Nvidia at $6T Has a Bond Problem No AI Story Can Fix

This is the part of the story that never makes the headline.

Everyone’s watching Nvidia march toward a $6 trillion market cap while the AI narrative runs white-hot. Headlines scream about chip demand, data center buildouts, and the next generation of models. All real, all massive.

But here’s what nobody wants to talk about: the 10-year Treasury is sitting above 4%, and that changes everything about how you value a company whose biggest earnings are five years out. Interest rates don’t just affect your mortgage or credit card—they quietly rewrite the entire math behind tech stock prices.

When bonds pay real money again, Wall Street has to answer a harder question: why wait seven years for Nvidia’s earnings when I can lock in 4.5% risk-free today?

Strip away the hype and the mechanism is actually pretty simple.

How Interest Rates Quietly Reprice Every Growth Stock

A stock’s price today is just the sum of all its future cash flows, discounted back to the present. That discount rate matters more than most people realize—and interest rates are the foundation of that rate.

When the risk-free rate (the 10-year Treasury) was 1.5% in , a dollar earned in was worth about 87 cents today. Now, with rates above 4%, that same dollar is worth closer to 73 cents. The further out the earnings, the bigger the haircut.

Tech stocks—especially high-growth names trading at 40x or 50x forward earnings—load almost all their value into years five, seven, ten. They’re selling a story about the future, not today’s profit. That’s fine when money is free. It gets expensive when the alternative is a 4.5% yield you can clip without the volatility.

This isn’t theory. From late through , the Nasdaq dropped over 33% while the Fed hiked rates from near zero to over 5%. It wasn’t because AI stopped being real. It was because the present value of got marked down hard.

Why Are Tech Stocks Still Rising If Rates Are High?

Good question. If higher interest rates punish long-duration assets, why is Nvidia up and the Nasdaq near all-time highs?

Two things can be true at once. Rates reprice valuations lower, but earnings growth can overwhelm that drag. Nvidia isn’t expensive *despite* rates—it’s expensive because the market believes revenue will triple in four years. If that happens, the company grows into its valuation even with a higher discount rate.

But here’s the shift: the tolerance for “maybe” just disappeared. In , unprofitable SaaS companies traded at 20x revenue on the promise they’d figure out margins later. Today, if you’re not profitable or at least close, you get cut in half. The names still working are the ones printing actual cash—Nvidia, Microsoft, Meta. The rest got repriced violently.

The cost of waiting for future profits just doubled, and only the names with the strongest cash flow can afford to pay it.

What Happens If Rates Stay Higher for Longer?

If the 10-year stays above 4% for the next two years, tech multiples have more room to compress. Not because AI is fake, but because the opportunity cost of holding a volatile stock instead of a bond keeps rising.

Look at what happened in the late ’90s. The best tech companies in the world—Microsoft, Cisco, Intel—were growing fast and profitable. Then interest rates rose from 4.5% in 1998 to over 6% by , and the Nasdaq lost half its value. The businesses were fine. The valuations were not.

It’s not a perfect comparison—AI infrastructure spending is more durable than the dot-com bubble. But the mechanism is the same: higher discount rates mean future earnings are worth less today, and at some point, the multiple breaks.

The Rotation Nobody Talks About

Here’s the uncomfortable truth: for the first time in over a decade, bonds are a legitimate alternative to stocks. A 10-year Treasury at 4.5% with zero volatility looks pretty attractive when the S&P trades at 21x earnings and tech stocks can drop 15% in a week.

Institutional money is already moving. Bond fund inflows hit their highest levels since . That’s not panic—it’s math. If you’re a pension fund or endowment, locking in 4–5% for ten years solves a lot of problems, especially if you think equity returns over the next decade will be closer to 7% than 15%.

Asset	Yield / Expected Return	Volatility
10-Year Treasury	4.5%	Low
S&P 500 (historical avg)	~10%	Moderate
Nasdaq / Tech-Heavy	~12%+ (if growth holds)	High

The gap between risk-free and risky just narrowed. That’s not bearish by itself, but it does mean tech has to work harder to justify the premium.

Does This Mean Sell Everything and Buy Bonds?

No. It means understand what you own and why you own it.

If you’re holding Nvidia or Microsoft because you believe AI infrastructure spending will compound for years, higher interest rates don’t erase that thesis—they just make the bar higher. The winners will still win, but the multiple you pay today matters a lot more when the risk-free rate is 4.5% instead of 1%.

If you’re holding speculative names with no earnings and a ” story,” that’s where the pressure shows up first. The market doesn’t kill growth stocks when rates rise—it just stops paying 40x sales for them.

The repricing is already happening. You just have to look past the Magnificent Seven to see it. Unprofitable tech is down 40–60% from highs, even as the Nasdaq recovered. That’s not a coincidence—it’s interest rates doing exactly what the textbook says they should.

Why do higher interest rates hurt tech stocks more than other sectors?

Tech stocks load most of their value into distant future earnings. When rates rise, those future dollars get discounted more heavily—a dollar earned in is worth about 15% less today at 4% rates versus 1.5%. Value stocks and dividend payers, by contrast, deliver cash now, so the discount effect is smaller.

At what interest rate level do bonds become a better bet than stocks?

There’s no magic number, but historically, when the 10-year yield exceeds the S&P 500’s earnings yield (earnings divided by price), bonds start to look competitive. Right now, the S&P trades at roughly 21x earnings, or a 4.8% earnings yield—almost neck-and-neck with Treasuries. That’s the narrowest gap in over a decade.

Can tech stocks keep rising if rates stay high?

Yes, if earnings grow fast enough to offset the higher discount rate. Nvidia can still double from here if revenue triples in four years—the math works. But the market will demand proof, not projections, and any stumble gets punished harder when the risk-free alternative pays 4.5%.