Bubble, crash, collapse: recent market headlines have read like a doomsday thesaurus for artificial intelligence.
The recent technology stock sell-off generated much excitable commentary. Fuelling the apparent gloom were warnings from OpenAI’s Sam Altman, who admitted AI investors may be “overexcited”, and an MIT study suggesting 95 per cent of corporate AI pilots have delivered no measurable return.
A recent downbeat note from Apollo economist Torsten Slok, arguing today’s AI boom is frothier than the dotcom bubble, also did the rounds.
However, the bursting-bubble narrative feels overcooked. Firstly, stocks can’t rise every day, and this looked more like routine profit-taking than panic.
READ MORE
Secondly, valuations are not uniform. Yes, some names – Tesla, Palantir – look incredibly expensive. Others don’t.
Google, for example, trades on a cheaper multiple than the S&P 500.
Thirdly, elevated trailing price-earnings (PE) ratios – the favoured choice of tech stock sceptics – don’t tell us much. Even forward PE ratios can be misleading, says Ritholtz Wealth Management’s Matt Cerminaro.
Adjust for earnings growth using PEG ratios and the so-called magnificent seven (excluding Tesla) start to look cheap, he says, second only to energy in relative value.
[ Why Greeks are in pole position when it comes to artificial intelligenceOpens in new window ]
Of course, one may question whether current tech earnings are sustainable. Still, Cerminaro’s conclusion – “tech’s not as expensive as you think” – seems reasonable.
Expensive on the surface, perhaps, but hardly the stuff of a bubble bursting.
