When South Korean chip giant SK Hynix launches its planned $29 billion listing in the US, investors will tell themselves they are buying one of the world’s most important semiconductor companies.
But what they will really be buying is a rosy artificial intelligence forecast. Specifically, they will buy into the view that AI will become one of the biggest investment and infrastructure booms in modern economic history, and that we are still in the early days.
This is what makes SK Hynix’s successful deal so important. It’s not just another big list. It’s a window into how Wall Street now thinks about AI.
For most of the market’s modern history, investors evaluated companies based on a set of relatively straightforward questions. How fast are revenues growing? What are the limits? How much money can the business generate? Is management doing its job?
Of course, these questions have not disappeared. But in the AI market, they have become secondary. The big question has become much bigger: How big will the AI economy ultimately be?
Look what happened to SK Hynix. On paper, it is a manufacturer of memory chips. Historically, this meant a cyclical, capital-intensive business prone to booms and busts. Investors are concerned about prices, inventories and supply gluts.
Today, investors see something completely different. They see a company standing at one of the most critical bottlenecks in AI.
Without advanced memory chips, there is no AI generating boom. There are no boundary models of AI. There are no high-scale AI data centers. There is no next generation AI infrastructure.
This underlying reality has transformed SK Hynix from a semiconductor stock into something much more powerful: a leveraged proxy for global AI spending. The same phenomenon is happening in the markets.
Nvidia is no longer valued primarily as a chip maker. The services that AI data centers serve are no longer primarily valued as services. Data center operators are no longer valued as real estate businesses. Electrical appliance manufacturers, refrigeration companies and even nuclear power developers have, to varying degrees, traded in AI.
Entire market sectors are being repriced based on a single macroeconomic assumption: that AI will require an unprecedented infrastructure build-out.
Maybe the investors are right. The world’s largest technology companies are expected to spend hundreds of billions of dollars this year alone on AI-related capital expenditures. Each new generation of models requires more computing power, more electricity, more network capacity and more memory than the last.
If AI develops as its proponents expect, today’s spending boom may seem modest in hindsight. But, as always, there is another way to interpret what is happening.
What if SK Hynix is no longer a company, at least in the way investors traditionally understand companies? What if a macro trade is made? This doesn’t have to sound radical.
Investors already treat banks as leveraged bets on interest rates, energy companies as proxies for oil prices and gold miners as leveraged positions on silver bullion.
Increasingly now, they are treating AI infrastructure companies as backed bets on the premise that demand for AI will be far greater than almost anyone currently predicts.
This creates tremendous opportunities and tremendous risks. The history of financial markets is full of examples where investors correctly identified transformative technologies while misjudging the surrounding economy.
Railroads transformed economies. The Internet transformed society. But both also generated periods of spectacular overinvestment because investors assumed revolutionary technologies guaranteed extraordinary returns.
They don’t, but what they do guarantee is tremendous capital expenditure.
Tech firms are spending more because they expect demand to explode. Suppliers like SK Hynix are expanding to meet customer demands for more capacity. Investors prize that expansion because it confirms the narrative that demand will remain strong.
It’s a self-reinforcing cycle. None of this means we are in a bubble inflated by AI. AI may ultimately justify every dollar currently invested, and perhaps much more.
But SK Hynix’s successful listing reveals an important shift in market psychology. Investors are no longer asking whether AI companies can build profitable businesses. They are asking whether AI itself will become the defining economic project of the 21st century.
This is no longer a company-specific investment thesis. It’s a macroeconomic bet, and one that Wall Street is making with increasing conviction.
Nigel Green is the founder and CEO of the deVere Group.
