The City is humming with a familiar sound: the whir of foam being frothed into a latte. But the real froth is in the markets, where artificial intelligence stocks have been trading at multiples that would make a tulip bulb blush. Let's call it what it is: a bubble. And bubbles, dear reader, are not sustained by fiscal responsibility but by the speculative whims of momentum traders and the desperate yield-chasing of pension funds.
The latest Q4 earnings season has exposed a fissure in the AI narrative. Companies like Nvidia and Palantir have delivered stellar numbers, sure. But look closer. Their forward price-to-earnings ratios are drifting into absurdist territory. The market is pricing in perfection, but the economy is delivering sticky inflation and a central bank that is anything but dovish. The Bank of England, to its credit, is still waging war on prices, and the Federal Reserve is showing no signs of unsheathing the rate-cut sword. That means the cost of capital is staying high, and high-multiple stocks are the first to get the oxygen cut off.
I've seen this play before. In the late 1990s, it was dot-com. In 2007, it was subprime. The script is depressingly similar: a new technology that promises to transform everything, a chorus of analysts who have never met a projection they couldn't stretch, and a public that confuses a stock's rise with economic wisdom. Today, the star is generative AI. But the business models are still shaky. The only one making real money from AI chatbots, it seems, is the hardware vendor selling the shovels.
Capital flight is already stirring. Bond yields are wiggling higher as investors demand compensation for inflation risk. The yield on the 10-year Gilt has ticked up, and the dollar is strengthening as risk appetite wobbles. When the risk-free rate becomes more attractive, the equity risk premium shrinks, and suddenly a stock trading at 50 times earnings looks less like a bargain and more like a liability.
Of course, the AI bulls will argue that this time is different. They always do. They will point to the transformative potential, the productivity gains, the inevitable dominance. But I remind you of the law of large numbers. Even if AI changes the world, it cannot change the fact that overpaying for growth is a recipe for disappointment. The market is a voting machine in the short term, but a weighing machine in the long. Right now, the votes are coming in fast and furious, but the weights will eventually pull the prices down.
A correction, in my view, is not just likely; it's necessary. It would clear the froth, reset valuations, and allow real, profitable businesses to emerge from the rubble. The Bank of England would do well to ignore the calls for rate cuts to support stock prices. Let the market correct. Let the weak hands get washed out. The patient capital, the kind that has endured past crashes, will find opportunities in the ashes.
So my advice to the retail punter who thinks AI stocks are a one-way bet: hedge your bets. Take some profits. And if you can't bring yourself to sell, at least buy a put option. The music may stop sooner than you think. And when it does, do not say you were not warned.
