Nvidia, the company at the center of the global artificial intelligence investment boom, is facing a new and potentially serious challenge.
Reuters reported on November 20 that the US Securities and Exchange Commission is investigating potential accounting issues at the semiconductor giant.
The investigation comes amid growing scrutiny of Nvidia’s sales to customers in Asia, particularly Singapore, and questions about whether advanced American AI processors have ultimately been used by Chinese companies despite US export restrictions.
The US Commerce Department has also introduced new rules intended to restrict Chinese companies’ access to advanced American computing power through data centers located outside China.
These developments coincided with a dramatic collapse in AI-related stock prices on November 20.
Nvidia fell approximately 13.8%. The Nasdaq plunged 5.28%, while the S&P 500 lost 3.70%.
The combination of regulatory investigations, tighter export controls and collapsing equity valuations creates new risks for Nvidia and the entire AI investment ecosystem.
But the latest Federal Reserve data offer an important counterpoint: commercial bank credit continues expanding rapidly, particularly lending to businesses.
The central question is whether the stock market correction will remain confined to financial asset prices or develop into a broader contraction in credit, capital expenditure and economic growth.
Table of Contents
- Nvidia Faces an SEC Accounting Investigation
- The Singapore Revenue Controversy
- US Restrictions on China’s Access to Overseas AI Computing
- Nvidia’s Financial Strength and the Risks to AI
- The AI Selloff Becomes a Market Correction
- Federal Reserve Data Show Continued Bank Credit Growth
- The Magnificent Seven Lose Their Earnings Lead
- S&P 500 Consensus Earnings Growth
- Consumer Inflation Expectations Fall
- China, Technology Leadership and Military AI
- Global Financial Markets and Economic Growth
- Corporate Developments and Investment Risks
- Macro Watch Bottom Line
1. Nvidia Faces an SEC Accounting Investigation
Reuters reported on November 20 that the Securities and Exchange Commission is investigating potential accounting issues at Nvidia.
The report cited two people familiar with the matter.
The Wall Street Journal had previously reported the existence of an SEC investigation, but had not established its focus.
Reuters subsequently reported that SEC enforcement staff were examining possible accounting issues.
The precise scope of the investigation has not been publicly established.
Neither the SEC nor Nvidia provided a substantive public response to Reuters’ requests for comment.
An investigation is not evidence of wrongdoing. No accounting violation has been established by the reporting available at this briefing’s cutoff.
Nevertheless, the development is important because Nvidia has become one of the most financially significant companies in the global economy.
Its processors are essential to the expansion of advanced AI computing.
Its revenues and profits have grown at extraordinary rates.
And its customers have committed enormous amounts of capital to building the infrastructure required to deploy its technology.
Why Accounting Transparency Matters
Investors need to understand not only how much revenue a company reports, but also the financial arrangements through which that revenue is generated.
This is especially important when a company invests in, lends to or otherwise supports customers that purchase its products.
Such arrangements are not necessarily improper.
However, they can create complicated financial relationships.
Investors need to determine whether reported revenue reflects financially sustainable demand and whether customers can meet their obligations without continuous external financing.
The SEC investigation raises the importance of those questions.
2. The Singapore Revenue Controversy
One of the most important issues receiving renewed scrutiny concerns Nvidia’s geographic revenue reporting.
Nvidia previously reported revenue based on the billing locations of its direct customers.
Singapore became an increasingly important location in those disclosures.
For the fiscal quarter ended July 2025, Nvidia attributed approximately $10.2 billion of revenue to Singapore.
That represented roughly 22% of total quarterly revenue.
The Singapore figure had increased substantially from a year earlier.
However, in the following fiscal quarter, Nvidia changed its geographic reporting methodology.
It began classifying revenue according to the headquarters locations of its direct customers rather than their billing addresses.
Singapore consequently disappeared as a separately reported geographic category.
Why the Change Matters
Nvidia explained that the new methodology better represented the geographic profile of its customers.
That explanation is plausible.
Many international companies use Singapore as a regional billing and administrative center even when the products purchased are ultimately delivered elsewhere.
Indeed, Nvidia’s earlier disclosures had indicated that customers frequently used Singapore for invoicing while products were shipped to other destinations.
Nevertheless, the change makes it more difficult for investors to compare historical Singapore-related revenue with subsequent periods.
It also makes it harder to determine how much of Nvidia’s revenue is associated with customers whose billing operations, corporate headquarters and ultimate computing users are located in different countries.
The reporting change does not establish that Nvidia concealed sales or violated export controls. But it has increased the importance of understanding the ultimate destination and use of its products.
How Much Revenue Comes From Asia?
Nvidia’s latest geographic disclosures provide useful information, although they do not identify the ultimate users of every processor.
For the quarter ended October 25, 2026:
| Customer Headquarters Location | Revenue |
|---|---|
| United States | $72.9B |
| Taiwan | $30.3B |
| China, including Hong Kong | $11.1B |
| Other | $1.6B |
| Total | $115.8B |
Approximately 37% of quarterly revenue was attributed to direct customers headquartered outside the United States.
Taiwan and China accounted for most of that amount.
However, these figures must be interpreted carefully.
A sale to a company headquartered in Taiwan does not establish that the processors were used in Taiwan.
Similarly, a sale to a company headquartered in the United States does not establish that the computing capacity was ultimately used by an American company.
Nvidia itself explains that the headquarters locations of direct customers can differ from shipping locations and ultimate end users.
Consequently, the geographic revenue breakdown cannot establish how much Nvidia computing capacity is ultimately used by Chinese companies.
The Megaspeed Investigation
The controversy also involves Singapore-based Megaspeed, which has reportedly been a substantial Nvidia customer.
US authorities have investigated whether Megaspeed was involved in violations of export restrictions concerning advanced AI chips and China.
Megaspeed has maintained that it conducted lawful business.
The allegations remain distinct from the SEC’s reported investigation into Nvidia’s accounting.
Neither investigation establishes wrongdoing by Nvidia.
Nevertheless, the overlapping concerns about export compliance, customer relationships and financial disclosure have intensified scrutiny of the company.
3. US Restrictions on China’s Access to Overseas AI Computing
The US Commerce Department has introduced a major new export-control rule designed to restrict Chinese companies’ access to advanced American computing power through data centers outside China.
The rule was published in the Federal Register on November 19 and became effective that day.
It addresses a potential weakness in earlier export controls.
Previously, restrictions on shipping advanced processors directly to China did not necessarily prevent Chinese companies from renting computing capacity from data centers located in other countries.
A Chinese AI developer could potentially obtain access to American processors without physically importing them into China.
How the New Rule Works
The Commerce Department’s new rule introduces licensing requirements for certain advanced chips and overseas data centers that provide significant computing capacity to covered companies headquartered in China and other designated destinations.
The restrictions also apply when the ultimate parent company is located in a covered jurisdiction.
Existing overseas data centers providing covered services are given a 90-day period to obtain the necessary licenses if they wish to continue.
The rule includes exceptions and specific treatment for certain commercial applications.
It is therefore not a blanket prohibition on every Chinese company’s use of every overseas computing service.
Five Countries Face Additional Controls
The regulation also establishes additional licensing requirements for certain advanced semiconductor shipments to five countries:
- Singapore
- Malaysia
- Thailand
- Indonesia
- Brazil
These countries were identified as presenting elevated risks of diversion.
The rule includes an exception for qualifying shipments to companies headquartered in the United States, subject to reporting requirements.
The Economic Implications
The new restrictions could affect several parts of the AI supply chain.
Nvidia and other American semiconductor suppliers may face additional compliance costs and reduced demand from certain customers.
Data center operators may have to change their customer relationships or obtain licenses.
Chinese AI developers may lose access to some overseas computing capacity.
And companies investing in Southeast Asian data centers may have to reassess the economics of projects designed to serve Chinese customers.
The United States is moving beyond restricting physical chip exports to controlling access to the computing power those chips provide.
That represents an important expansion of the technology competition between the United States and China.
The Strategic Consequences
The restrictions are intended to preserve American technological leadership and limit China’s access to advanced computing capabilities.
But they will also strengthen Beijing’s incentive to develop domestic alternatives.
China will invest more heavily in semiconductor design, manufacturing, advanced packaging and domestic computing infrastructure.
The AI arms race is therefore becoming more expensive for both countries.
4. Nvidia’s Financial Strength and the Risks to AI
The regulatory developments must be considered alongside Nvidia’s extraordinary financial performance.
On November 17, Nvidia reported results for its fiscal third quarter ended October 25.
Revenue reached approximately $115.8 billion, an increase of 103% from a year earlier.
Data center revenue increased 110% to approximately $107.6 billion.
The company’s reported quarterly profits and cash generation were also extraordinary.
| Nvidia Metric | Q3 FY2027 |
|---|---|
| Total revenue | $115.8B |
| Revenue growth YoY | +103% |
| Data center revenue | $107.6B |
| Data center revenue growth YoY | +110% |
| GAAP net income | Approximately $65.3B |
| GAAP net income growth YoY | Approximately +105% |
Nvidia’s financial results demonstrate that demand for advanced AI processors remains enormous.
The company is generating substantial cash flow from that demand.
But the financial position of Nvidia’s customers is a different question.
Some customers are highly profitable hyperscalers.
Others are specialized infrastructure companies that depend heavily on borrowing, equity financing and long-term contracts.
The Risk to Future Revenue
If financing becomes more difficult for Nvidia’s customers, they may reduce or postpone orders.
Tighter export controls could also reduce certain international sales opportunities.
And greater regulatory scrutiny could increase compliance costs or require changes in customer relationships.
These risks do not negate Nvidia’s technological leadership or its extraordinary current profitability.
But they increase uncertainty about the sustainability of future revenue growth.
Nvidia’s financial strength depends partly on the continued ability of its customers to finance extraordinary levels of capital expenditure.
That is why developments affecting the entire AI financing ecosystem matter so much for Nvidia.
5. The AI Selloff Becomes a Market Correction
US stocks plunged on November 20.
The S&P 500 declined 3.70% to 6,910.42.
The Nasdaq Composite fell 5.28% to 22,617.73.
The Dow Jones Industrial Average lost 1.93% to 48,147.33.
| Asset | November 20 Change |
|---|---|
| Nvidia | −13.8% |
| CoreWeave | −12.1% |
| Anthropic | −9.1% |
| Nebius | −16.1% |
| Meta | −5.9% |
| Amazon | −3.9% |
| S&P 500 | −3.70% |
| Nasdaq | −5.28% |
Both the S&P 500 and the Dow entered correction territory, falling more than 10% from their recent record closing highs.
The Nasdaq suffered its worst weekly decline since March 2020.
Why the Decline Is Significant
The selloff occurred despite Nvidia’s extraordinary earnings and Anthropic’s successful IPO.
Investors are increasingly questioning the financial sustainability of the AI investment boom.
The Nvidia regulatory developments added another source of uncertainty.
But the decline was broader than Nvidia.
Specialized infrastructure companies also suffered severe losses.
This suggests that investors are reassessing the financial risks associated with enormous capital expenditure commitments throughout the industry.
The Financial Transmission Mechanism
Falling equity prices make it more expensive for companies to raise new equity capital.
They can also reduce collateral values and increase borrowing costs.
For companies with strong free cash flow and substantial financial reserves, these pressures may be manageable.
For companies dependent on repeated financing, they can become serious constraints.
A sustained decline in AI-related valuations would increase the risk that financially weaker companies are forced to reduce investment.
That would weaken demand throughout the global AI supply chain.
6. Federal Reserve Data Show Continued Bank Credit Growth
The Federal Reserve released its latest H.8 bank-credit data on November 20.
The figures cover the period through November 11.
Total commercial bank credit reached approximately $20.13 trillion, increasing $62.3 billion from the previous week.
Loans and leases reached approximately $14.26 trillion.
The Fed’s October annualized growth figures show particularly strong business lending.
| Credit Category | October Annualized Growth |
|---|---|
| Total bank credit | +6.5% |
| Loans and leases | +8.9% |
| Commercial and industrial loans | +11.7% |
| Real estate loans | +2.6% |
| Residential real estate loans | −0.3% |
These are seasonally adjusted annualized growth rates for October, not year-over-year changes.
What the Data Reveal
Commercial bank credit was still expanding rapidly before the latest stock market collapse.
Commercial and industrial lending was particularly strong.
Residential lending, by contrast, was weak.
This is consistent with an economy experiencing strong business investment alongside a struggling housing sector.
The Creditism Implications
Credit expansion creates purchasing power and supports economic growth.
The continued expansion of bank credit is therefore an important source of support for the economy.
But the latest figures predate the most severe stock market selling.
They cannot establish whether the correction is beginning to restrict new lending.
The critical question is whether falling equity prices and weaker investor confidence will eventually cause credit growth to slow.
A stock market correction can occur without a recession.
A severe contraction in credit would pose a much greater threat to economic activity.
7. The Magnificent Seven Lose Their Earnings Lead
New FactSet research published November 20 reveals an important shift in corporate earnings growth.
The Magnificent Seven reported aggregate third-quarter earnings growth of 27.0%.
The other 493 companies in the S&P 500 reported blended earnings growth of 32.6%.
| Group | Q3 Earnings Growth YoY |
|---|---|
| Magnificent Seven | +27.0% |
| Other 493 S&P 500 companies | +32.6% |
This represents a significant change in earnings leadership.
For several years, the largest technology companies have generated a disproportionate share of S&P 500 earnings growth.
Their strong profitability helped support high valuations.
The latest figures suggest that earnings growth is broadening beyond the technology giants.
Important Accounting Qualifications
The figures are affected by several unusual items.
Alphabet recorded a substantial nonoperating loss on equity securities.
Meta reported a significant litigation-related charge.
Amazon’s prior-year earnings included large investment gains.
Consequently, the comparison should not be interpreted as proof that the underlying operating businesses of the Magnificent Seven have suddenly deteriorated.
Nevertheless, the reported shift in earnings leadership is important.
It weakens one argument used to justify exceptionally high valuations for the largest technology companies.
8. S&P 500 Consensus Earnings Growth
The latest previously verified full-index FactSet consensus estimates were published November 13.
The November 20 FactSet analysis provides new comparisons between the Magnificent Seven and the rest of the index, but a complete updated set of annual consensus figures was not verified for this briefing.
The following estimates are therefore unchanged reference figures.
| Period | Consensus EPS Growth YoY |
|---|---|
| Q3 2026, blended | +31.9% |
| Q4 2026, estimated | +28.3% |
| Full-year 2026 | +33.3% |
| Q1 2027 | +19.6% |
| Q2 2027 | +3.0% |
| Full-year 2027 | +16.2% |
These figures anticipate a substantial slowdown in earnings growth during 2027.
That slowdown matters because interest rates remain elevated.
Even if profits continue growing, slower expected growth and higher discount rates would put pressure on equity valuations.
The next regular FactSet Earnings Insight report is scheduled for December 4.
9. Consumer Inflation Expectations Fall
The University of Michigan released its final November consumer sentiment survey on November 20.
The consumer sentiment index increased from 48.0 in October to 51.5 in November.
More importantly, one-year inflation expectations fell from 4.7% to 3.7%.
Longer-term inflation expectations were approximately 3.3%.
Why This Matters
Inflation expectations influence wage negotiations, business pricing decisions and household spending.
A decline reduces one potential source of persistent inflation.
That is encouraging news for the Federal Reserve.
However, inflation expectations remain above the Fed’s 2% inflation objective.
Consumer sentiment also remains weak.
Federal Reserve Governor Lisa Cook indicated on November 20 that monetary policy was not sufficiently restrictive to address inflation.
Consequently, the improvement in inflation expectations is unlikely, by itself, to eliminate the prospect of further interest-rate increases.
The Risk to AI Financing
Higher interest rates increase borrowing costs and reduce the present value of future earnings.
That is particularly important for AI infrastructure companies making enormous investments today in anticipation of future revenues.
The combination of higher financing costs and declining equity valuations could force weaker companies to reduce investment.
10. China, Technology Leadership and Military AI
The new US export-control rules are part of a broader strategic competition with China.
The United States is attempting to preserve its leadership in advanced semiconductors and computing infrastructure.
China is attempting to reduce its dependence on American technology.
The restrictions on overseas computing capacity demonstrate that the competition now extends beyond the physical location of advanced processors.
It also involves who controls and uses the computing power they provide.
The Economic Consequences
China will have stronger incentives to develop domestic AI processors and computing infrastructure.
That will require substantial investment in semiconductor manufacturing, advanced packaging, networking and data centers.
These investments will support Chinese industrial activity.
But they will also create financial obligations and increase demand for capital.
The competition for AI leadership is therefore becoming a competition in financial capacity as well as technological capability.
Military AI
On November 20, Treasury Secretary Scott Bessent said that President Trump and Chinese President Xi Jinping had agreed on the importance of retaining human involvement in decisions concerning military AI.
That represents a potentially constructive diplomatic development.
However, the reported understanding is not a binding arms-control agreement.
Military AI development will continue, including investment in autonomous systems, intelligence, surveillance, electronic warfare and advanced communications.
The AI arms race and the defense spending boom are becoming increasingly interconnected.
11. Global Financial Markets and Economic Growth
The AI-related stock market correction has spread internationally.
Asian and European equity markets suffered substantial weekly declines.
Bitcoin fell below $60,000, while gold remained under pressure from the stronger dollar and expectations of further US interest-rate increases.
The global consequences matter because AI infrastructure investment supports economic activity in many countries.
Taiwan and South Korea manufacture critical semiconductors and memory.
Japan supplies manufacturing equipment and materials.
European companies provide electrical equipment, industrial machinery and energy infrastructure.
Countries throughout Southeast Asia are attracting investment in data centers.
Financial Contagion Versus Economic Contagion
Financial contagion occurs when falling asset prices spread across markets and countries.
That is already happening.
Economic contagion develops when falling asset prices cause companies and households to reduce spending, investment and borrowing.
That process takes longer.
The central question is whether the financial-market correction will cause companies to postpone or cancel investment projects.
US GDP Outlook
The Atlanta Fed’s latest GDPNow estimate remains 3.2% annualized growth for the fourth quarter.
That estimate was published November 18.
No new GDPNow update was released on November 20.
The next update is scheduled for November 25.
The new developments since the latest estimate are mixed.
Bank credit continues expanding.
Consumer inflation expectations have declined.
But equity prices have fallen sharply, AI financing conditions are becoming less favorable, and manufacturing indicators suggest slower demand.
If AI capital expenditure slows substantially, it would weaken an important source of US and global economic growth.
12. Corporate Developments and Investment Risks
One important corporate development on November 20 involved Intuit.
The company reduced its annual revenue growth forecast and announced an agreement to sell Mailchimp to Bending Spoons for approximately $3.68 billion.
Intuit acquired Mailchimp in 2021 for nearly three times that amount.
Its shares fell approximately 12.8% on November 20.
Comparable revenue and net income growth figures from the latest earnings release were not independently verified for this briefing.
The Lesson for AI Investors
The transaction illustrates the risk of paying exceptionally high prices for expected future growth.
A technology can be commercially successful without producing an attractive financial return for investors who paid too much to acquire it.
That distinction is particularly relevant to the AI investment boom.
Investors must distinguish between the long-term economic value of artificial intelligence and the valuations currently being assigned to companies involved in developing it.
13. Macro Watch Bottom Line
The most important new development is that Nvidia, the financial and technological centerpiece of the AI investment boom, is facing growing regulatory scrutiny at the same time that confidence in AI-related financial assets is deteriorating sharply.
The reported SEC investigation into potential accounting issues adds uncertainty to Nvidia’s financial outlook.
The renewed scrutiny of Singapore-related revenue raises questions about the transparency of geographic sales disclosures and the ultimate use of advanced processors.
And the new US export controls could restrict access to Nvidia’s technology through overseas data centers.
These developments are important individually.
Together, they create additional risks for Nvidia and the broader AI ecosystem.
The Investigation Does Not Establish Wrongdoing
It is essential to distinguish allegations from established facts.
The SEC investigation has been reported by Reuters, but the precise accounting issues have not been publicly identified.
Nvidia’s change in geographic reporting methodology is documented.
But that change does not establish that the company violated accounting rules or export controls.
Similarly, large revenues associated with Asian customers do not prove that processors were ultimately used illegally by Chinese companies.
The central issue is whether investors and regulators have sufficient information to understand the ultimate customers, financial relationships and export-control risks involved.
The New Export Controls Could Have Significant Consequences
The Commerce Department is now targeting overseas computing services as well as physical chip shipments.
That could affect Nvidia’s customers, data center operators and Chinese AI developers.
It could also alter investment plans in Southeast Asia.
The restrictions are intended to preserve American technological leadership.
But they will also accelerate China’s efforts to develop domestic alternatives.
The global AI arms race will become more expensive.
Financial Markets Are Flashing a Warning
The sharp decline in Nvidia and other AI-related shares demonstrates that investor confidence is weakening.
The successful Anthropic IPO and Nvidia’s extraordinary earnings have not prevented the selloff.
Investors are increasingly questioning whether the industry’s enormous capital expenditure commitments will generate adequate returns.
This is particularly important for companies that depend on repeated borrowing and equity issuance.
But Bank Credit Is Still Expanding
The latest Federal Reserve data show that commercial bank credit continued growing rapidly through November 11.
That is an important counterpoint to the stock market correction.
The banking system has not yet shown clear evidence of a broad credit contraction in these figures.
However, the data predate the most severe market decline.
We must therefore watch closely for signs that falling equity valuations and weaker investor confidence are beginning to restrict credit creation.
The Implications for 2027
The AI revolution remains a potentially transformative technological development.
But the current investment boom is also an enormous financial undertaking.
It depends on credit expansion, equity financing, customer demand and sustainable operating cash flow.
Nvidia’s regulatory problems introduce additional uncertainty.
The new export controls could change the geography of AI investment.
And the stock market correction could make financing more difficult for weaker companies.
The decisive question is whether these developments will cause a substantial slowdown in AI capital expenditure.
If they do, the consequences will extend well beyond the technology industry.
Semiconductor manufacturers, electricity utilities, construction companies and equipment suppliers around the world would experience weaker demand.
That would reduce economic growth and increase pressure on corporate earnings and credit markets.
For now, the latest bank-credit figures show that credit expansion is still supporting the economy.
But financial-market confidence is deteriorating rapidly.
The AI investment boom is entering a period in which regulatory compliance, access to capital and sustainable cash flow will become increasingly important determinants of success.
The companies with the strongest balance sheets and most reliable cash flow will be best positioned to withstand the adjustment.
Those dependent on continuous financing will face the greatest risks.
That is why, in the next phase of the AI revolution, cash flow will be king.
