In the wake of predictions that Microsoft and OpenAI will spend $100bn or more on a new AI super data centre, Ranmore Fund Management, the global equity manager led by Sean Peche, and MAPFRE AM, the investment arm of Spain’s largest insurance company, take a look at the valuations involved, and note opportunities that may benefit investors looking to the AI theme, but at better prices.

Sean Peche, Portfolio Manager at Ranmore Fund Management, said: “At Ranmore, we think Value investing is the best way to ‘play AI’

"The common thinking is the Mark Twain argument, that – “in a gold rush it’s a good time to be in the picks & shovels business” And since Nvidia’s chips and Microsoft’s OpenAI are deemed the AI ‘picks and shovels’, that’s how investors are getting exposure.

"BUT we think this only holds true, if:
a) you’re unsure who’d find ‘gold’
b) you pay a fair price for a ‘pick and shoveller’

"If you knew who'd find gold, wouldn’t you rather own THAT business?"

"So, who is finding ‘AI gold’?", Peche continued.

"Well, according to a Salesforce Survey in 2023, “82% of business leaders say generative AI will lower overall business cost and 80% say generative AI will increase revenue” – ie, most companies.

"And if AI can help cut costs, then the real beneficiaries could be companies with low profit margins

"Think about it - if operating margins are only 2%, then costs are 98% of sales

"That’s a lot for AI to work with – cutting costs by 1% of sales could increase earnings by 50% in this example

Except most Quality / Growth portfolios don’t own businesses with 2% operating margins...

"Take the food retailers for example: Those with loyalty card programs have vast amounts of data -that’s ‘AI gold’. They know:
- what we buy
- when we buy
- where we buy
- how much we pay.

"They know how our buying patterns change over the month, our response to promotions, and even the weather at purchase. AI is helping them drive revenue growth with personalised promotions and dynamic pricing, which improves their demand forecasts lowering wastage & logistics costs. And better inventory control improves cash flow.

Peche further said: "Now consider banks that many Quality & Growth investors also shun, in contrast to Amazon which only knows our discretionary spend on Amazon.

"But banks know our income AND every $ of our spend, on EVERY ‘platform’ - Real ‘AI gold’. So, banks are using AI to:
- more accurately price additional credit - increases revenue
- improve customer service via Chatbots & improve realtime fraud detection - lower costs.

"Maybe the market is waking up to this because Kroger’s share price is up 25% this year, beating Amazon +19%.

"The MSCI World Banks Index is up 13% this year, beating Microsoft +12%.

"Meanwhile, Apple and Tesla are down 11% and 30% respectively…

"Remember too that Nvidia’s largest customers are the large tech companies. So, Nvidia’s revenue is THEIR capital expenditure...

"Last quarter MSFT’s capex rose 58% y/y, Alphabet’s rose 45%.

"So, if you think Nvidia’s revenue will still explode, you better start worrying about their customers’ ‘free’ cash flow.”

César Gimeno, fund manager at MAPFRE AM, also commented on this AI theme.

“The recent stock market boom in the Artificial Intelligence industry has seen a vast number of companies, both large and small, as well as retail and institutional investors, trying to surf the wave of this new trend and generate economic returns. The fact that the Nasdaq 100 has surged by over 48% in the past year is nothing to sneeze at.

"However, in many cases, the use of artificial intelligence is not necessarily the only or the best alternative that companies have for developing their business.

"To better understand how to harness this technological development, we need to delve deeper into this industry’s value chain. As in most economic sectors, we can divide this industry into two parts: producers and consumers.

"There are two major areas when it comes to the producers. Firstly, there’s the hardware enablers: here the clear example is Nvidia, whose GPUs (graphics processing units) are the fundamental physical element used to build large-scale deep learning algorithms. Then, we have solution and software enablers, such as large language models (Microsoft’s ChatGPT or Google’s Gemini).

"However, it’s important to bear in mind that these are the most representative examples, where investors have already positioned themselves, recording significant gains over the past year (Nvidia +235% and Microsoft +64%, compared to +27% on the S&P500). This, in turn, has led to very demanding valuations, with Nvidia’s price-earnings ratio (PER) at 66x and Microsoft’s at 37x, compared to 21x for the S&P.

Gimeno continued: "That’s why it’s worth looking at other areas/values that are just as significant. One example would be the production of GPUs, where Nvidia is merely a designer and needs to leverage other companies to be able to manufacture these devices. It’s a sector where TSMC is the undisputed leader, and whose valuation is significantly more attractive than that of the former, with a PER of around 20x.

"In turn, producers are highly reliant on key equipment to manufacture these chips. This segment has well-known examples in Europe, such as ASML and its photolithography equipment, as well as some less familiar ones like its counterpart ASM, whose silicon wafer processing solutions will become increasingly important.

"From the consumer’s perspective, it is vitally important to bear in mind how AI solutions will be used in their value proposition. To this end, we can divide companies into two major groups: those with a clear strategy regarding the inclusion of artificial intelligence solutions in their value proposition and those without one.

"It’s evident that the focus in this case should be on the latter. However, this is only a necessary condition, not a sufficient one. A second critical factor that we must consider is profitability— not only in absolute terms, but also in relative terms. History has shown us how the implementation of new technological developments can cannibalize more profitable pre-existing businesses, as was the case with analogue and digital photography.

"Artificial intelligence will be transformational, and we can already identify some of the winners of the future. However, there are still less obvious opportunities to benefit from this unstoppable trend.”