
Given the whirlwind November to date encompassing the Blue Wave mid-terms datacentre backlash, Jensen Huang’s resignation as CEO of Nvidia and new US artificial intelligence (AI) chip export controls, Nvidia’s first earnings report of the post-Jensen era was highly anticipated. Given ongoing internal and external investigations the company made clear they would not be answering questions on Huang, but both the new interim CEO and CFO, Colette Kress, as well as the new Chairman, Harvey Jones, made an impassioned defence of the company, its incredible human talent, strong innovative roadmap, and AI ecosystem support while also reiterating Jensen would not be reinstated.
An outsized earnings beat & raise accompanied by increasing their recently issued next year revenue guidance to 75% growth was as much as investors could ask for, even if the debate is more around 2028 AI capex. To sum it up, Kress signed off the earnings call saying, “We’ve got this”.
This reminder of Nvidia’s extraordinary financial prowess highlights the even more depressed valuation, which already embeds significant concern around future AI capex and a post-Jensen discount. While the company will need to appoint a new permanent CEO, conclude investigations, and rebuild trust with employees, investors and regulators there is no doubting that Huang leaves an incredible legacy and Nvidia will not be the first or last major technology company to transition away from its founder.
Highlights
- Is Jensen coming back? With a multitude of voices across the spectrum calling for his return, in his first public statement yesterday since his resignation Huang offered to come back. But the new Chairman, Harvey Jones, reiterated the board did not intend to reinstate him and they continue to assess external and internal candidates, two of whom were on the call in interim CEO and CFO Colette Kress, as well as Ian Buck who leads their datacentre business. With the ongoing investigation the company said they would not answer questions on this topic.
- Next year guidance raised but debate on 2028 ongoing: With the recent CEO drama, mid-term datacentre backlash, financing concerns, and new US AI chip export restrictions on datacentres in South-East Asia and Brazil, the key debate was whether the company would reiterate last quarter’s strong next year revenue guidance of 70% growth. In fact, Kress raised it to 75% growth and off a higher base given strong near-term results and guidance. That 75% growth contemplates the new export rules as well as datacentre readiness with emphasis placed on demand being well ahead of their ability to supply. The market debate is now more on calendar year 2028 given most datacentre builds for next year are already baked in and that concern will not be answered anytime soon but Kress pointed to customers just this week asking for more supply and unfulfilled demand next year spilling over into 2028.
- Will Nvidia continue to support the AI ecosystem? Given concerns on AI financing notably for neoclouds and Nvidia’s importance as a financial backstop to the ecosystem, there were question marks as to whether post Huang Nvidia would be as supportive. Kress was nuanced in her response, on the one hand reiterating financial support but emphasising they will be selective. She also committed to returning all excess cashflow to shareholders beyond strategic uses, including the current US$227bn of buyback authority. It was also made clear in their enhanced disclosures that financial guarantees to customers do not kick in until datacentre projects are financed, constructed and operational. It was noted that the major project, PORTS-Pike in Ohio for OpenAI set for phased starting in 2029 with a maximum long-term commitment of $105bn. Yesterday, Goldman and Apollo launched a US$42bn financing vehicle for SpaceX backed by Nvidia GPUs, with the associated asset backed debt remaining attractive to investors given, as Nvidia highlighted on the call, continued compute shortages. Hopper GPUs from 2022 are now being recontracted on multi-year agreements at up to 2x their original price while Blackwell rental rates are up over 60% since launch.
- Supply chain commitments continue to grow: While memory stocks continue to selloff and sit at valuations that imply a significant correction from 2028 onwards, Nvidia’s supply commitments swelled to US$390bn, up US$111bn quarter on quarter. This was driven mainly by new memory long term agreements highlighting the disconnect between Nvidia’s demand outlook and current investor scepticism.
- China update: A topical subject given recent newsflow, but Nvidia’s datacentre sales last quarter included only 2% to China and their guidance contemplates this figure at zero. Nvidia also announced the board was setting up a National Security and Export Oversight Committee to enhance engagement with the US government, including all relevant agencies and to demonstrate Nvidia is serious about ensuring strict adherence to US export controls.
Asset-backed debt: Debt secured by specific assets that can be used to repay investors if the borrower fails to meet its obligations.
Buyback: A repurchase of outstanding shares by a company to reduce the number of shares on the market and increase the value of remaining shares.
Capital expenditure (capex): Money invested to acquire or upgrade fixed assets such as buildings, machinery, equipment or vehicles in order to maintain or improve operations and foster future growth.
Cash flow: The net balance of cash that moves in and out of a company. Positive cash flow means more money is moving in than out, while negative cash flow means more money is moving out than in.
Compute: The processing power required to run computational workloads, including artificial intelligence models and applications.
Data centre: A facility housing computing infrastructure, such as servers, data storage and networking equipment, used to process, manage and store data.
Export controls: Government restrictions on the export of certain goods, technologies or services to particular countries or entities.
GPU (graphics processing unit): A specialised processor originally developed for graphics that is also widely used for the intensive computing required by artificial intelligence applications.
Neocloud: A specialist cloud computing provider focused primarily on supplying high-performance computing infrastructure, particularly GPUs, for artificial intelligence workloads.
Supply commitments: Contractual agreements or commitments to purchase components or manufacturing capacity from suppliers over a specified period.
Valuation: The process of determining the value of an asset or company.
All opinions and estimates in this information are subject to change without notice and are the views of the author at the time of publication. Janus Henderson is not under any obligation to update this information to the extent that it is or becomes out of date or incorrect. The information herein shall not in any way constitute advice or an invitation to invest. It is solely for information purposes and subject to change without notice. This information does not purport to be a comprehensive statement or description of any markets or securities referred to within. Any references to individual securities do not constitute a securities recommendation. Past performance is not indicative of future performance. The value of an investment and the income from it can fall as well as rise and you may not get back the amount originally invested.
Whilst Janus Henderson believe that the information is correct at the date of publication, no warranty or representation is given to this effect and no responsibility can be accepted by Janus Henderson to any end users for any action taken on the basis of this information.