Nvidia and AMD Just Bought $4 Billion of Meta’s Stock—And They’re Fighting the Very Thing Meta Is Building

 Nvidia and AMD Just Bought $4 Billion of Meta’s Stock—And They’re Fighting the Very Thing Meta Is Building

**By [moonlight editory team] | October 27, 2026 | 12-Minute Walkthrough**

---

## The Short Version (For the Busy American)

Meta just sold a chunk of itself to the companies that supply its chips, and the irony is almost too thick to cut.

As part of Meta’s massive **$95 billion equity raise** to fund its AI buildout, **Nvidia, AMD, and Broadcom** have indicated they’ll buy up to **$5 billion** of Meta’s Class A stock. Reports peg the actual commitment closer to **$4 billion** from Nvidia and AMD alone . The chipmakers aren’t doing this out of charity. They’re doing it because their biggest customer needs cash to keep buying their chips—and because Meta is simultaneously building **custom silicon** designed to replace them.

Meta’s in-house **MTIA** chips, developed with Broadcom, are already shipping. The company plans **four generations** of the accelerator by 2027, scaling to **multiple gigawatts** of capacity . That’s a direct assault on the merchant GPU model that made Nvidia a trillion-dollar company.

**The punchline:** Nvidia and AMD are investing in their own disruption. They’re buying equity in a customer that’s building a cheaper, specialized alternative to their products—because the alternative to investing is watching that customer run out of money and stop buying chips altogether.

---

## What Actually Happened: The Circular Capital Web

Let me break down the transaction, because the structure is revealing.

### The Raise

Meta launched a **$35 billion** offering of common and preferred stock, later upsized to **$37.5 billion**. A separate **$60 billion at-the-market program** in its new non-voting Class C shares will follow in early 2027 . It’s Meta’s first stock sale since 2013—a stunning reversal for a company that bought back **$26.2 billion** of its own shares in the first nine months of 2025 and **zero** in 2026 .

The money is going straight into AI infrastructure. Meta’s capital expenditures are running at **$135 billion to $145 billion** this year. Free cash flow turned **negative $2.12 billion** last quarter—the first negative quarter in over a decade .

### The Chipmaker Backstop

Nvidia, AMD, and Broadcom’s **$5 billion** indication is non-binding. They can buy more, less, or none at all . But the signal matters. Ryan Lee of Direxion put it plainly: the chipmakers **“need major cloud-services providers like Meta to continue spending on chips”** .

The AMD relationship is the most entangled. In February, AMD agreed to supply **6 gigawatts** of MI450 chips to Meta in a deal reportedly worth **more than $100 billion**. AMD also granted Meta warrants for up to **160 million AMD shares** at one cent each . If AMD buys Meta stock, the two companies will own pieces of each other—a **mutual hostage arrangement** that aligns their incentives in ways that traditional supplier relationships don’t.

### The Custom Chip Threat

Here’s what makes this circular dance so precarious. Meta isn’t just buying Nvidia and AMD chips. It’s building replacements.

**MTIA**—Meta’s Training and Inference Accelerator—is being developed with Broadcom on a **2-nanometer process**. The first gigawatt of capacity is committed, with plans to scale to multiple gigawatts. Broadcom CEO Hock Tan said the product is **already shipping** .

Meta’s roadmap includes **four MTIA generations by 2027**: MTIA 300, 400, 450, and 500. They’re designed for Meta’s specific workloads—recommendation systems, ranking, and generative AI inference—where a general-purpose GPU is overkill and a specialized ASIC is cheaper and more efficient .

The strategic goal is explicit: **reduce dependence on Nvidia and AMD** .

---

## The Human Touch: Why This Matters Beyond the Boardroom

Let me put this in terms that actually connect to your life.

### Your 401(k) Owns This Web

If you own an S&P 500 index fund, you own Meta, Nvidia, AMD, and Broadcom. The circular financing arrangement means their fates are now more tightly bound than ever.

When Nvidia buys Meta stock, it’s using its own capital to support a customer’s balance sheet so that customer can keep buying Nvidia chips. If Meta’s AI investments pay off, Nvidia wins twice—once on chip sales, once on equity appreciation. If Meta’s AI investments fail, Nvidia loses on both. Your retirement account is exposed to that leveraged bet.

### The “Circular Capital” Concern

This isn’t just a Meta story. Across the AI industry, suppliers and customers are becoming shareholders in each other. SpaceX just arranged **$40 billion** in GPU financing through an Apollo affiliate that will buy equipment and lease it back . Anthropic locked in **$450 billion** in cloud commitments with Google, Amazon, and Microsoft.

LSEG researchers warned last week that these **“ownership webs”** make it harder to identify where **independent demand** originates and who ultimately bears the financial risk . When Nvidia’s revenue growth depends partly on companies it has invested in, the line between **genuine customer demand** and **financed demand** blurs.

### The Funding Crunch Is Real

Here’s the context that explains why this is happening now. Meta’s cash and securities stood at **$82.45 billion** at the end of Q3, barely above its **$80.92 billion** in long-term debt . Its free cash flow is negative. Its capex is exploding. And it has roughly **$825 billion in future spending commitments** .

The debt market, as Elon Musk warned over the weekend, has **“almost entirely shut off”** for AI companies without independent revenue. Meta has independent revenue—its ad business grew **26% to $64.6 billion** last quarter—but even that isn’t enough to fund the AI buildout alone .

So Meta is selling equity. And its chip suppliers are buying, because they need Meta to stay solvent enough to keep spending.

---

## Frequently Asked Questions

### Why are Nvidia and AMD buying Meta stock?

Because they need Meta to keep buying their chips. Meta is one of the largest buyers of AI accelerators in the world. If Meta’s cash flow can’t fund its capex, it buys fewer chips. By investing in Meta’s equity offering, the chipmakers are **supporting their customer’s balance sheet** and ensuring the spending continues .

### How much are they actually buying?

Meta said Nvidia, AMD, and Broadcom have indicated interest in up to **$5 billion aggregate** of Class A shares. Reports peg the Nvidia/AMD commitment closer to **$4 billion** . The indications are **non-binding**—they can buy more, less, or none .

### Are Nvidia and AMD fighting Meta’s custom chips?

**Yes, indirectly.** Meta is developing **MTIA**, its own custom AI accelerator, with Broadcom. The goal is explicitly to **reduce dependence on Nvidia and AMD** . Nvidia and AMD are investing in Meta anyway because the alternative—Meta running out of cash and cutting chip orders—is worse than the long-term threat of custom silicon.

### What is MTIA and why does it matter?

MTIA is Meta’s **in-house AI chip**. It’s built on a **2-nanometer process** with Broadcom and is already shipping at a **1-gigawatt scale**, with plans to expand . Meta has **four generations** planned by 2027. For workloads like recommendation systems and inference, custom ASICs can be **cheaper and more efficient** than general-purpose GPUs .

### Is this “circular financing”?

**Yes, and regulators are watching.** The LSEG analysis warned that “ownership webs” make it harder to trace where demand originates and who bears the risk . When suppliers invest in customers, revenue growth can reflect **financed demand** rather than genuine end-market demand. It’s not illegal, but it complicates the picture for investors trying to value AI stocks.

### Why is Meta raising so much money?

Because **AI infrastructure is expensive**. Meta’s capex is **$135-145 billion this year**, and its free cash flow is **negative $2.12 billion**. It has **$825 billion in future spending commitments**. Its ad business is strong, but not strong enough to fund the buildout alone. The **$95 billion** raise—$35 billion now, $60 billion later—bridges the gap .

### Does this mean the AI bubble is bursting?

**Not necessarily.** Meta’s ad revenue is growing 26%. Nvidia’s chips are in genuine demand. But the **financing structure** is under stress. When companies have to sell equity and their suppliers have to buy it to keep the spending going, it signals that **the debt markets are no longer willing to fund the boom alone**. That’s a warning sign, not a death knell.

### Should I buy Meta stock?

I can’t give personalized investment advice. What I can tell you is that **dilution is real**—the $95 billion raise represents roughly **6% of the company** . But the ad business is strong, the AI investments are generating measurable returns in engagement and ad performance, and the chipmaker backstop provides some validation. The risk/reward depends on your time horizon and your belief in Zuckerberg’s superintelligence bet.

---

## The Path Forward: What to Watch

### The MTIA Rollout

Watch for announcements about MTIA deployment at scale. If Meta can shift a meaningful share of its inference workloads to custom silicon, it reduces its dependence on Nvidia and AMD—and improves its margins. That’s the long-term threat the chipmakers are hedging against .

### The Class C ATM Program (Early 2027)

The **$60 billion** at-the-market program in non-voting shares is the next funding tranche. If Meta sells aggressively, dilution continues. If it holds back, it signals confidence in its cash flow trajectory .

### The Chipmaker Follow-Through

Nvidia and AMD’s indications are non-binding. Watch for the final pricing and allocation. If they follow through and buy the full $4-5 billion, it signals commitment. If they pull back, it signals caution .

### The LSEG Scrutiny

The “ownership web” concern isn’t going away. Watch for more research from analysts and regulators about whether AI revenue growth is **genuine demand** or **financed demand**. If the distinction becomes a market narrative, AI stocks could face a repricing .

---

## Conclusion: The Circular Bet

Here’s what I want you to take away from this deal.

Meta needs money to build AI infrastructure. Its chip suppliers need Meta to keep buying chips. So the chip suppliers are buying Meta stock to help Meta keep buying chips.

It’s a **circle**, and it works as long as the AI payoff arrives before the money runs out.

Nvidia and AMD are hedging. They’re investing in a customer that’s building their replacement. They’re doing it because the short-term risk of Meta running out of cash is greater than the long-term risk of Meta succeeding with custom silicon.

For American investors, the message is uncomfortable: **the AI trade is becoming a closed loop**. The companies selling the shovels are buying equity in the miners. The line between supplier and customer is blurring. And the whole system depends on the AI revolution generating returns before the financing runs dry.

Meta’s ad business is strong. Nvidia’s chips are in demand. But the **circularity** is a warning. When the people selling you the tools have to buy your stock to keep you buying, the boom is no longer self-sustaining.

It’s financed. And financed booms have a way of ending badly.

---

## Disclaimer

This article is for informational purposes only and does not constitute financial advice, investment advice, or a recommendation to buy or sell any securities. The author is not a licensed financial advisor. All investments involve risk, including the potential loss of principal. Past performance does not guarantee future results. Readers should conduct their own research and consult with a qualified financial professional before making any investment decisions. Data and figures cited are accurate as of the date of publication but may change. The author holds no position in any securities mentioned.

---

**Tags:** #Meta #Nvidia #AMD #Broadcom #MTIA #CustomSilicon #AIStocks #StockMarketToday #FinancialNews #MarketAnalysis #Investing #CircularFinancing #EquityOffering #Capex #FreeCashFlow #AIInfrastructure #DataCenters #SemiconductorStocks #StockMarket #Nasdaq #TechStocks #InvestmentStrategy #RiskManagement #Dilution #ClassCShares #AI Trade #MarketVolatility #EarningsSeason #Zuckerberg #GPU #ASIC