- In Huang, we trust !
- In Europe, growth keeps up despite challenges
US EQUITIES
In Huang, we trust !
- Nvidia released its third-quarter financial report on Wednesday evening, ending October 2025. As the largest company by market capitalization in the S&P 500, valued at $4.5 trillion, its results and outlook were eagerly awaited by investors. This publication is once again extraordinary, but the stock has decreased.
- Revenue reached $57 billion, up +62% year-over-year and 22% quarter-over-quarter. Net income stood at $31.8 billion in Q3-25, up 59% year-over-year, with a net margin of 56%. For the fiscal year ending January 2026, revenue is expected to reach $213 billion, with net income of $114 billion. Before the emergence of ChatGPT in November 2022, Nvidia reported $27 billion in revenue and $8.4 billion in net income. In just three years, Nvidia has multiplied its revenue by 7.9x and its net income by 13.6x. Over the same period, the stock price surged from $14 at the end of 2022 to today’s level, a 12.9x increase.
How did this company become such a giant?
- In 2006, Nvidia introduced CUDA, a computing architecture and coding interface enabling parallel computing for scientific calculations. Rather than processing instructions sequentially, CUDA allowed developers to accelerate computing time by executing instructions in parallel across Nvidia’s chips.
- In 2012, AlexNet, an AI program designed by Krizhevsky and Sutskever using CUDA and running on standard graphics cards, won the ImageNet image recognition competition. This marked the resurgence of convolutional neural networks, first introduced by Yann LeCun in 1989. Nvidia quickly ramped up investments to seize this new opportunity. By 2013, it generated $200 million in data center revenue out of a total $4 billion, largely from gaming graphics. On the eve of the ChatGPT boom, the data center segment was already thriving, with $10.6 billion in revenue, representing half of total sales in 2021.
- Nvidia’s success stems from several factors. First, neural networks and deep learning models gained traction in scientific communities starting in 2012. Second, Nvidia’s unmatched ability to orchestrate parallel computing through its chips and software proved essential for AI. Meanwhile, Moore’s Law, predicting processor capacity doubling every 18 months at constant cost, began to slow in the mid-2010s. Accelerated computing became indispensable. By late 2022, AI had its “iPhone moment”, visibly entering everyday life.
How can we continue to create stock market value?
- In the short and medium term, the scale of Nvidia’s largest customers, bordering on hubris for some, remains a growth driver. The company captures most of the investment in hardware to train large language models (LLMs) and handle chat requests from OpenAI, Alphabet, Anthropic, Microsoft, xAI, and others. Nvidia addresses this demand with systems currently connecting tens of thousands of chips, soon to scale to hundreds of thousands or even millions. For years computing ingenuity has delivered 5x to 10x performance gains annually. Looking ahead, new foundational models beyond LLMs will emerge with significant commercial potential. Yann LeCun, a French AI pioneer and former Chief AI Scientist at META, resigned to launch his own startup developing JEPA, a model for understanding the physical world. Nvidia also continues to expand in robotics, particularly autonomous vehicle chips, which currently generate less than $2 billion in revenue.
MACROECONOMY
In Europe, growth keeps up despite challenges
- Halfway through November, the European Commission published its updated economic forecasts for the Euro Area and the European Union (details in the table below). (real) GDP growth has been revised upwards for 2025 to +1.3%(compared to +0.9% initially forecast in May) and is expected to rise to +1.4% in 2027.
- At the same time, inflation remains unchanged at +2.1% in 2025, while the terms used to describe its future trend now indicate that it will hover around 2% over the forecast horizon (compared to an initial forecast in May of +1.7% in 2026).
- Unsurprisingly, the strong stockpiling trend followed by US companies in anticipation of tariffs has supported European exports and explains a large part of the upward revision to the forecast.
- Looking ahead, the latest high-frequency indicators (surveys, etc.) remain favorable despite a turbulent global environment (global trade set to slow):
o Household consumption should rebound, supported by a resilient labor market (ageing population dragging the unemployment rate down), the recovery in households’ purchasing power (wage increases are now slowing but still outpace inflation) and the gradual decline in the high savings rate;
o The investment momentum should continue and intensify. The effects of the implementation of the German recovery plan (EUR 500 bn fund for infrastructures and climate, exemption from the debt brake for defense spending) will be compounded by the use of the remaining funds from the Recovery and Resilience Facility (post-COVID NextGenerationEU plan) and the gradual increase in the defense budgets of various countries.
Summary of the forecasts for the Euro Area
