Dear investors,
For the last several months, the same question has kept resurfacing. Are markets overestimating AI? Have valuations become excessive? Is the comparison with the dot-com bubble justified?
The question is a legitimate one. But to me, it reflects a recurring tendency in financial markets. Whenever a major transformation comes into view, attention shifts away from the world it may create and toward the real and imagined ways in which it might disappoint.
Over the course of my investment career, I have seen this pattern unfold time and time again: with the internet, with the rise of emerging markets, and with China's gradual opening to the rest of the world. Each time, the sceptics were ultimately right about one thing: excesses always emerge. But the biggest investment mistakes have rarely been made by believing in these transformations. They have been made by abandoning a ship while it is leaving the harbour.
In January 2024, when AI had not yet become the dominant topic it is today in financial discussions, I was already highlighting the importance it would come to assume. Two and a half years later, many people seem more concerned about the height of the wave than about the strength of the tide that is carrying it.
Yet, there is no denying that the wave is already high. Augmented Intelligence is now estimated to account for 37% of capital spending in the United States and 17% in Europe. Inevitably, excess computing capacity will emerge. This will accelerate the decline in the cost of AI and make it more accessible – a development we should welcome. However, while its diffusion throughout the economy will increase productivity and creativity, it will also bring about profound disruptions that will unsettle many established positions. Will major purchasers of data-processing capacity, companies whose market positions today appear unassailable, such as Microsoft, be able to pass on the cost of these investments? How will established software companies cope with the remarkable ability of new language models such as Claude to analyse and simplify? And finally, how should we assess AI's enormous requirements for memory and electrical power? Every major technological disruption reshuffles the deck, producing both winners and losers. It is up to us to rise to the challenge.
What about the markets? The easing of tensions in the Middle East that we anticipated in my previous letter has indeed materialised, leading to a significant decline in energy prices and alleviating inflationary pressures that had been so widely feared. This period of regaining calm is boosting consumers' purchasing power and should also temper the inclination of the major central banks to continue raising interest rates. Seemingly overlooked by the media, the ongoing shift in the Ukraine conflict to Russia's detriment suggests that a ceasefire may not be so far away, which would significantly improve Europe's economic outlook. We remain concerned, however, about the slow deterioration of the Chinese economy, constrained by Beijing's refusal to stimulate domestic demand, since doing so would, in practice, require meaningful political liberalisation.
The uncertainties have, of course, not disappeared. They never do. But at a time when many are already watching for turning points, it seems to me more useful to focus on the underlying forces at work.
I wish you a peaceful and prosperous summer.
Best regards,