Nvidia Without Jensen
Updated 13 Nov 2026•6 min read
November 2026
Nvidia is down 12% since the start of the week (its sixth consecutive weekly decline and its longest losing streak since 2012), while the semiconductor index (SOXX) is up 13%. The gap was driven by certain allegations on Jensen Huang (CEO of Nvidia), which culminated in his resignation on 12th November.
So what exactly happened?
On 29th October, The New York Times reported that Nvidia CEO Jensen Huang had been in a personal relationship with Alice Huang, an executive at Singapore-based data-centre firm Megaspeed whom US intelligence reportedly believes to be a Chinese spy. Consequently, Nvidia’s board opened an investigation into the matter. Further reporting then revealed a second relationship, with a senior Nvidia executive, which raised conflict-of-interest concerns within the company itself. Huang subsequently resigned as CEO, president and director. CFO Colette Kress was named interim CEO, and former Synopsys CEO Harvey Jones became the board’s first independent chair.
So where does the company stand today?
We believe Nvidia is not the same without Jensen Huang. As co-founder, he led the company for c.33 years and ran an unusually flat structure, with c.60 direct reports. No clear successor has been identified.
With this in mind, we set out the following bull and bear cases for Nvidia from here.
How this could benefit Nvidia
Balance sheet de-risking. Nvidia has backstopped financing for several neoclouds, which leaves it with contingent exposure to some of the most leveraged parts of the AI build-out. If new management steps back from these arrangements, the balance sheet becomes cleaner and less exposed to a downturn in the neocloud segment.
Capital allocation reset. Under Huang, Nvidia deployed c.$40–60bn of free cash flow each quarter into the AI ecosystem through investments and financing. A new leadership team may slow this and return more capital to shareholders via buybacks, which we would be directly accretive to shareholders.
Valuation. Nvidia trades at roughly 11–12x NTM earnings (its lowest multiple since 2012) while consensus expects revenue to grow c.70% next year. That leaves the stock cheap against its own history and implies the market is pricing in substantial execution and leadership risk.
Potential Drawbacks for Nvidia
Supplier economics at risk. A meaningful part of Nvidia’s margin strength reportedly came from favourable terms with key suppliers such as TSMC and SK Hynix. Huang’s standing in Taiwan, where he is widely regarded as a national figure, helped secure these terms. Without him, Nvidia may lose some of that leverage, putting pressure on component costs and gross margins.
Risk on relationship capital. Huang had a lot of influence in the build of the AI ecosystem, with close ties to frontier labs, hyperscalers, customers and suppliers. We think these relationships helped Nvidia secure a disproportionate share of both customer demand and constrained supply capacity. That relationship capital is hard to transfer, and competitors may use the transition to win share and allocation.
Loss of the ecosystem orchestrator. Huang was known for his vision, commercial instinct and speed of decision-making, and he personally orchestrated many of Nvidia’s ecosystem deals and financing arrangements. That gave Nvidia an unusual degree of influence over the wider AI value chain.
Regulatory and China exposure. The DOJ has reportedly opened an antitrust investigation into Nvidia, alongside separate probes into Megaspeed linked to the alleged diversion of chips to China. Analysts estimate c.10% of Nvidia’s revenue comes from China, and likely more including indirect and grey-market channels. Heightened scrutiny could weigh on this revenue, and any fines or penalties would add further pressure on earnings.
CUDA moat under pressure. The leadership transition also comes at a time when AI coding agents have began to erode Nvidia’s CUDA software moat, which has historically locked customers in through high switching costs. In its IPO roadshow materials, Anthropic said its Claude Mythos model took around three days to get a model running on AMD’s MI355X GPUs, versus the 6–12 months this typically takes. If porting workloads off Nvidia becomes this fast, customers could treat GPUs as more interchangeable, which weakens Nvidia’s pricing power.
So who stands to gain?
Chip rivals. Custom silicon and rival GPU makers stand to (1) gain share, (2) foundry capacity, and (3) talent if Nvidia’s grip loosens. However, with the supply chain fully booked, any shift is likely at least six months out.
Memory manufacturers. A weaker Nvidia has less leverage to negotiate component terms, which shifts pricing power on high-bandwidth memory towards suppliers. More accelerator vendors also mean more buyers competing for the same constrained supply.
Scale-up optical networking providers. Nvidia currently controls parts of scale-up networking through its proprietary interconnects. If custom chips take share, an open scale-up market emerges that isn’t yet in street estimates for 2027 or 2028.
Closed model players. Nvidia was among the loudest backers of the US open-source push, and that advocacy now weakens alongside a tougher regulatory backdrop. Slower encroachment from Chinese open-source models could benefit closed-model leaders such as OpenAI and Anthropic.
Who loses?
Neoclouds. Their borrowing rests partly on Nvidia’s guarantees and promises to rent back unused capacity. A successor who favours buybacks over pre-buys and backstops would mean tighter funding and weaker demand, though some vendor financing comes from elsewhere.
The wider AI ecosystem. Nvidia recycles much of its c.$40–60bn of quarterly free cash flow into neoclouds, labs and suppliers. Even a pause could pull $50bn or more out of the system just as rates rise and parts of credit become less liquid.
