The world of one business – AI

Unlock Editor’s Digest for free

Roula Khalaf, editor-in-chief of the FT, selects her favorite stories in this weekly newsletter.

AI was once the hot new investing theme – a brilliant opportunity for fund managers of all sizes eager to put their capital to work on transformative technology.

That’s still the case, sort of, if we look past the warnings of a robotic apocalypse, the destructive “rogue” hacking agents that increasingly look like a feature rather than a bug, and the legal risks that are beginning to circulate in AI labs. (“Oops, our agents did it again” is no longer an excuse.) Despite all this, no one can reasonably doubt the ecosystem’s ability to make money once in a generation, from chips to data centers, and that’s what professional investors are paid to think about.

But the AI ​​business is no longer new and it is no longer “a” theme. Instead, it has morphed into “the” theme – a vast, suffocating mass that blankets the global financial system, starving everything else of oxygen, distorting market reactions, and leaving every portfolio, everywhere, alarmingly dependent on the fortunes of a single bet. Stocks, real estate, infrastructure, energy, emerging markets, some vintages of private equity. . . they are all, to a large extent, the same profession.

The sheer dominance of this thing over global stocks was clearly demonstrated this week. The price of oil continued to smolder at around $100 a barrel, and traders again sounded their warnings that logistical constraints in the Strait of Hormuz opened up a very real chance of a push to $200. Bond prices were stuck in the deepest rut in decades, with the yield on benchmark 10-year U.S. government bonds – the all-important borrowing cost in global markets – sitting comfortably above 5 percent. Either of these, or certainly both, would generally cause stock markets to decline.

No way. Instead, the S&P 500 index of U.S. stocks hit a new record high, as did the tech-heavy Nasdaq. We are all now accustomed to a little market dissonance, but this is, to say the least, extremely strange.

Cutting up clues in weird and wonderful ways to fit a narrative is cheating. But we still have to do it to understand what’s going on. So yes, the S&P hit a record high. But if we exclude the AI ​​component of the index, the remaining share has fallen 7 percent since the end of August. You get a similar picture if you look at the S&P on an equal-weighted basis, which helps blunt the outsized impact of huge tech stocks. The gap between the heavier typical index and its more impartial cousin is nothing new, but it has now reached its highest point in almost 24 years. Three-quarters of S&P stocks fell last month – technology stocks and energy stocks that reflect data center construction are really doing all the work.

Who cares? Many people don’t do this, as long as the index line goes up. The problem is that professional investors struggle to figure out how to diversify properly, how to protect themselves if something goes wrong – a crucial part of their job of managing other people’s money. Many are spooked by how portfolios turn out to be correlated to a range of supposedly diverse asset classes.

For now, the whole world is all in. Non-U.S. money has been funneled into dollar-denominated assets at an extraordinary rate over the past decade, and increasingly, it is not even hedged against currency risk. In other words, the AI ​​boom has almost accidentally become a major support not only for the US stock market but also for the dollar.

Without a doubt, the profits of American companies, not only in the technology sector but especially in the technology sector, justify this enthusiasm. If anything, U.S. stocks are pretty cheap right now — stock prices haven’t kept pace with spectacularly fast earnings growth, apparently held back in part by rising bond yields.

But we still, collectively, have a very poor idea of ​​how AI is going to play out. To some extent it’s anyone’s guess, but private equity firm Bain said in a recent report that annual spending on AI infrastructure could reach $1.5 billion by 2031. Extraordinary stuff. But what’s even more extraordinary is that he adds that “to maintain this level of investment, we would need an AI market approaching $6 trillion per year.” Really? An industry that represents one and a half times the total annual economic output of the UK each year? This one has difficulty passing the smell test.

Policymakers also seem unsure of what they are dealing with here. The Chairman of the US Federal Reserve, Kevin Warsh, believes that the productivity resulting from the use of AI will reduce inflation relatively quickly. But others at the central bank, most recently Mary Daly and Lisa Cook, are moving to an opposing view, that construction poses an upside risk to inflation in the coming months.

So we end up with a surprisingly two-speed US stock market, beloved by almost every investor on the planet, dominated by AI development that is, in practical terms, almost impossible to hedge or avoid. We don’t really understand the long-term economic implications, but the underlying business assumptions seem pretty heroic. And we have all, deliberately or not, tied our fortunes to the idea that everything would turn out just fine. We better be right.

(email protected)