One reason Nvidia fell so much is that it has positioned itself as the mascot, the face of the AI trade. The company is at the epicentre of the AI boom, with the most powerful—and premium—product range, financial deals, investment support and broader commitments involving many companies to keep orders coming.

CEO Huang has been a central figure at Nvidia: besides having a far-reaching vision of what AI will look like—and therefore staying a step ahead in terms of business decisions—he has also been involved in politics and helped shape decisions regarding the White House’s China policy within the broader technology war.

The risk is: if he has to leave, who will replace the man in the black leather jacket, and what will the broader implications be for Nvidia’s sales, earnings and its ecosystem? His departure remains hypothetical, however, and Nvidia’s board has expressed confidence in his leadership while launching an independent investigation.

What does it mean for Nvidia’s stock price?

I don’t have a model to price the Jensen risk or quantify how much his potential departure would affect sales and earnings. But we could argue that yesterday’s selloff was perhaps an overreaction, and the rebound among competitors was certainly due to rising volatility due to rising stress regarding AI trade.

And the Huang distraction doesn’t change the fact that the broader AI trade faces more tangible issues with material implications for company valuations: too much spending, increased leverage, delayed returns on investment and circular deals.

To those, we can now add potential political risks under the incoming Democratic-led Congress and fading enthusiasm for the buzziest AI trades of the past three years. The billion-dollar question is: how hard will it be to revive market appetite sustainably following a spectacular pullback?

Coreweave reports

Today, Anthropic and its valuation is coming back to headlines (latest reports suggested a valuation of $1.1trln versus $2trln previously), and after the bell, CoreWeave—there you go, Nvidia-backed CoreWeave—will report its results. Analysts expect revenue of around $3.54 billion, roughly 160% higher than a year earlier, but another loss, too.

Spectacular growth, in other words, at a substantial cost. And that substantial cost is no longer something investors are willing to shoulder without asking detailed questions. Remember, CoreWeave’s shares and bonds sold off sharply last month amid growing concerns about its debt, while co-founder Brannin McBee spoke of tapping insurers’ 'unlimited' pools of capital to finance further expansion. Unlimited financing? No, sir, said investors, showing that their patience—and their willingness to fund growth at any price—has limits.

So today, investors will watch whether strong demand is translating into revenue as new capacity comes online, whether construction remains on schedule, and how much additional financing the company needs to deliver growth. The size of the order book matters, but so does the cost of turning those orders into operating capacity—and eventually cash flow.

For the broader AI ecosystem, it will take more than just reassuring guidance to revive appetite: AI infrastructure companies cannot grow their businesses without investing massively, requiring substantial financing through debt, equity or customer prepayments. Meanwhile, delays, rising borrowing costs or greater caution around new customer commitments could reinforce concerns that tomorrow’s expansion will be slower or less profitable than expected.

Having exposure to cloud companies is interesting—the business is here to stay—but we may see consolidation in the sector, favouring players with easier access to financing and therefore a greater ability to contain financing costs.

And the financing gap is striking: selected dollar bonds maturing in 2031 recently yielded around 5.1–5.3% for Microsoft, Alphabet and Amazon, compared with around 7% for SpaceX and more than 16% for CoreWeave.

This is a big gap. These are secondary-market yields, rather than the interest rates paid on all their existing debt, but they illustrate how differently investors price their credit risk—and why that matters when companies need to raise more money!

US inflation watch

Speaking of borrowing costs, investors will be closely monitoring today’s US CPI inflation update. The good news is that prices for US goods and services may have risen at a slower monthly pace in October than in the previous two months, as energy prices eased. The bad news is that US diesel prices didn’t reflect that pullback, and underlying inflation pressures remain a concern: both headline and core metrics are expected to have risen on annual basis.

The war in Iran is not over, and oil markets are not ready to return to ‘normal’ just yet. Therefore, a softer CPI reading could ease expectations of a December hike and offer some relief to borrowing costs and valuations. But whether December hike bets will recede enough to reverse the rise in yields and the dollar remains to be seen. Fed funds futures currently assess around an 84% chance of a 25bp hike in December.

Beyond inflation, what is interesting is that the US dollar’s appreciation this year has also been supported by strong growth expectations, partly thanks to US government spending and AI expansion. Relatively contained longer-term inflation expectations suggest that the rise in nominal yields has not simply reflected higher expected inflation.

Therefore, today’s inflation data will be an important piece of the puzzle, but it will not alone determine where the US dollar is headed next. A potential AI slowdown and a more constrained fiscal agenda under the incoming Democratic-led Congress could temper US growth expectations and, over time, inflation pressures.

Softer growth expectations could take some shine off the greenback even if expectations for the Federal Reserve (Fed) remain hawkish. The USD could attract some safe-haven inflows if we see renewed stress in technology stocks and broader risk assets. Yet any USD strength driven by a tech selloff could prove short-lived, as weaker AI-led growth expectations eventually come home to roost.