Two major capital machines just collided, and the last time investors watched a new technology compete this aggressively for funding, many people owned the right idea but still lost money.
You may think America’s $40 trillion debt pile and the AI spending boom belong in separate conversations. One sits in Washington and the bond market. The other sits in Silicon Valley and your portfolio.
Put them together, though, and the picture changes.
The U.S. government has crossed $40 trillion in total debt, while NVIDIA (NASDAQ: NVDA) and major financial institutions are building financing platforms designed to mobilize more than $500 billion for AI infrastructure.
That gives us two enormous capital demands hitting the same global market. Washington needs investors to keep lending it money. AI needs investors to finance chips, data centers, power and computing infrastructure. Both want more capital while U.S. 30-year Treasury yields recently hit their highest level since 2007, with borrowing costs also jumping in major markets including Japan and Germany.
So if you own expensive technology stocks, AI names, REITs or highly indebted companies, this stops being someone else’s bond-market problem. Your portfolio is already sitting near the blast zone.
The Bond Market Can Slash Your Stock’s Price
The first blow does not require an earnings miss, a recession or bad news from the company you own. Look at the 10-year and 30-year Treasury charts. The 10-year yield climbed from roughly 1.5% in 2021 to above 4% by 2024, while the 30-year moved from around 2% to well above 4% over the same period.
That changes the comparison every investor makes. Why should I pay 40X earnings for a stock when the government suddenly pays me much more to lend it money?
This is where high-multiple technology stocks can get hit even while their businesses continue growing. The company may beat earnings and raise revenue, yet investors can still decide that 30 times earnings makes more sense than 40.
The most exposed names sit where investors have pushed the payoff far into the future, Microsoft (NASDAQ: MSFT), Alphabet (NASDAQ: GOOGL), Amazon (NASDAQ: AMZN), Meta Platforms (NASDAQ: META), and the smaller AI names in the AI infrastructure
The higher the price of money, the less investors may pay today for money they expect years from now.
30 year treasury yield
10 year treasury yield
Then Higher Cost of Money Reaches Earnings
The second blow can arrive later, which makes it easier to miss while the market focuses on falling stock prices. Watch highly leveraged REITs, utilities carrying enormous capital programs, cash-burning technology companies, capital-intensive manufacturers and businesses with major refinancing needs.
A company that locked in cheap debt years ago can sit comfortably for some time. Another company that must refinance billions next year walks straight into the new rate environment.
Higher government yields also work through the economy. Mortgage rates rise. Auto loans become more expensive. Corporate borrowing costs increase. Consumers eventually have less room to spend. And then we get to AI.
AI Is Joining the Same Fight for Capital
The AI trade started as a revenue story, but the next stage increasingly looks like a financing story. NVIDIA recently partnered with major financial institutions to create financing platforms aimed at mobilizing more than $500 billion for AI infrastructure.
Since the early boom rewarded companies for selling chips. The next phase requires somebody to fund the data centers, power infrastructure and computing capacity that keep those chips running…which has already become a point of tension.
Reuters reported that NVIDIA paused parts of a financing initiative that offered credit support to AI cloud companies in exchange for a share of their revenue after concerns emerged around the structure. Investors now need to watch who pays for the buildout, who carries the financing risk and whether the return justifies the capital going in.
The Telecom Boom Shows How This Story Can Go Wrong
The late-1990s telecom boom offers the part of this story that AI investors should find most uncomfortable. You see, the internet really did change the world. Fiber-optic infrastructure became essential. Yet companies still spent enormous sums building capacity faster than the economics could support. So yeah, the technology won, but many investors did not.
Research on technology investment cycles shows how a genuine breakthrough can still attract too much capital before the expected returns arrive. Now look back at the current AI buildout. Companies can keep selling chips, building data centers and spending billions. Investors can still overpay for that infrastructure if financing costs rise while the expected returns arrive later than promised.
Check Your Portfolio Before the Bill Arrives
Right now, Washington and the AI industry are reaching into the same pool of capital, and somebody will eventually pay more to get it.
Economists call that “crowding out”, which sounds academic until you put real names and real money behind it. The U.S. government needs investors to absorb trillions in debt. Meanwhile, NVIDIA ?and its financial partners are talking about mobilizing more than $500 billion for AI infrastructure. The Federal Reserve Bank of Minneapolis? laid out the basic problem decades ago: when government borrowing and private investment compete for available capital, one side can push the cost of funding higher for the other.
Now bring it back to your holdings. If you own a heavily indebted REIT, a cash-burning software stock or a company that needs constant outside funding, you are not watching this collision from the sidelines. Your company may need to borrow in the same market where Washington already demands enormous amounts of money and AI infrastructure keeps pulling in more.
That is the squeeze. The business with strong cash flow can keep writing its own cheque. The business drowning in debt has to ask somebody else for one.
And when everybody suddenly wants more money from the same lenders, the company you own had better have a damn good reason for getting it first.
