Tomorrow afternoon, after the closing bell, NVIDIA will tell us whether the most expensive bet in the history of capitalism is still working.
Wall Street is expecting roughly 77% year-over-year revenue growth. That number is not arbitrary. It happens to be the exact same growth rate at which the four largest hyperscalers, Microsoft, Meta, Amazon, and Alphabet, are scaling their 2026 capital expenditures. Combined: around $700 billion this year alone, with about 75% of it pointed straight at AI infrastructure.
That symmetry is not a coincidence. It is the entire trade.
The number behind the number
Let’s put $700B into perspective. It is more than the annual GDP of Switzerland. It is more than the combined market cap of every airline on Earth. It is roughly what the entire US federal government spends on Medicare in a year. And it is being committed by four companies, in a single calendar year, almost entirely on the assumption that the chips they buy from NVIDIA and the data centers they fill with them will pay back many times over.
The breakdown roughly looks like this. Amazon at $200 billion. Alphabet around $180 billion. Meta near $125 billion. Microsoft around $115 billion. Almost double what these same firms spent in 2025. And about 75% of that is going into AI servers, GPUs, custom chips, networking, and the power infrastructure to run all of it.
Why tomorrow’s print matters more than usual
In a normal cycle, an NVIDIA earnings report is a single stock event. You watch the beat, you watch the guidance, you trade it or you don’t. This one is different for three reasons.
First, the macro backdrop is the worst it has been in years. Headline PCE just printed at 3.5% year-over-year, up from 2.8% the month before. The Fed has been holding at a 3.5 to 3.75% target range and the market has fully priced out rate cuts for 2026. Bank of America came out blunt last week: no cuts this year. Inflation is sticky, partly because oil is sitting near $107 a barrel thanks to the ongoing Iran conflict and Strait of Hormuz disruption. In that environment, every long-duration growth bet has to defend its valuation in real time.
Second, the competition is finally real. Amazon’s internal chip business is reportedly running at a $20 billion-plus annualized revenue run rate, growing in triple digits. Google has TPUs. Microsoft has Maia. AMD has been quietly stealing share at the edges. NVIDIA still has the moat, but the moat is narrower than it was twelve months ago and tomorrow’s call will tell us how worried Jensen actually is.
Third, last week’s Micron warning rattled the entire memory chain. If high-bandwidth memory demand is showing cracks at the margin, that flows back into NVIDIA’s Blackwell ramp and into 2027 guidance.
What a Wealtharian investor actually watches
You do not need to be a sell-side analyst to read this print. You need to watch four things.
1. Data center revenue versus guidance. If data center growth comes in noticeably below the implied 77% pace, that is the first crack in the capex story. Every hyperscaler watching tomorrow has the option to slow down their own spending in Q3.
2. Gross margins. NVIDIA has been running at gross margins most companies dream about. If margins compress, even by a point or two, that is the market quietly telling you customers are starting to negotiate.
3. Forward guidance commentary on Blackwell and Rubin. The Blackwell ramp is in full swing and Rubin is on deck. Any softness in language around customer commitments is a tell.
4. China. Regulatory drama, export restrictions, and the new generation of compliant chips. Any color here will move the stock more than most people expect.
The contrarian read most people miss
Here is the angle nobody is going to put on cable news tomorrow. The big risk to the AI trade is not that NVIDIA misses. The big risk is that NVIDIA beats, raises, and the stock still sells off. That has happened before in mega-cap tech, and it is usually the signal that the “easy money” stage of a secular trade is ending and the next leg is going to demand actual earnings, not just narrative.
The setup right now feels uncomfortably like that. Records were just made on the S&P last week. Then the index pulled back as yields spiked. Then memory names cracked. Then oil pushed higher again. A clean beat-and-raise might rally NVDA 5% after hours and fade by Friday. If that happens, do not panic. It is just the market rotating from “AI is going to print money” to “show me the cash flow.” Wealth is built by paying attention to which one the market is in.
The bigger wealth angle
Step back from the ticker for a moment. $700 billion in a single year, by four companies, is one of the largest concentrated capital allocations in corporate history. It is comparable in scale to the buildout of the US interstate highway system, except compressed into twelve months. Whatever is being built with that money will reshape labor markets, productivity, energy demand, and yes, the ways ordinary people can build wealth, for the next decade.
The single most important question this earnings season is not “what will NVIDIA do” but “what does this much money chasing this much compute mean for everyone else.” It means lower-cost AI tooling for individual entrepreneurs. It means displaced workers in some industries and new income streams in others. It means utility stocks suddenly look like growth stocks because power is the real bottleneck. And it means there is a window, maybe two years, maybe three, for individuals to position themselves on the right side of the productivity wave instead of getting flattened by it.
Tomorrow’s earnings will move NVDA. The narrative they set will move trillions.
What to do this week
Do not trade the earnings. Almost nobody wins that game. Instead, use tomorrow’s print as a free pulse check on the entire AI trade. If the language is confident and capex commitments hold, the secular thesis is still intact and you can continue to dollar-cost average into broad AI exposure through ETFs or a diversified basket. If guidance gets cautious or hyperscalers start hedging on language, that is your cue to rebalance, take some chips off the table, and watch the power and infrastructure names that benefit no matter who wins the chip war.
The investors who built generational wealth in the dotcom era were not the ones who guessed Cisco’s quarterly print. They were the ones who saw the bigger buildout, sized their exposure correctly, and stayed disciplined through the noise. The same thing is happening now, just in a different generation of infrastructure.
Want to track your own path to financial independence through the AI cycle? The Wealtharian Wealth Tracker lets you monitor your net worth, FU money progress, and investment milestones in one place. Try it free →