Jensen Huang has left Nvidia. The company announced his resignation as president, chief executive officer and board member on November 12, with Chief Financial Officer Colette Kress becoming interim CEO.
Much of the attention has focused on the controversy surrounding his departure. At MacroXX, we believe the more useful discussion is about the bigger picture.
Nvidia is not just a company that sells chips. Its technology sits at the heart of a huge wave of spending on artificial intelligence. Its business connects technology companies, investors, lenders, energy suppliers and governments. Nvidia also helps support the financing that allows parts of this expansion to happen.
That makes this more than a story about one company or one executive.
For MacroXX, the central question is simple: will AI generate enough income and real-world benefits to justify the money being spent on it?
The MacroXX approach
At MacroXX, our goal is simple: to make economics and financial markets easier to understand. Too often, market commentary seems written exclusively for professionals, with unfamiliar jargon, complex charts, and analysis that feels far removed from everyday life.
We take a different approach. Through clear explanations and practical examples, we connect economic developments to what matters for investors—helping readers understand what is happening, why it matters, and how it could influence their investment decisions.
1. Spending money on AI is not the same as making money from AI
The AI boom starts with things that companies must buy and build.
They need powerful chips, large buildings to house computers, cooling equipment and reliable electricity. Together, these form the physical backbone of AI.
That spending creates business today. Builders receive contracts. Equipment suppliers receive orders. Chipmakers sell processors.
But there is a second stage: the companies paying for all this must eventually earn enough from it.
At MacroXX, we find it useful to compare this with opening a restaurant. Renovating the building, buying kitchen equipment and hiring staff all create economic activity. But none of that proves the restaurant will attract enough paying customers to succeed.
AI faces a similar test.
Building computing capacity is one thing. Using it to generate lasting income is another. That gap helps explain why the AI spending boom can support the economy while also creating financial risks. IMF chief Kristalina Georgieva has warned about the global economy’s growing dependence on the sector.
For investors, two questions therefore matter:
How much money is being spent on AI?
How much money will businesses earn because of AI?
A strong answer to the first does not automatically provide a strong answer to the second.
2. Following the money: when suppliers help fund their customers
Some of the companies supplying AI technology also invest in the businesses buying it.
This is often called circular financing.
Here is a simplified example:
A chip supplier invests in an AI company.
The AI company uses some of that funding to buy computing services.
The provider of those services buys chips from the original supplier.
Money moves through a network of connected businesses, helping them expand together.
These relationships are not unique to Nvidia. Microsoft has both an investment relationship and a cloud-computing relationship with OpenAI. Amazon and Google have invested in Anthropic, which also uses their computing infrastructure. Nvidia invests in businesses that use its technology.
There can be good reasons for these arrangements. A promising AI company may need expensive computing equipment before it has enough customers to pay for it. Financial support can help it get started.
The question MacroXX focuses on is what happens next.
Does the company eventually earn enough from customers to pay its own bills? Or does it keep needing new investment to continue buying computing services?
Imagine a manufacturer helping a customer finance a large order. That can be sensible if the customer’s business grows. But if the customer survives only because the manufacturer keeps providing support, the sale tells us less about independent demand.
That is the concern—not that every connected investment is suspicious, but that some spending may depend on continued financial support rather than a self-sustaining business.
3. Why a slowdown could spread
These connections can help the AI industry grow quickly. They can also allow problems to spread.
Consider a possible chain of events:
An AI business earns less than expected.
Investors become less willing to provide more money.
The business cuts its spending on computing.
Its computing provider buys fewer chips.
Suppliers and lenders become more cautious.
This is a possible risk, not a prediction.
Companies can be linked in several ways at once: as customers, suppliers, investors and financial partners. If one relationship weakens, others may come under pressure too.
The MacroXX takeaway is to look beyond the sales figures.
“Are chip sales growing?” is an important question. But so is: “Can the businesses behind those sales afford to keep buying?”
4. A $500 billion headline does not mean Nvidia has spent $500 billion
Large announcements can make the story harder to understand.
An investment, a loan and a financial guarantee are different things:
An investment means putting money into a business in exchange for ownership.
A loan means providing money that is supposed to be repaid.
A guarantee means promising to cover certain losses or payments if specified conditions arise.
A fundraising target describes money a project hopes to attract. It is not necessarily money already available.
These differences matter when reading about Nvidia’s $500 billion AI infrastructure financing platform.
According to Reuters, the platform was designed to attract money from outside institutional investors, with Nvidia’s financial guarantees limited to a maximum of 25%. As of November 16, it had not announced a deal. The headline figure should therefore not be read as $500 billion already spent by Nvidia.
A useful comparison is a property development announcing plans to attract $1 billion from investors. That does not mean the developer has already raised—or personally supplied—the full amount.
At MacroXX, we believe the details behind a financial headline matter more than the size of the number.
The practical questions are:
How much money has actually changed hands?
Who is providing it?
What has Nvidia promised?
Who would lose money if a project struggled?
5. China still wants the technology—but access is complicated
Nvidia also sits at the center of competition between the United States and China.
The United States has restricted sales of advanced AI chips to China. These rules limit which products Nvidia can sell and to whom. Reuters reports that restrictions introduced in 2022 were later expanded to include modified chips developed for the Chinese market.
But restricting sales does not remove the desire to buy.
A Reuters investigation published on November 13 identified more than $3.2 billion in Chinese computing projects and equipment agreements built around restricted Nvidia processors. That amount included different stages of development, as well as buildings, power systems and other equipment. It was not $3.2 billion in Nvidia chip sales.
For a business planning an AI facility, this creates a practical problem: what happens if the chips it intends to use become difficult or impossible to obtain?
It may have to delay construction, change suppliers or redesign the project. Those changes can increase costs and make the original investment less attractive.
For MacroXX readers, the investment lesson is straightforward: strong demand does not guarantee a completed sale. A customer may want the product and have the money to buy it, but government policy can still stand in the way.
6. AI can be useful without every AI investment being successful
It is tempting to divide the debate into two camps: either AI changes the world, or the whole boom is a mistake.
At MacroXX, we do not believe investors need to choose between those two extremes.
AI could become widely used while some companies make disappointing profits. Customers might benefit from cheaper services, while suppliers face tougher competition. Businesses might save time, while investors discover that they paid too much for the expected growth.
A useful technology is not automatically a good investment at any price.
AI businesses could attract enough paying customers to justify the spending. Alternatively, AI use could keep growing while high costs and competition hold back profits. Funding could also slow before businesses earn enough to support themselves, forcing them to postpone projects and reduce spending.
These are possibilities, not forecasts.
The important distinction is between a technology being useful, a company making money and an investor earning a good return. Those three things are related, but they are not interchangeable.
7. What Huang’s departure changes—and what it does not
A leadership change matters when a company plays such a large role in its industry.
Huang helped shape not only Nvidia’s products, but also its partnerships and financing strategy. Reuters reports that Nvidia’s financing platform is expected to continue under Kress, although the transition could bring a pause and a more cautious approach to risk.
That is worth watching. But it does not replace the bigger economic questions.
The MacroXX framework is to focus on five areas:
Customer income: are AI businesses earning enough to pay their computing bills?
Spending discipline: are new projects being built because customers need them, or mainly because financing is available?
Financial commitments: how much risk are suppliers taking to support their customers?
Government restrictions: can companies obtain and sell the technology their plans depend on?
Market expectations: are business results keeping up with the growth investors already expect?
These questions are less dramatic than the latest controversy. They are also more useful for understanding the investment case.
The MacroXX perspective
Nvidia’s leadership has changed. The basic economic test has not.
The AI boom is building enormous computing capacity. Its lasting value will depend on what businesses and customers can do with that capacity—and whether they will pay enough to support it.
At MacroXX, that is the bigger picture we want to keep in focus: not simply whether AI will matter, but whether the benefits will justify the bill.
This newsletter is for informational and educational purposes only—not financial advice. Use of this information is at your own risk, and the authors are not responsible for any resulting losses or damages.



Agree, the CEO change is noise next to whether AI actually pays off. Still adding to $NVDA slowly on dips like this