🧥 NVIDIA Isn’t Just Selling AI Chips Anymore — It’s Financing the AI Economy
Its latest 13F reveals something more interesting than a stock portfolio: NVIDIA is increasingly becoming the banker, supplier and infrastructure partner of the AI boom.Hey Wealth Builders 👋 This has been talked about NVIDIA many times before in any major news outlet or commentary: spectacular growth, Blackwell, CUDA, the Cisco-vs-AI-bubble debate, Google’s TPUs, Meta’s diversification, and even those strangely calm “controlled exit” sell-offs. But NVIDIA’s latest Q2 FY2026 13F filing adds another layer to the story. And this one is fascinating. As of June 30, 2026, NVIDIA’s disclosed U.S. equity portfolio had exploded from $18.37 billion in Q1 to roughly $63.4 billion. The biggest new twist? SpaceX. NVIDIA now owns roughly 122.8 million SpaceX shares worth about $21 billion. (Financial Times) Suddenly, “NVIDIA is a chip company” feels a little… 2023. 😏 🔎 The Portfolio Is Actually an AI Infrastructure MapHere is what NVIDIA owned at June 30: The SEC filing confirms the seven Q1 holdings, while the latest filing adds SpaceX and materially increases the portfolio's value. (Securities and Exchange Commission) Look at the pattern. Intel = compute architecture/manufacturing. CoreWeave + Nebius = compute customers. Coherent + Nokia = moving the data. Synopsys = designing the chips. SpaceX = AI + communications + enormous future compute requirements. This isn't Warren Buffett-style “I like this business” investing. It looks much more like ecosystem engineering. 💰 And Here’s Where It Gets Really InterestingNVIDIA isn't merely investing in companies that may benefit from AI. Some of these companies are also customers, suppliers or strategic partners of NVIDIA. For example, NVIDIA invested another $2 billion in CoreWeave at $87.20 per share, while CoreWeave committed to expanding its NVIDIA-powered AI infrastructure. (Securities and Exchange Commission) NVIDIA also invested $2 billion in Coherent at $256.80 per share, alongside a multibillion-dollar purchase commitment and capacity rights for optical products. (Securities and Exchange Commission) Synopsys? $2 billion at $414.79 per share. (Securities and Exchange Commission) Nokia? $1 billion at $6.01 per share. (Nokia Corporation | Nokia) Intel? NVIDIA originally invested $5 billion at $23.28 per share. That Intel investment is now worth roughly $30 billion. That's about a $25 billion mark-to-market gain before considering dividends or other effects. (Tom's Hardware) Jensen Huang isn't just buying stocks. He appears to be trying to shape the ecosystem in which NVIDIA sells its products. And that is both brilliant… …and potentially dangerous. 🌀 The Question Nobody Should Skip: Who Ultimately Funds the Demand?Imagine NVIDIA invests $2 billion into an AI cloud provider. The cloud provider uses the money to build data centers. It buys NVIDIA GPUs. NVIDIA books revenue. NVIDIA generates cash. NVIDIA invests more money. The customer buys more GPUs. Congratulations. You have built an AI flywheel. 🚀🚀🚀 But now ask the uncomfortable question: How much of the flywheel is powered by independent end-user demand, and how much is powered by NVIDIA's own capital? That's not an accusation of accounting fraud. It is a question about economic quality. A customer buying GPUs because thousands of external customers are paying for its AI services is very different from a customer expanding mainly because its strategic supplier is helping finance the expansion. Both can produce legitimate revenue. But they don't necessarily deserve the same valuation multiple. And this is where our earlier NVIDIA discussions connect. We previously asked whether AI demand resembles Cisco in 2000. The answer was: not necessarily. NVIDIA has enormous profits, genuine demand and extraordinary cash generation. But now we have another question: How much of future AI demand is truly organic? That's the footnote worth reading. ⚠️ Three Risks I Think Investors Should Watch1. Double exposureIf AI spending crashes, NVIDIA could get hit from both directions. GPU demand falls. AND the value of its AI-related investments falls. That's not diversification. That's correlation wearing a fake moustache. 🥸 NVIDIA itself warns that investments can create significant volatility, that private investments may be illiquid, and that some companies may never become profitable. (Securities and Exchange Commission) 2. Customer concentration doesn't disappearOur earlier NVIDIA analysis highlighted hyperscaler competition, custom ASICs and Google's TPUs. That thesis still stands. NVIDIA's FY2026 filing says two direct customers represented 22% and 14% of revenue, while indirect AI demand is also concentrated. (Securities and Exchange Commission) So even while NVIDIA is investing throughout the ecosystem, it remains dependent on a relatively concentrated group of buyers. 3. The 13F isn't the whole storyThis may be the biggest mistake investors make. A 13F is a window, not the entire house. It doesn't capture the full universe of private investments, commercial commitments, guarantees, partnerships or strategic arrangements. NVIDIA disclosed $17.5 billion of private-company and infrastructure-fund investments in fiscal 2026, plus $3.5 billion of land, power and shell guarantees to early-stage companies. It also reported $27 billion of investment commitments as of April 2026. (Securities and Exchange Commission) So don't look at $63 billion and conclude: “That's NVIDIA's entire investment portfolio.” It isn't. It's the visible part. 🧮 So Should We Value NVIDIA Differently?Yes — but not because P/E suddenly became useless. P/E and P/S remain useful. But they shouldn't be the only lens. I would use a four-layer NVIDIA valuation framework: 1️⃣ Value the operating business. 2️⃣ Value the investment portfolio separately. 3️⃣ Stress-test ecosystem revenue. 4️⃣ Then decide the ecosystem premium. NVIDIA may deserve a premium because it is increasingly controlling pieces of the AI stack. But the premium should be based on incremental economic value, not simply on how impressive the portfolio looks. 🎯 What Should Retail Investors Actually Do?Don't copy NVIDIA's portfolio. You don't have Jensen Huang's balance sheet. 😅 Instead, copy the questions:
This also connects perfectly with our earlier “picks and shovels” and Physical AI newsletters. The opportunity may not always be the obvious AI winner. Sometimes the money is in the plumbing. ☕ Why Wealth Builder Can Help With This ProblemInvestors today don't suffer from a lack of information. They suffer from too much of it. 13Fs, earnings calls, AI hype, valuation debates, macro news and 47 different people shouting “BUY THE DIP!” before breakfast can turn investing into financial whiplash. Newsletters like Wealth Builder, together with carefully curated passive-income and investing publications, help turn that noise into understandable frameworks, ideas and research questions. Instead of simply telling you what happened, the goal is to explore why it matters, what assumptions may be hiding underneath, and what investors can watch next. You still do your own due diligence—but you don't have to start every research journey from scratch. 👉 Want more high-signal ideas from fellow investors and like-minded newsletters? Explore them here: 🚨 The Big TakeawayThe NVIDIA story has evolved. First it was: “NVIDIA sells GPUs.” Then: “NVIDIA powers AI.” Now perhaps it is: “NVIDIA helps finance, supply and shape the AI ecosystem.” That could create an extraordinary competitive moat. It could also create extraordinary financial interconnectedness. And that's why the smartest question isn't: “Is NVIDIA overvalued or undervalued?” It is: “How much of NVIDIA's growth is organic — and how much is NVIDIA helping finance itself?”Because when the supplier becomes the investor, the customer, the banker and the ecosystem architect… the valuation game changes. 🔥 WATCH THE LOOP.#NVIDIA #NVDA #AIInvesting #AIInfrastructure #13F #WealthBuilding #StockMarket #Investing #Semiconductors #PhysicalAI #PassiveIncome #FinancialLiteracy ⚠️ DisclaimerThis newsletter is for educational and idea-sharing purposes only and does not constitute investment, financial, tax or other professional advice, nor a recommendation to buy or sell any stock, ETF, cryptocurrency or other asset. Markets are uncertain and investments can lose value. Please conduct your own due diligence and consult qualified professionals where appropriate. Use this newsletter as one perspective among many, not as a substitute for your own research and judgment. This is part of my own learning, research and self-improvement process. 📚 Notes & Sources
|
Wealth Builder
Most investors stare at revenue growth like it's the final boss. It isn't. Growth tells you how fast a company is moving. It doesn't tell you what that growth costs. A company growing 50% while burning mountains of cash can look spectacular right up until the cash runs out. Another growing 25% while throwing off enormous amounts of cash may quietly become a compounding machine. That's the real reason the Rule of 40 is useful. Not because 40 is magic. But because it forces us to ask the...
The biggest retirement mistakes don't happen at 65. They happen at 35. Imagine sitting across from 1,000 retirees in America, Singapore, Europe, Australia, or Japan. Ask them one simple question: "If you could rewind your financial life, what would you change?" Almost nobody says, "I wish I'd bought more NVIDIA." Or... "I should've sold before that market crash." Instead, the answers are surprisingly similar. "I thought I had more time." "Healthcare cost far more than I expected." "I...
Half of retail investors say they've copied famous traders. The dangerous part isn't who they're copying. It's what they do AFTER they see the trade. We have all played this game. Buffett buys it. “I should buy it.” Nancy Pelosi buys it. “I should REALLY buy it.” 👀 Elon Musk tweets about it. “WHERE IS MY BROKER APP?” 😂 But a fascinating new survey suggests this isn't just internet entertainment. A July 2026 MarketWise survey of 1,005 U.S. retail investors found that 50% said they had copied a...