This year there are three types of US stock market investors, and they are all at the Nvidia party.
The first group are inside the increasingly crowded room, wondering if they should risk downing one more cocktail or leave now to avoid an almighty hangover tomorrow.
A second, smaller but louder, group are on the dance floor, whooping it up and proclaiming that whatever the price of the drinks, they are not leaving anytime soon.
Finally, the largest crowd are those standing outside on the pavement, getting a bit cold waiting in the red-velvet rope lined queue, hoping they might get in when others leave.
The AI phenomenon of 2023 has so far concentrated most of the returns on just one group of shareholders in a way the dotcom mania never did. Nvidia is up well over 200% while companies such as Microsoft, seen by many as another obvious winner thanks to their early investment in ChatGPT is also up a healthy 39%, but it seems pale in comparison.
Still, there are surely many more beneficiaries to be uncovered as the adoption of AI starts to be felt in all sorts of industries. Here are a few examples that might allow buyers to participate and still sleep at night too:
In mega cap land, Meta Platforms (aka Facebook to most of us) is the champion of a free and open-source version of AI, winning backers who don’t want to see power concentrated in a few hands, or at least those who don’t want to have to pay for it either. Meta stand to harvest all the data that their AI produces, whilst Amazon plans a lite version of AI, allowing users to apply the technology piecemeal, rather than all encompassing.
If Nvidia’s forecasts are correct, then this means that an entire refresh of personal and corporate hardware must be due as well as that purchased by data centres. So, what ‘fish’ can we catch to go with those ‘chips’? Dell Computers blew past forecasts last week and gained 20% after predicting a new wave in PC demand, as your next laptop will contain some version of AI.
This rising tide would lift other long-established companies, forgotten by the FAANG loving crowd for many years now, such as IBM, Intel and another personal favourite, Seagate Technologies – maker of that dinosaur product, the hard disk drive. AI will produce levels of data that has to be stored somewhere and the largest storage depot in the world (Amazon’s S3) is housed on these very things. These companies trade on multiples of earnings that are less than the equivalent in revenue for the likes of Nvidia.
A large language model (LLM) such as ChatGPT costs a bomb to run, mainly in electricity costs and gives off enough heat to power about 10 million homes. Data centres running this level of computer power are now outstripping the ability of air cooling to manage this heat and thus liquid cooling systems are going to be required potentially triggering substantial growth over the next few years.
Delighted to have a chance to wave the British flag for once, I can tell you that a small Sheffield-based company called Iceotope looks to be an early leader in this field – sadly, it is a private company, or I would have bought some shares by now.
Julian Wheeler is a partner and US equities specialist at Shard Capital. The views expressed above should not be taken as investment advice.