per Wikipedia. [1, 2]

This is around three times the global average. America is an incredible engine for the creation of prosperity. Shareholder capitalism of the type so enthusiastically practised in the USA works incredibly well and also sits well with democracy and freedom. It is a pity that many people cannot see the bleeding obvious even when it is in front of their noses.

Before I talk further about the Asian memory stocks, read what a BlackRock executive, Robert Kim, said about the impact of AI.

Other industries may be in store for the same reckoning that software has seen, says Tony Kim

AI has the whole world economy in its hands, says a BlackRock executive.

Artificial intelligence isn’t just driving the stock market; it’s rewiring the entire global economy.

That’s our call of the day from the head of BlackRock’s fundamental equities global technology team, Tony Kim, who says the AI trade has staying power because it’s totally rewiring the global economy.

“The trade continues because it’s the equivalent of 10 Manhattan Projects going off at the same time. This is the new reality,” Kim said on a panel at the Milken Institute’s global conference in Los Angeles on Monday. “You have five foundation labs in the U.S., five in Asia and China, and each one is the equivalent of a Manhattan Project.”

Of the $110 trillion in global GDP, AI capital expenditure represents $1 trillion this year, with another $7 trillion to $8 trillion added in the next five years, he said. The message is that AI is driving a fraction of the global economy now, but that will only get bigger. “I think people need to just get over it. This is the new reality.”

To make his global-reshaping point, Kim dug into an explanation of the so-called AI stack. The bottom layer is the physical part of AI – compute, chips, data centers and power – where much of AI value lies and capex is being allocated. “Nine of the 10 biggest companies in the world build chips. The only one that doesn’t is [Saudi oil giant] Aramco,” he said.

Kim said the global economy is transitioning to a “token economy,” with AI output the product. “Tokens are revenue, and to build the token economy you need to build these token factories.”

The second layer is the intelligence or model layer, with SpaceX, Anthropic, OpenAI and others worth $1 trillion each if not more. At the top, which is two-thirds of global GDP, he said, are services and applications, where “the jury is out.”

To date in 2026, he said, $8 trillion in value has been added to both the foundation model and compute layers, but the top layer has lost $1 trillion to $2 trillion in market cap.

“So you’re seeing the market also rewire itself to this stack. Why does this continue? It is the refactoring of the global economy to what I call this AI-token economy. And many of the truths we held before AI are being questioned,” he said.

As for the state of software, following the sector’s implosion earlier this year, Kim said that it’s “both dead and we’ll have more software than ever.”

He explained that the 20-year era of software-as-a-service, or SaaS, ended in 2023. “In the world post-AI, in the last three or four years, we’ll have more software created now than ever in all humanity – 24/7 the generation of tokens and code will be near-infinite. You’ll have billions and billions of agents generating code.”

“That’s what we’re seeing. We’re seeing the market adjudicate moats. What’s the moat? It’s not just the moat of software; it’s basically every industry in the S&P 500. It started in software, but what you thought was perceived moats and duration, quality and all of these things are being questioned. If you have a world where there’s infinite code, DIY customization, where is your moat?” Kim asked.

Kim said companies that achieve scale, such as OpenAI’s potential $1 trillion market valuation, will “basically suck the market cap out of all others.

“My view is that value is accruing to those who exhibit these power-law [advantages]. It’s coming at the expense of those who do not. I think you’re going to see more and more dispersion, in every industry,” he said.

As for the losers? “There’s a backlog of a 1,000 unicorns that were built in the pre-AI era that will never exist,” Kim said.

I think he is spot on with this rewiring idea. It is all change, and there will be huge winners and huge losers.

On the Asian front, South Korean giant Samsung Electronics’ Memory business is superstrong.

Overall results for Samsung were incredible.

There is no doubt that Samsung is doing well with its large Memory business, but it is maybe not as exciting as some other opportunities in the sector; maybe safer, though, as it has just joined the $1 trillion-plus club.

This is another incredible chart. The whole memory sector looks like this. The company is on a roll, to put it mildly.

SK Hynix, which is another South Korean company, stands at the heart of the AI revolution.

As the AI era accelerates, SK hynix is evolving from a memory supplier into a “Full Stack AI Memory Creator.” This represents our commitment to go beyond meeting customer demands — to become a co-designer, co-architect, and ecosystem partner that drives breakthroughs in AI computing together with our customers. By integrating advanced DRAM and NAND technologies, SK hynix will continue to build the core value of the AI era through its full-stack AI memory innovations.

I am very impressed with SK hynix, but they are not yet on the IG platform. They are pursuing an ADR (authorised depositary receipt) to be traded in New York, and as soon as they have that, they will be listed by IG.

The last of my Asian trio is Tokyo-quoted Kioxia, which has a close relationship with SanDisk.

The chart is super strong and looks similar to the SanDisk chart. Fundamentals are equally strong.

Results are imminent from Kioxia, but judging by the share price performance, they are expected to be good. They are available for trading on the IG platform because they are quoted in Tokyo. Japan is nine hours ahead of the UK, so trading on the TSE begins at midnight UK time.

Kioxia is an amazing company which invented NAND and is a leader in advancing the technology. The market for their products is expanding at almost geometric speed.

Kioxia leads in technology, in scale, including a joint venture with SanDisk, and in partnerships such as the SanDisk one.

They see their role on a grand scale.

Share Recommendations

QQQ3

Samsung Electronics SMSN

SK hynix 000660

Kioxia. 285A

Strategy – Buy The Winners From Global Rewiring

The Memory business is being transformed by the AI revolution. Many investors worry about a cyclical peak, but the companies, both in America and the Far East, believe they have a sustainable opportunity to advance a key technology driving the global data centre build-out. Their shares have been skyrocketing, but this reflects explosive growth and a new perception of their role in technology.