Mega-cap leaders have overcapitalised in winner-takes-all race for AI dominance, warns Tom Aylott, bringing their future relevance into question
Some of the world’s leading technology companies have seen their share prices soar over the past year thanks to excitement surrounding artificial intelligence (AI), but the hype they’ve drummed up could instead become their fatal flaw. Orbis Global Balanced manager Alec Cutler said mega-cap tech stocks are facing an “existential event” after having overpromised and overspent on AI in a winner takes all race for supremacy in the space.
The likes of Alphabet, Amazon, Meta, Microsoft and Nvidia have all sworn to be the leaders of the AI revolution, but they can’t all take the crown. If they fail to live up to these hefty assurances – which Cutler expects most of them to – then they will be left with a lot of underwhelmed shareholders.
“They sold investors last year on the fact that they’re going to cut costs and rationalise it and make efficiency gains, but this is an existential event for them,” Cutler said. “They have to win in AI. This is a winner take most space so they’re all going for it, but only around two of them is really going to make it. Yet they’re all spending as if they’re going to be one of those two.”
The necessity for these companies to become leaders in what could be the most profitable technology of the next decade is essential to their survival. Cutler noted how these multi-billion-dollar companies have swelled to such enormous sizes that they need to find new growth drivers to ensure they continue expanding. There is a risk now that these companies not only plateau without AI dominance but decline in relevance.
But their pledges to remain at the forefront on innovation seem to have convinced markets. Cutler said: “We did our work and looked into it, and what we figured out is these companies are completely wrong. There’s a gold rush element to it, and there’s also an existential risk element to it. Right now, the market is seeing this as a positive for the primary players such as Microsoft, Google, Apple, Meta and Amazon because it’s being framed as a new revenue generator for them. But these are companies that need to find something really big in order to grow and sustain their billion-dollar market caps.”
And after investing record-high levels of capital expenditure on AI, these companies will also be at a substantial monetary loss if it doesn’t pay off. Meta, for example, plans to spend up to $40bn this year alone in building up its AI infrastructure, and Microsoft has already splashed out over a quarter of that ($14bn) in just the first three months of 2024.
Cutler said: “I don’t know how it’s going to make everyone so more productive that it’s going to be worth the capital they’ve invested and produce a return on the billions of dollars they’ve spent buying chips that will be obsolete in two years.”
Yet so much of what it takes to stay ahead in the race for AI depends on who can spend the most money, according to Richard de Lisle, manager of the VT De Lisle America fund. He said it is a matter of scale and those who can afford to keep throwing money at AI are the only ones able to stay in the race. We have already witnessed some early AI leaders fall back due to having less spending power than their peers, such as Tesla. After soaring 101.7% in the AI rally of 2023, its share price took a turn this year (dropping 28.3% year-to-date) as investors questioned its ability to keep up with other tech behemoths.
“Elon Musk came out and said Tesla is an AI company now. Well, the problem with that is in order to be one, he would have to spend billions to be as strong as the likes of Meta and he doesn’t have that capital. Mark Zuckerberg does, but Tesla can’t do that,” de Lisle said. “I don’t think Tesla will be a good performing stock going forward just because of the capital requirements. Regardless of what you think about Musk as a person or electric cars, Tesla is undercapitalised and in order to play the game, everyone’s suddenly got to spend more money.”
Upping capital expenditure to all-time highs has been enough to maintain their seats at the table, but it doesn’t ensure they’ll be a leader in future. Many tech giants are more vulnerable to being toppled and replaced by new entrants now than ever before, and their high volumes of spending are an attempt to simply keep pace, according to de Lisle.
“They’re all terrified,” he explained. “These companies wouldn’t be spending all these billions otherwise. It’s because they’re scared. There’s a group of stocks that will win, but it’s not necessarily the group of stocks that we’ve got at the moment. Some will lose and fall off, and then there’s another group that will come in. People aren’t really sure what those are yet – they are looking for them now.”
Investors are keen to hold on to companies that could be the next tech champions, but “people lose sight of how quickly the leading group changes”, according to de Lisle. And as AI develops and businesses fine tune their applications, the rate at which these market leaders switch places is accelerating. Just a couple of years ago, markets were focused on the FAANGs, which were quickly replaced by the magnificent seven, and even those are now shifting. Trying to keep up with the rapid changes is becoming an ever more challenging task for active managers.
“The whole AI revolution is going to increase the speed of change of the top companies,” de Lisle said. “It argues for indexation in a way because you might say, well I can’t keep up with this, so I better buy a tracker for the S&P 500 and that will sort it out for me. Because if a company is getting bigger and more important, it will get a bigger weighting in the index. You don’t have to work out whether Tesla is going to fall or not. It’s an automatic thing and in that way, you get covered from the increasingly rapid change in technology.”
Amid all this uncertainty at the top of the market, there are ways to benefit further down the supply chain. De Lisle said “the only thing that’s pretty clear is that the leading stock in a year’s time will be Nvidia,” so working out what companies it relies on and that it will pull up with it is key. By that logic, he bought shares in Supermicro, which climbed a whopping 922.9% since the start of last year.
Yet there is one thing that all these tech companies – regardless of who wins the AI race – need a lot more of to support their operations, and it is something both de Lisle and Cutler have been allocating to. All the infrastructure needed to support AI at scale requires enormous amounts of energy. “That’s really what it comes down to,” de Lisle said. “It’s all very wonderful, but they need an awful lot of power. That is what’s at the end of the food chain, really.”
Cutler said the composition of the leading tech companies struggling for dominance in AI could swing quickly and often, but businesses addressing this universal need for power will thrive regardless of who is leading in the race. He added: “They’re all spending like crazy, so when you have all these combatants trying to be a winner, do you really want to invest? Do you want to bet on which will be the two winners that are able to monetize to the degree that it’s going to pay off? Or do you bet on the ammo providers?
“These companies will be paying a lot for electricity and will need to build an entirely new grid surrounding their facilities, so it’s been a wonderful shot in the arm for the critical energy infrastructure theme. Not only will they need to rewire a very densely wired server, but you now have to rebuild the entire grid around it to supply the amount of wattage that’s now needed by that facility.”