Hot chips

Microchip stocks are soaring. Fuelled by the growth of established players, like Intel and Arm, and relative newbies, such as Nvidia, the sector continues to stoke global equity markets. 

Cynics point to the chip boom having all the hallmarks of a bubble. But it’s possible, given the enthusiasm for all things AI, that this upcurve still has a long way left to run. 

Expectations defied

Since it floated on the Nasdaq in New York last autumn, Arm’s share price is up about 190%. After posting positive results last week, it’s doubled. 

Arm’s performance has surprised those who doubted its pre-IPO target price. We wrote at the time about how James Anderson, one of the UK’s best-known tech investors, warned it was “not clear that Arm is a critical player” in growth markets such as AI. He wasn’t alone, with others claiming how it still looked a little too much like an old fashioned, British hardware business.

Anderson has a brilliant record, but he’s being proved wrong here, with AI’s near-hyperbolic investment narrative the driver for Arm’s rise. Also, interestingly, by listing on the Nasdaq, Arm offered itself up to the mercy of the world’s largest market of retail investors, most of whom love to jump on the next hot stock (or asset or coin or… whatever). It’s worth considering how this added to its inflows – but it does leave the stock vulnerable to a ‘meme-stock’ shift in sentiment. 

Designing shovels

Despite the ups and downs of being a darling of the retail investment crowd, with the AI market hotter than ever, Arm has secured its spot at the heart of the AI landscape because of its moat. Arm is interesting, and pretty unique, because it doesn’t make microchips – rather, it designs chips and sells the licenses to leading AI companies. 

This week, Arm CEO, Rene Haas, emphasised how his company is benefiting from the “profound opportunity” of enormous demand for AI applications. All the major large language models (Nvidia’s Grace Hopper, Microsoft’s Cobalt, Amazon’s Graviton) are based on Arm’s V9 chip design. This business line is driving Arm’s revenue, and proving to doubters that the company is more than just an old-fashioned tech design company.

As per the old investment adage, in a gold rush, sell shovels and picks. Arm’s services are required to compete in the AI arms race. Gold is pouring their way as a result. 

Nvidia evolving

Another key global chip player, Nvidia, marches inexorably upwards, too. Its share price has multiplied 5x in 12 months and is up c50% year to date. Many believe, despite its gains, Nvidia is reasonably priced as its fundamentals remain solid and the demand for its chips looks only set to grow as AI adoption proliferates and the tech evolves.

Crucially for Nvidia, when ChatGPT sent the world AI-mad around this time last year, it was in the right place at the right time, with its H100 chip the best for training large language models. In the 12 months since, the company has developed and invested in its technology, and rolled out a broad range of new chips and hardware, whilst growing its software and services business. 

Nvidia remains ambitious, seeking to go beyond software and hardware and become a broad tech provider, the “Apple of AI”, perhaps. Investors like this reluctance to rest on its laurels. 

A booming bubble 

Unless you’re super deep in the tech stock weeds (I’m not!), and know something lurking that I don’t, it’s hard to see why, despite their prices, these stocks don’t still have big upside potential. With rumours that Altman’s OpenAI wants to raise $7trn (that’s a ‘tr’), cash will continue to pour into AI. This will filter through to the earnings of companies like Arm and Nvidia. 

Yes, the charts look bubble-y. But bubbles continue to inflate if well-run, fundamentally-solid businesses stay on course and demand remains. As an amateur observer, it seems to me that demand will only grow as we further explore AI’s transformative effects. Thus, it doesn’t seem sensible to bet against the companies providing the hardware, software and services powering the globe’s AI capabilities. In markets, bubbles only go bad if they burst. 

What’s more, at some point, and in some way, these capabilities will grow even greater as AGI nears. When this occurs, businesses underlying the tech will benefit again. As ever, those investing in the design and manufacture of picks and shovels will profit.