A unique and, frankly, paradoxical situation is emerging in the semiconductor and artificial intelligence market. While some sectors are basking in record cash flows, others—for the first time in history—are showing negative dynamics in this key metric.
Analyzing the latest data, I see a striking contrast. Four leading chip manufacturers—Nvidia (NVDA), Micron (MU), Broadcom (AVGO), and Applied Materials (AMAT)—will collectively generate $430 billion in free cash flow over the next 12 months. This is more than triple the figure from two years ago. In effect, these companies are turning into veritable "cash machines," capitalizing on unprecedented demand for computing power.
At the same time, the picture for the largest AI companies looks diametrically opposite. Five tech giants—Amazon (AMZN), Alphabet (GOOGL), Meta (META, recognized as an extremist organization in Russia), Microsoft (MSFT), and Oracle (ORCL)—will show negative aggregate free cash flow for the first time in history. To grasp the scale: at its peak in 2024, this figure exceeded $260 billion for them. The reason for this reversal is the explosive growth in capital expenditures on AI. The AI-related spending of these five companies will collectively rise to $1.8 trillion in 2026 and 2027.
The Question of Cycle Sustainability
This data inevitably raises the question of the long-term sustainability of the current investment model. Economist Remy Bourgeau rightly notes that sooner or later, the tech giants will have to start making money again, rather than just burning it. The key question: what will happen to chip manufacturers when this colossal cash flow from AI giants dries up?
The Chinese factor also deserves special attention. Chinese giants are essentially continuing to sell their models worldwide for a fraction of American prices, while actively purchasing Chinese chips. In my opinion, this is precisely the essence of the current investment cycle: it cannot be endless.
My analysis: We are witnessing a classic "innovation cycle," where the "shovel and pickaxe" manufacturers (chipmakers) reap superprofits at the initial stage of the boom. However, the current dynamics of AI giants' capital expenditures are clearly unsustainable. As soon as the market realizes that the returns on these trillion-dollar investments do not meet expectations, the correction could be very painful. Investors should prepare for volatility, not linear growth.