I completed my first ever mountain climb last week - a tiddler, admittedly, at c. 450m / 1500 ft, but still. The way up was tough - but the only times I nearly slipped and fell were on the way down, which felt like it took forever. Are AI-related #stocks also at a peak - and about to embark on an uncomfortable descent? A couple of recent examples give pause for thought.

A company called Broadcom is one of the big winners of the AI boom. It makes a particular type of computer chip, essential for the networks in data centres. This year it passed $1 trillion in market capitalisation. The company has been growing at incredible rates. 

Last month, it released its latest quarterly results. Revenues from its AI division were up 143% year-on-year to $11 billion. It forecast the next quarter’s sales to be up an astonishing 200% year-on-year to over $16 billion. It reiterated its target for AI revenues in 2027 to hit $100 billion.

Unbelievable set of results. A company making billions yet growing at rates as if it were a tiny startup.

The share price reaction? It fell 15%.

That’s right. Down 15%. The value of the company went down by $300 billion – the third-largest fall, in terms of value, by any company in a single day.

What could possibly explain this reaction? According to market commentators, the share price fell because, even though it met or exceeded all analyst forecasts and its own projections, investors were disappointed that Broadcom did not announce even greater increases in revenue and even higher projections for next year.

This sounds like it doesn’t make sense (and it really doesn’t) but what we have here is a situation where Broadcom’s shares were priced for *more than* perfection. Anything less than that and the market reaction was savage.

The tale of Broadcom has been repeated this week: a South Korean chipmaker, SK Hynix, announced even more outrageous results last night: profits up *557 per cent* to $42bn - in just the last 3 months. The share price fell 18% in response.

It’s not that AI is not a transformative technology with far-reaching consequences. The question is whether the value ascribed by the market to the architects, builders and suppliers of that technology is reasonable or not.

Will the companies spending hundreds of billions building the models one day generate enough from users to pay for all their investment?

Or, in the case of Broadcom and SK Hynix, will those AI model builders have the money to keep buying more and more of their chips?

Negative share price reactions to mind-blowing financial results looks worryingly like a trend. Investors who have gone all in on #AI should beware: it's a long way down from the top.

Monmouth Capital #investment #markets