Did the "Magnificent 7" just turn themselves into commodity producers?
The parallels between their data center land grab now and the shale oil boom-and-bust cycle of the 2010s are striking to longtime commodities analyst Jeff Currie, who is now at Carlyle. He and I talked about this on the show yesterday, and I wrote last week about the "revenge of the old economy" trade that we are seeing again.
But what we're witnessing now is even more than just a scramble for physical goods, and a super capex cycle. Big tech companies are transforming themselves altogether, and what they're turning into is something that their investors--and all of us--need to understand.
Currie warns that what they're becoming is something he is very familiar with: commodity producers. The trouble with that is the "Mag 7" exploded in value over the past fifteen years precisely because they were not cyclical, asset-heavy competitors producing a commoditized good. Now, arguably, they are.
As we wrote last month, and as Ed Yardeni's team correctly foresaw, the problem with AI is that instead of leaving the Mag 7 "as seven independent kingdoms, protected by large moats...prosper[ing] with its own unique monopoly," it has now pitted these companies in an arms race directly against each other.
What's worse, Currie warns, is that they are increasingly funding this race with debt, and producing an asset that is becoming commoditized--and could potentially drop in value.
One way to track this is the "H100 rental index," produced by Silicon Data--a proxy for demand for AI compute. Much like the shale boom was predicated on $100 oil (which never happened), if rental prices were to sink to thirty or fifty cents an hour instead of staying in the $1-2 range, Big Tech could find itself in a similar predicament, Currie warns.
Energy companies wound up destroying 41 cents of every dollar invested during the shale boom, wiping out $2.6 trillion in equity. "The biggest winners of the shale revolution were U.S. citizens and their government," he wrote.
If the "Mag 7" companies don't want to meet a similar fate, they'll have to show investors that they are each building differentiated, market-dominating AI tools with a high return on investment. And even if they can do that, the "steel in the ground" that is now part of their business model could still leave them trading at lower valuations.
See you at 1 p.m!
Kelly