In May, the money gathered in Beverly Hills.

They come every year for the Milken conference—Davos for capitalist whales. 

When someone speaks from that stage, you can bet the people who can move markets are listening. And this year, BlackRock's Larry Fink told them what's coming next.

A new asset class, he said. One born of a boom so vast that the world will run short of capital to build it before demand ever cools.

Compute. 

Don Wilson, who built DRW Holdings, one of the world's biggest trading firms, went further: compute, he said, will become the largest commodity on earth.

Both were pointing at the same asset, set to launch later this year.

Two words: compute futures.

Maybe you haven't heard of it. But soon you won't be able to turn an online corner without hearing the name.

Because, as you read this, the largest players—CME, ICE, and others—are getting it ready for primetime.

Once live, it will reengineer how the AI buildout gets financed, priced, and steered. But first, it will settle the oldest question in the market—real or hype—the only way that ever sticks: in public, with money.

We've seen this dynamic play out a hundred times before. 

Sometimes it ended badly, other times it supercharged markets. But both times, it created untold fortunes for the people who saw it first… and understood what they were looking at. 

The Spot and the Curve

Strip the jargon and a futures contract is a promise. 

A price agreed today for something settled later. Two sides who disagree about tomorrow, bound to a single number.

In one corner, the hedger—the farmer with a crop in the ground, the airline staring down jet fuel. His nightmare is a price swing that erases a year of profit overnight, so he trades away the shot at a windfall for one thing: certainty. Lock the price, sleep at night.

In the other corner, the speculator, who owns no crop and no planes, and who wants the risk the hedger is desperate to shed—because risk is where the profit hides.

Risk passing from the person who fears it to the person who craves it. That's the whole machine.

Run enough of these contracts across enough dates—next month, next quarter, two years out—and something appears that no single price can give you.

A forward curve.

This curve is the market's collective bet on where the price is headed, printed in public, moving every second. 

A spot price tells you what compute costs today. The curve tells you what the smartest money thinks it'll cost when the next wave of data centers switches on—a forecast the whole world can see, and bet against.

Today, the only piece an ordinary investor can touch is a yes/no bet on Kalshi—a coin flip on where the price lands by Friday. Useful, but it's a firecracker next to what's coming. 

The New Oil

Computing power already moves like a commodity. 

One workhorse chip, the H100, rented near eight dollars an hour at its peak. Then it crashed to a dollar or two. Then it ripped forty percent higher again.

The beating heart of the largest buildout in history, swinging like a memecoin, with no way for anyone to lock it in.

For roughly a hundred years, oil was only bought and sold for immediate delivery at whatever the going rate was that day (spot). 

It wasn't until futures contracts let people trade tomorrow's oil at a locked-in price that crude became a financial asset the whole world could hedge, speculate on, and price around, turning it into the most economically and politically powerful commodity on earth.

Wheat, gold, electricity, interest rates—same road. 

Compute is standing exactly where oil once stood. And the same Chicago houses that turned grain and crude into tradeable commodities are doing it again, right now, to the fuel behind artificial intelligence.

The Gauge and the Lever 

Now, the crux of this entire thesis: 

Give the boom a real price and it starts remaking the boom itself.

Financing comes first. A data center borrows billions against depreciating chips—the fragility that drops these companies double digits on a single rumor. Let it sell compute forward and it hands lenders hedged cash flow and borrows cheaper. 

But cheaper capital builds more supply, which drives compute prices down, which can gut the very operators the cheap money was meant to save. The tool that steadies the boom can also accelerate its glut.

Then coordination. A public forward curve would become a signal the whole buildout reads—telling power companies how many megawatts to plan, chipmakers how many fabs to tool, developers where to break ground. A GPU-hour price becomes one of the batons conducting hundreds of billions in power, silicon, and real estate.

Then steering. Once the curve is live, it becomes a live input into every compute-dependent stock you own. A falling curve will weigh on a neocloud's valuation—choking its ability to raise money and shrinking what it can actually build—even as it hands the renters on the other side a windfall.

First these contracts measure the boom. Then they start to steer it. Which is exactly why the people watching this market matter more than the people ignoring it.

How This Plays Out 

The AI boom has been an argument—real, or a bubble? Today that fight gets settled with opinions, because the only price that matters is a snapshot.

It tells you what compute cost yesterday. It can't tell you what the market believes about the wave of data centers coming in 2027—and it gives a skeptic no way to bet against the story.

A futures market changes both. 

It prices the future, and it lets the doubters put money on it. 

If the price holds or climbs as capacity floods online, the buildout is earning its keep. If it craters, the doubters get their evidence—in public, on a live tape—the way the ABX called the subprime crash months before Wall Street would admit it. 

Pretty soon, that referee will walk onto the field. 

The only question: whether you're watching the right screen when the tape and the story finally disagree.