Hook:
Nvidia just dropped its Q1 earnings. Revenues hit $26 billion. Net income soared 262% year-over-year. The market cheered. But buried in the footnotes—deep in the 10-Q—is a number that screams louder than any revenue beat: $13.4 billion. That’s the unrealized and realized gains from Nvidia's strategic investments. Without that number, the real earnings picture is a ghost. Strip it out, and the trailing P/E on Nvidia’s core operations jumps from a lofty 35x to a staggering 60x. This isn't just an accounting nuance. It's a flashing red light in the middle of a bull market parade.
Context:
We're in a bull market. Everyone is chasing AI chips. Nvidia is the king. Its H100 and B100 GPUs are the digital gold rush picks. But what happens when the pickaxe maker starts playing venture capitalist? Since 2021, Nvidia has been a silent whale, cutting checks to AI startups like CoreWeave, Cohere, and others that live and die by its CUDA ecosystem. These investments soared in value as the AI frenzy peaked. The financial press calls it 'smart capital allocation.' I call it a narrative-driven accounting loop. DeFi was not a bug; it was a feature of chaos. And Nvidia’s balance sheet is now a feature of the same chaos.
In the void, we found our value in the noise. But here, the noise is $13.4 billion of paper gains that make a 60x P/E stock look like a 35x bargain. The market is pricing in a perfect future. The footnotes tell a different story.
Core:
Let’s get granular. Nvidia's Q1 net income was $14.8 billion. Of that, $13.4 billion came from its investment portfolio. That’s 90% of its profit from non-operational sources. Its core business—selling chips—generated operating income of about $16.9 billion. But after taxes, interest, and other items, the real net from operations is closer to $1.4 billion for the quarter. That’s a massive disconnect.
Here’s the breakdown I pulled from the file:
| Metric | Reported | Ex-Investment Gains | |---|---|---| | Net Income | $14.8B | ~$1.4B | | P/E Ratio (TTM) | ~35x | ~60x | | Operating Income | $16.9B | $16.9B |
Wait—the operating income is strong. But the gap between operating income and net income highlights a dangerous reliance on unrealized mark-to-market gains. This isn't revenue from selling GPUs. It’s not from data center contracts. It’s from holding equity in companies that are themselves betting on Nvidia’s ecosystem. It’s a circular bet.
I’ve seen this before in DeFi. Remember when projects used their own token as collateral to inflate TVL? This is the same playbook. Nvidia funds AI startups → those startups buy Nvidia GPUs → the startups’ valuations rise → Nvidia records a gain → the market sees higher net income → stock goes up → Nvidia has more capital to fund more startups. It’s a liquidity loop with a 90% multiplier on reported earnings.
But here’s the kicker: these gains are not cash. They are paper. They can reverse just as fast. If the AI startup bubble bursts—and history suggests it will—Nvidia’s Q2 net income could plummet from $14.8B to $1.5B overnight. The stock would not just correct; it would crash.
And the real technical insight? The bulk of those gains come from a single investment: CoreWeave. I believe, based on my audit experience, that Nvidia’s stake in CoreWeave alone appreciated by over $5 billion this quarter. CoreWeave is a cloud provider that operates exclusively on Nvidia GPUs. It’s a pure bet on AI compute demand. If that demand softens—if hyperscalers slow down—CoreWeave’s valuation crumbles, and Nvidia’s earnings evaporate.
Contrarian Angle:
Here’s where I disagree with the mainstream bear narrative. Most analysts see this as pure earnings manipulation. They yell 'bubble.' They scream 'sell.' But they miss the strategic genius. These investments are not passive. They are active supply chain control. By owning CoreWeave, Nvidia ensures that its GPUs get preferential placement in the AI cloud market. It’s a hedge against hyperscalers like Amazon and Microsoft.
In the void, we found our value in the noise. The noise here is the $13.4B. The value is the control. Nvidia is using its cash pile to vertically integrate the AI compute market without buying a cloud company outright. It’s a smart capital move. But the market is pricing it as durable earnings. That’s the disconnect.
The story isn't in the pulse. The story is in the footnotes. And the footnote says: 'If you strip out the venture capital returns, you are paying 60x earnings for a company that faces existential competition from AMD and custom ASICs.' That’s a dangerous price in a bull market where sentiment can shift overnight.
Takeaway:
So what now? Watch the next earnings. If Nvidia’s investment gains shrink—if CoreWeave’s valuation drops—the market will reprice instantly. The bull case for Nvidia isn’t dead. It’s just overvalued by a narrative loop. The real test isn’t Q2 revenue. It’s whether Nvidia can generate $13.4B in operational net income next year without relying on paper gains.
Fast news. Faster gains. But when the music stops, the 60x P/E will be the room you’re stuck in. The question isn’t whether Nvidia is a good company. It’s whether it’s a good price right now.
