AI has driven investors to hallucinations

WHAT will happen first? Will artificial intelligence (AI) kill your investments or your job? The competing stories about how the next big technological revolution will play out are putting markets in a spin.

Last year, the key debate in financial circles around AI centred on whether it was a bubble, whether the technology itself was genuinely useful and monetisable enough to justify the spectacular valuations of companies at its core.

It’s hard to overstate how obsessed investors were on this issue. At one point last summer, chipmaker Nvidia’s results day was considered on a par with the monthly US jobs report – the biggest beast of the global economic data calendar – for its role in setting market direction and maintaining the upward momentum in stocks.

Quarter after quarter, Nvidia’s earnings knocked it out of the park and its share price fizzed higher.

Now we are in a new phase. Nvidia put out earnings numbers last week, and again they beat expectations, but the market largely shrugged, suggesting we have finally passed peak excitement over this one enormous company. And we face a different question – not whether the technology is good enough, but whether it’s too good, to the point where maybe lots of jobs and companies will soon just not need to exist.

Even Nvidia’s bumper earnings in the final quarter of last year – and its expectation for some US$78 billion in earnings in the current quarter – could not liven up US stocks from a very wobbly patch, with the valuation of some companies hit hard by the fear that AI could seriously compromise their business models.

Software stocks have taken a particularly harsh beating, and fears are stirring from an unusual source in the form of a blog post from a relatively obscure outfit called Citrini Research. This rumbled stocks when it sketched out an imagined vision of the future where AI has led to mass unemployment and monstered the markets.

Citrini could be right. It could be wrong. Many in the finance industry think it’s the latter. “Successive waves of technological change have not produced runaway exponential growth, nor have they rendered labour obsolete,” noted Citadel Securities’ Frank Flight in a rather pointed blog that mirrored and picked apart the Citrini argument.

But that’s not really the point. The issue is that investors have found themselves sitting in a stock market that has crammed money into a rather narrow, very US-focused AI trade, and they are now having doubts over whether they have rewarded the right companies, in the right country, doing the right stuff, and what the ramifications of the technology really are.

SEE ALSO

Nvidia plans new chip to speed AI processing, WSJ reports

US stocks: Wall Street notches monthly declines on combined AI, tariff, geopolitical uncertainty

“A market that moves 3 per cent on a blog is a market that does not know,” said Helen Jewell, international chief investment officer for fundamental equities at BlackRock. In response, she noted, investors are having a rethink. “How do I make sure my portfolio is not too exposed to any factor, geography or risk?” she added. The only answer, boring as it sounds, is to diversify.

That line of thinking is one of the reasons why we are seeing the benchmark S&P 500 index of US stocks, heavily weighted towards AI heavy hitters with large market capitalisations, stuck in a rut.

In contrast, a version of that same index where stocks are equally weighted regardless of size is still pushing higher. On top of all the political stress in the US, it is also one of the reasons why European and Asian stocks are performing so much better than their US peers this year.

Market volatility of the type we have seen in the past week is unusual by recent standards. We have become accustomed in the past few months to US stocks that drift smoothly higher. But now we are going to have to learn to live with turbulence while investors settle on a new vision for how this all plays out.

Morgan Stanley pointed out in a note last week that shake-outs are common in periods of disruption.

“What’s going on now is typical of a major investment cycle,” the bank’s analysts wrote. “Volatility bands tend to widen and there are intermittent periods along the way where markets question both the pace of capital spending and which areas of the market could be disrupted.” The launch of the iPhone in 2007, for example, shook up sectors from gaming to cameras.

One thing that is clear is that AI is here to stay. The bank’s analysis of more than 10,000 earnings and conference transcripts shows that nearly a third of companies adopting AI have seen at least one quantifiable benefit, up from 24 per cent in the third quarter of last year. (Naturally, it used AI to do this analysis, rather proving its own point.)

The attention economy rewards people with loud voices making bold and confident pronouncements, especially pronouncements laced with doom.

Last week proved that money managers have taken note of this, too. The wisest path, though, is to screen out overconfidence and listen to analysts and investors who admit that when it comes to guessing the future, they are hedging and spreading their bets, because they just do not know. FINANCIAL TIMES