In November, the European stock market rose by 1.0%, while the US indices denominated in Euros were down this time: -0.6% for the S&P 500 and -2.2% for the Nasdaq.
While the economy shows signs of weakness on all fronts, whether in China, Europe, or the United States, in both consumption and industry, the market’s attention this month has once again focused on artificial intelligence and technology, the only island of growth in a sea of stagnation. The massive investment plans in data centers continue to fuel the debate between those who claim there is an AI bubble and those who argue that the market is not particularly exuberant given the fantastic prospects this technological revolution offers us.
At Clartan, we have chosen a somewhat uncomfortable middle ground: neither bubble nor complacency. Indeed, a quick comparison with the famous TMT bubble of the 2000s shows us that the current situation has nothing to do with the excesses of that time. We can find convincing points of comparison here and there, between Cisco and Nvidia for example, but the crazy and indiscriminate atmosphere of that era, so characteristic of bubbles, simply does not exist today. We would struggle to find the new France Telecom trading at 220 euros in today’s market (renamed Orange, it is worth 14 euros now).
As an illustration, let’s take the example of Oracle, which, boosted by large orders from Open AI for Cloud hosting, jumped +36% in trading in September. It turns out that the stock has lost all those gains since, as investors realized that a large portion of the orders came only from Open AI, a startup that is not yet self-financing, and that these investments would therefore be made on credit, on Oracle’s balance sheet. It didn’t take long for short sellers to attack the company, which had exposed itself by proposing a business plan consuming several tens of billions in cash over several years. In summary, the market continues to do its job and clearly distinguishes between those who reinvest according to their means, like Microsoft, and those who have eyes bigger than their stomach, like Oracle or, to a lesser extent, Meta.
We could also mention the cases of Google and Nvidia. The market was highly impressed by the release of Google’s latest LLM Gemini, trained with internal chips called “TPU” that are said to be much more efficient for training than Nvidia’s “GPU.” The difference in performance between the two stocks over the past few weeks is striking.
We are therefore clearly witnessing a market that discerns, seeks to separate winners from losers as quickly as possible, and ultimately does its job of evaluating the cost of capital for each company. The general confusion about the possibility of a “bubble” comes from the fact that no one really knows how much overall computing capacity will ultimately be needed to run AI. While waiting for this question to be clarified in the coming years, we believe it is not unreasonable to allocate a portion of our portfolios to this theme, while of course remaining disciplined in the selection of stocks.
The Clartan funds had flat performances in November. Valeurs rose by +0.7%, Europe by +0.1%, and Ethos by +1.1%. Patrimoine also increased by +0.1%, Flexible by +0.3%, and Multimanagers fell by -0.8%.