Advertising space
Should investors really get so excited about AI?
Technology investment

Should investors really get so excited about AI?

The launch of ChatGPT has captured the imagination of investors around the world, but they should be careful.
Imagen del autor

1 JUN, 2023

By Simon Edelsten


Any company happening to mention that it has some sort of commercial involvement in artificial intelligence seems to see its share price jump at the moment.

The launch of ChatGPT has captured the imagination of investors around the world, but they should be careful. AI is not new – specialist funds and ETFs in this space go back several years. To make money from AI you have not only to choose the right companies but pay a fair price.

But what might that be? Look at the company’s costs, growth prospects, tax, and, with technology stocks especially, how much management and staff are being paid in shares. The impact of this last number can be hidden but it is the investor who ultimately pays.

You will have to make some assumptions with all your calculations, so a rough rule of thumb may help. Analysts tend to value tech stocks based on how much revenue they are generating in sales revenue. No matter how exciting the growth story might be, when a share trades roughly above five times sales you might want to start being more wary.

Which brings me to Nvidia. AI requires enormous processing power – we are going to be asking our computers to think so much harder. Nvidia’s graphics processing units appear to be among the best chips available to power this function. Business is booming and that means Nvidia’s shares have shot up in price – by nearly 200% already this year.

As I write Nvidia is valued at $1 trillion. Its management estimate that sales will hit close to $40 billion this year. That means the company is priced at 25x sales. For that to be justified sales have to carry on growing sharply – by my calculations, it will take an eightfold hike in sales (assuming margins are not compressed) to justify today’s share price.

Nvidia designs chips; it doesn’t make them. It leaves that job to companies like TSMC – the Taiwan Semiconductor Manufacturing Company – the world’s largest chipmaker. TSMC (which we own) is valued at six times its sales.

Intel is not the Pathfinder it was when nearly every personal computer came with a sticker boasting “Intel inside”. It makes central processing units (CPUs) – not the more powerful units that Nvidia sells. Speed matters, but not all the time. There will be trade-offs. We expect CPU sales to rise. The market values Intel shares at just 2.2 times sales. I bought some at that price. 

And do not forget Microsoft and Google (Alphabet). Their tech wizards have been laboring away in the lab with soldering irons and magnifying glasses designing their own chips. The US defense department is already using Microsoft’s Rapid Assured Microelectronics Processor in the company’s Azure cloud. Google is expected to use its “tensor processing units” in its cloud servers within a couple of years. 

But AI is more than just a chips story. Ultimately, what AI does is process data very quickly. Expect companies that hold data to find more ways to monetise it. The Microsoft (Bing) and Google search engines collect vast repositories of the stuff.

There must be money to be made in AI consultancy too. Accenture (another holding) is the world’s biggest IT consultant.

AI may not exactly transform the fortunes of all of these businesses, but it does not seem unreasonable to expect it to provide a tailwind.

AI will reach far and wide in the coming year. Technology will move on rapidly, which means we need to watch out for disruptors. Nvidia may be leading the way now, but Intel reminds us how that lead can be snatched away. Next-gen graphene chips are on the horizon offering 10 times the speed of today’s chips. Nvidia’s growth shows there is a lot to play for in this market.

Investors should also be wary of the stocks they hold that could be threatened by AI. I believe that AI will destroy as many businesses as it creates. But that’s the topic for another day!

For now, the key point is that it does not matter how excited you are by the prospects for AI and its impact on the companies in your portfolio, in the long term a share price depends on how much profit a company makes. Valuations can drift for a while but not forever. At some point, there is a reckoning. Buy well!

Advertising space