Nvidia stock falls after board opens CEO probe

Nvidia fell 4.27% to $178.67 on November 9 while the Nasdaq Composite rose 2.74%. The divergence was even sharper within semiconductors. AMD and Broadcom both rallied as the Philadelphia Semiconductor Index climbed around 6.9%. Nvidia had joined Friday’s rebound. By Monday’s close, it had given back the entire gain.

Between those sessions, Nvidia’s board announced an independent investigation following reports that Chief Executive Officer Jensen Huang had been targeted by a foreign intelligence effort. The timing and the strength of rival chip stocks make the inquiry the clearest new reason for Nvidia’s underperformance. The harder question is whether customers or regulators will change any decisions that matter to the business. The distinction will matter when Nvidia reports earnings on November 17.

How far Nvidia fell while chip stocks rallied

The two sessions before Monday help isolate what changed. On November 5, Nvidia fell 9.3% as artificial intelligence (AI) stocks led a broad selloff. Investors were already questioning whether the pace of data center construction would hold after the U.S. midterm elections. Nvidia gained 2.8% the next day, when technology shares began to recover. On Monday, that recovery continued across much of the market, helped by lower oil prices. Nvidia went the other way.

How Nvidia traded after the weekend

NVDAUSDT perpetual from Toobit, as of November 10, 2026, around 06:12 UTC.

Nvidia’s perpetual traded at 180.31 USDT, up 0.82% over 24 hours. It was back near its 24-hour high of 181.89 USDT ahead of the next U.S. stock session.

• Opening gap. The stock opened at $173.07, around 7.3% below Friday’s $186.64 close.

• Intraday recovery. Nvidia dipped to $170.24, then finished more than $8 above that low. Even the session high of $179.92 stayed below Friday’s close.

• Trading volume. Around 460 million shares changed hands, roughly two and a half times Friday’s volume.

The opening gap put much of the repricing before regular trading began. Buyers came in at lower prices during the session. The volume gives a sense of the activity without identifying whether the shares came from funds cutting exposure, short-term traders, or other sellers.

There was room for an unusually wide split. The Russell 2000, which tracks smaller U.S. companies, outpaced both large-cap benchmarks. Chip shares rallied even more. A market-cap-weighted index can absorb a decline in one large constituent when the rest of the market rises by enough. For anyone holding a broad-market fund, a semiconductor fund, and Nvidia directly, Monday exposed a different kind of concentration risk. The same stock can show up in all three holdings, and the gains elsewhere may have less influence on that portfolio than on the index.

A comparison with January 27, 2025, puts the divergence in context. When DeepSeek’s lower-cost AI model raised doubts about how much computing power developers would need, Nvidia fell around 17%, the Nasdaq dropped 3.1%, and the Philadelphia Semiconductor Index lost 9.2%. The selling spread because the question applied to equipment suppliers across the sector. On November 9, the chip index moved sharply higher. That makes a fresh industrywide collapse in expected chip demand a weaker explanation for Nvidia’s decline.

Why AMD and Broadcom could rally on the same day

AMD had news of its own on November 9. Its EPYC processors are used in Google Cloud’s TPU 8tx system. A host processor prepares data and coordinates the work, while the accelerator carries out the specialized computing. This arrangement gives AMD a place in the system even when Google provides the accelerator.

A cloud operator can buy AMD host processors, deploy accelerators it designed with a partner, and use Nvidia systems for other workloads. The three stocks therefore respond to different parts of the same spending cycle. Monday’s rally in competing chip stocks strengthens the case for a company-specific driver at Nvidia. Evidence of customers switching orders would be needed to show that one company’s gains came at another’s expense.

What Nvidia’s board is investigating

The board announced a review, Huang responded to employees, and the White House addressed the matter the next day. Each statement establishes something different.

• November 8, Nvidia’s board. A committee of independent directors would oversee the investigation with outside independent counsel. The board expressed confidence in Huang and his team. It said it had no reason at that point to believe national security or company secrets had been compromised.

• November 8, Huang. He apologized in a message to employees and committed to cooperating with the investigation.

• November 9, the White House. President Donald Trump said the Federal Bureau of Investigation was looking into the matter. He left the scope unclear.

The board’s assessment was preliminary. For the business, the next consequential disclosure would be a finding that changes how Nvidia handles sensitive projects, customer relationships, or approvals.

The distinction between ownership and operating authority is useful here. Nvidia’s proxy statement put Huang’s beneficial ownership at 3.58% of common stock as of March 23. His importance to customers comes chiefly from the job he holds. He represents Nvidia in major cloud partnerships and in conversations with governments that influence where advanced chips can be sold. Even if day-to-day shipments continue, a change in who handles sensitive approvals or customer relationships could create work for partners with deployment schedules already in motion.

The strongest case for limited disruption is that a large cloud installation takes years to plan. A customer may already have picked equipment, secured power, leased a site, and adapted software to a particular system. Replacing a supplier midstream can carry its own costs and delays. Nvidia’s board statement disclosed no change to customer agreements. If the review does affect the business, a deferred installation, revised shipment window, or tighter approval requirement would be a more useful sign than another volatile trading session.

A past export decision shows what a measurable business effect would look like. In April 2025, the U.S. government told Nvidia that it would need a license to export its H20 products to China. Nvidia subsequently recorded a $4.5 billion charge tied to excess inventory and purchase obligations. A specific licensing change affected a specific product and showed up in the accounts. The board inquiry would become a comparable operating issue only if it led to a concrete change in what Nvidia could ship, whom it could supply, or which costs it had to absorb.

Why cash flow is growing more slowly than sales

Nvidia reported $96.2 billion in fiscal second-quarter revenue, with a 75.0% gross margin. Its Data Center business accounted for more than nine dollars of every ten in revenue. The mix within that business shows how much depends on the buyers building the largest computing systems.

Nvidia revenue by market platform in millions of U.S. dollars from Nvidia’s Form 10-Q, as of July 26, 2026.

Data Center revenue rose 117% year over year to $89.0 billion. The $48.7 billion hyperscale business was its largest component. AI cloud providers, industrial customers, and enterprises made up the other $40.3 billion. In the same quarter, one direct buyer accounted for 16% of companywide revenue. That figure captures exposure to one purchasing relationship, which can serve several end users.

How sales and cash flow grew over six months

Operating cash flow rose sharply in dollar terms, but it fell relative to revenue. The increase in accounts receivable absorbed $24.6 billion of cash in the latest six-month period. Nvidia attributed part of that increase to longer payment terms on large agreements spanning several quarters. It said terms can run from 90 days up to a year for some investment-grade customers building large data centers.

When Nvidia has to commit to supply

Nvidia’s supply and capacity commitments rose from $119 billion to $279 billion in one quarter, primarily to secure memory and manufacturing capacity for its Data Center systems.

Nvidia commitments by fiscal year in billions of U.S. dollars from Nvidia’s second-quarter CFO commentary, as of July 26, 2026.

Around $92 billion of the $279 billion in supply and capacity commitments was scheduled for the remainder of fiscal 2027. Nvidia can adjust or reschedule some arrangements before firm orders, sometimes at a cost. Reserving supplies early helps keep new systems on schedule. The trade-off is that suppliers and customers can be working on very different timetables.

• A site opens late. A grid connection or other site delay can push equipment deliveries and associated revenue into a later quarter.

• A customer pays later. The system can count as revenue after delivery, while payment is still months away.

• An order gets cut. Nvidia may need to find another buyer for the equipment or redirect the capacity it reserved. That is a more lasting problem than a shipment moving by a few weeks.

Nvidia is also trying to widen access to AI infrastructure financing. Its August proposal with asset managers and banks targeted more than $500 billion of third-party capital over time, subject to definitive agreements. A project that secures financing is better placed to proceed with construction and equipment purchases. The announcement gives investors a financing route to watch. Completed funding arrangements and equipment deliveries will show how much business comes through it.

Where cloud spending reaches Nvidia and where it goes elsewhere

Alphabet spent $51.4 billion in capital expenditure during its third quarter. The largest part went to technical infrastructure.

Alphabet’s quarterly capital spending more than doubled from $24.0 billion a year earlier. The spending mix helps explain how much of that growth could reach a chip supplier.

• Servers and networking. Around 65% of technical infrastructure investment went here. The category includes equipment for both Nvidia systems and Google’s custom chips.

• Data centers. The other 35% of technical infrastructure investment went to facilities. A larger budget for buildings and equipment can support future installations while adding little to current chip sales.

Google Cloud’s contracted backlog reached $561 billion, and Alphabet expects to recognize just over 55% of it as revenue within 24 months. Much of the backlog relates to regular cloud contracts. It gives Alphabet a reason to bring computing capacity online, but it cannot tell Nvidia investors how many accelerators Google will buy. The platform can serve some of that demand from existing capacity, use chips it designed, or rent computing capacity from a third party.

Google also said third-party capacity it had arranged to bridge its computing needs came online only at the end of the quarter. That is a reminder of how far apart a spending decision and an operational data center can be. With power and grid connections among the constraints, an operator can have customers waiting and money committed while usable capacity is still months away.

How Nvidia can sell into a system with custom chips

On November 8, Nvidia, Rebellions, and SK Telecom announced a plan to connect the Korean chip designer’s processors using Nvidia’s NVLink Fusion technology. The proposed system splits the workload across two chip designs.

• Nvidia processors handle the initial work of processing the prompt.

• Rebellions chips generate the response. The company expects its systems to be ready for deployment in 2028.

If the design reaches deployment, a customer could use a locally designed chip alongside Nvidia networking, rack architecture, and processors. That gives Nvidia a route into some systems even when customers want more control over which chips run their models. How much Nvidia can earn depends on which components make it into the final installation.

What to watch in Nvidia’s November 17 earnings

Nvidia is scheduled to discuss results on November 17 at 5 p.m. Eastern Time. The fiscal quarter ended October 25, two weeks before the board announced its review. A strong result could show that shipments were on schedule through October. Management’s outlook for the next quarter will be more useful for assessing any changes since the announcement.

Nvidia’s August guidance for the October quarter

• Revenue. $108.0 billion, plus or minus 2%.

• Gross margin. 74.0%, within half a percentage point.

• China. The outlook excluded Data Center compute revenue from China.

Margins deserve a separate look. Nvidia’s guidance already implied that sales could rise around 12% from the July quarter while gross margin slipped by a percentage point. At $108 billion of revenue, a one-point change is worth around $1.1 billion in gross profit before operating expenses. An earnings beat driven mainly by higher shipments would tell a different story from one accompanied by higher margins and faster cash collections. The first shows scale. The second gives a better sense of how much Nvidia is earning and collecting as it scales.

• Shipments and deployment dates. Orders scheduled for later delivery can still support growth, but moving several major installations together would make near-term revenue more dependent on a handful of sites.

• Payment terms. If receivables keep growing faster than sales, look for which customers are getting more time to pay and how Nvidia is funding the gap between delivery and collection.

• The board’s next decision. Changes to who approves sensitive projects or handles key partnerships would help investors identify which parts of the business could be affected and on what timetable.

Monday showed that investors can still buy into the chip buildout while marking down Nvidia shares. If the inquiry stays contained and customers keep their rollout schedules, the selloff looks more like an immediate response to uncertainty. If customers defer orders or Nvidia has to change how it manages sensitive work, the effect becomes easier to trace through shipments, margins, and cash. Those are the developments that would turn one bad session into a different outlook for the business.

How to trade Nvidia stock on Toobit

To trade Nvidia on Toobit, open the TradFi tab under perpetual futures and select NVDAUSDT.

NVDAUSDT is a Tether-settled perpetual contract that gives you exposure to Nvidia’s share price. Ownership, voting rights, and dividends belong to shareholders.

Before placing an order, review the mark and index prices, funding rate, fees, margin requirements, leverage, and liquidation price. Check the current terms and availability for your account and region.

Trade Nvidia stock on Toobit now.

Risk warning

Stock futures involve leverage and liquidation risk. Prices can move quickly, and you may lose all the funds committed to a position. Review the current product terms and risks before deciding whether to trade.