The $3 Trillion Nobody Put on the Balance Sheet

Real Estate Insights. Success Mindset. One Step at a Time.

Issue #65 | August 24, 2026

🧠 Market Insights

The $3 Trillion Nobody Put on the Balance Sheet

Every quarter, the big AI spenders tell you what they spent. This week, the Wall Street Journal told you what they didn't. The Journal dug through the footnotes of the latest filings. Nine of the biggest tech companies are sitting on roughly $3 trillion in commitments tied mostly to AI. None of it shows up as debt on any balance sheet. That's five times what those same companies reported in capex last year.

The cleanest example sits on nearly 4,000 acres in Richland Parish, Louisiana. That's Meta's Hyperion data center, now reported above $200 billion. Meta financed it through a joint venture with Blue Owl Capital. Blue Owl owns 80%. Meta owns 20%. Morgan Stanley structured $27 billion in debt through a special purpose vehicle nicknamed Beignet. S&P rated it A+, leaning heavily on Meta's own credit. That debt never touches Meta's balance sheet. Meta's own auditor, EY, flagged the structure in the footnotes of the annual report.

Here's the part worth sitting with. Nearly 4,000 acres of Louisiana land isn't worth anywhere close to $27 billion. Not even with power secured. The real collateral behind Beignet isn't the dirt. It's the dirt plus Meta's promise to keep paying rent on it for twenty-plus years. Strip Meta's credit out of the deal and the land underneath is worth a fraction of that. What's left is a lot of specialized concrete and steel with exactly one buyer in the world who'd want it as-is.

Here's what that means if you're not selling to Meta. The same dynamic is showing up on much smaller deals. The ones landing on your desk and your sellers' desks every week. Somebody calls. The number's bigger than anything the parcel's ever fetched. They talk like the deal is done. They don't have anywhere near the capital Meta has behind Beignet.

Data Center Watch tracked $130 billion in US data center projects blocked or delayed in the first quarter of 2026 alone. That nearly matches all of 2025, in three months. Cancellations jumped to 25 in 2025, up from six the year before. Microsoft froze roughly 1.5 gigawatts of its own self-build projects. It canceled leases tied to power delivery delays. It kept raising its AI budget anyway. The bottleneck now is the physical layer. Transformers are running 40 to 60 week lead times. Interconnection queues stretch 5 to 7 years. Over 300 data center bills are moving through state legislatures this year alone.

None of that means the boom is over. Capex guidance keeps going up, not down. It means the same question we asked about Beignet applies to the call you're getting about your own dirt. How much of this is real? How much of it is a story wearing a number?

🐾 Dirt Dawg Take

Wall Street hides its risk in a $3 trillion footnote. Your sellers can't afford to hide from it. Every option and JV pitched at ground level carries the same question Beignet does. How much of this deal is dirt? How much of it is somebody's word?

✨ The Alpha Effect

We Love the Upside. That's the Problem.

Sam Zell put it plainly. Everybody wants to look at how good a deal can get. People love focusing on the upside. That's where the fun is. What amazed him was how little most people think about the downside. He wasn't talking about data centers. He didn't need to. It's the same story every cycle, wearing a different hat.

Here's what makes it worse than a simple blind spot. A boom isn't just something that happens to the market. It's something that happens to a room. Everybody around you is telling the same story. The broker just closed a bigger number than you've ever seen. The seller down the road got a call nobody expected. The headlines say the buildout has no ceiling. Asking the downside question out loud in that room feels lonely. Going along with the story feels safe. That's backward. It's exactly why the discipline is so rare.

Zell's test for every deal was simple. What's the outcome if everything goes wrong? Can I bear the cost? Can I survive it? Notice what that test isn't. It isn't a prediction. He wasn't claiming to know a crash was coming. He just refused to let the good story be the only one he let himself hear.

That's the real discipline. Not calling the top. Not predicting the crash. Just staying willing to hear the version where you're wrong. Everyone around you only wants to talk about the version where you're right.

🐾 The Alpha Shift

The sellers who come out ahead this cycle won't be the ones who chased the biggest number. They'll be the ones who stayed willing to ask what happens if they're wrong. Everyone around them will be too busy being right together.

🚀 One Step at a Time

Action Steps This Week

1. Choose Your Trade: JV or Straight Sale

A straight sale gets you paid now and gets you out. A JV keeps you in the deal for the upside. It also keeps you exposed to everything we just dug into: tenant concentration, stranded infrastructure, a buyer's balance sheet. Neither one is automatically the smart move. Know your own number for certainty versus upside before the first offer hits your desk. Not after.

2. Don't Let a "3-Year Option" Be a Free Look

An option with no real money behind it is a buyer parking on your land for free. They shop it around. They wait on financing. They wait to see if the power ever gets approved. Ask what's non-refundable. Ask how the extension fees escalate if they don't perform. Ask what milestones have to be hit to keep the clock running. If the answer is nothing, you're not in a deal. You're in a waiting room.

3. Know Who's Actually Calling You

Every buyer chasing your land falls into one of four buckets. The wildcatter has no real skin in the deal. He's trying to option your dirt cheap, hoping to flip the deal before anything's due. The developer raising funds has real experience and maybe a signed LOI. He still needs to close his own capital stack, and that's gotten harder. The developer with a strong balance sheet already has a tenant or off-taker in tow. No financing contingency. No guessing whether the lease materializes. Then there's the hyperscaler itself, buying direct. No middleman at all. Do your due diligence before you get attached to the number. It's the fastest way to know which one is actually sitting across the table.

Small steps compound. Know how much of any deal is dirt, and how much is somebody's word. Same test for a $27 billion SPV or a phone call about your back forty.

Here's to building your success, one step at a time...

from the dirt up.

Kris "Dirt Dawg" McGee 🐾

The Dirt Dawg Den