Weekly Summary
NVIDIA's leadership change unsettled the AI investment landscape, putting the economics of rapid infrastructure expansion back in focus.
Despite pressure on AI-related stocks, major US equity indexes ended the week higher, highlighting a divide between broader market performance and sentiment toward AI.
The mood around AI turned sharply darker this week. Jensen Huang’s departure from NVIDIA shook confidence in the company most closely associated with the AI boom, and the selling spread to businesses that depend on its chips, partnerships and financial support. Yet the broader market tells a more complicated story: the S&P 500 and Nasdaq still finished the week higher despite Friday’s declines. The damage was concentrated in the AI trade, where investors were already questioning how much growth they could safely count on.
NVIDIA announced Huang’s resignation on Thursday and named CFO Colette Kress interim CEO. The company said its operations, customer commitments and product roadmap were expected to continue without interruption. That matters. A leadership change does not make the chips less useful or erase demand for computing. But Huang was more than a chief executive. He helped persuade customers and investors that an enormous expansion in AI infrastructure was both necessary and achievable. His departure leaves the industry with less room to rely on that confidence alone.
The pressure on neoclouds, specialized providers of GPU computing capacity, was especially telling. CoreWeave and Nebius fell sharply again on Friday as concerns about AI financing intensified. NVIDIA has been more than a supplier to parts of this ecosystem; it has also helped finance infrastructure built around its products. Investors now have to consider whether a new leadership team might take a different approach to that support. This is a risk to assess, not evidence that existing commitments have been cancelled.
For these businesses, the distinction between demand and profitable growth is becoming harder to ignore. Building capacity requires money well before customers generate enough revenue to cover the investment. If financing becomes more expensive, or customers take longer to use the capacity they have ordered, returns can deteriorate even while AI adoption continues. The questions are practical: how much funding is already secured, how dependable are the customer contracts, and what cash remains after interest, power costs and equipment replacement? A large order book is encouraging, but it is not the same as cash in the bank.
We do not see this week as a reason to dismiss AI’s long-term potential. We do see a reason to separate progress in the technology from returns on the stocks. More capable models and wider adoption can coexist with overbuilding, tougher competition and disappointing investment returns. For our portfolio, that means looking beyond the size of a share-price decline and asking whether the underlying earnings outlook has changed. Several different AI holdings can still leave us exposed to the same spending and financing cycle.
NVIDIA’s November 17 earnings will be an important first test of confidence under the interim leadership. We will be watching for clarity on demand, the product roadmap and support for the broader ecosystem. Reassuring answers could help stabilize sentiment; uncertainty could prolong the pressure. Our outlook remains constructive but selective. A more durable recovery will need businesses to demonstrate that they can turn the AI opportunity into cash flow, rather than simply ask investors to believe in its scale.