By Liquiditrax ResearchPublished
Key finding. Today two companies riding the same AI wave report, and they are a controlled experiment in how that wave gets funded. Super Micro reports before the open, CoreWeave after the close. Both are expected to roughly double revenue. Only one is expected to make money.
The numbers. Super Micro consensus: earnings per share of $0.92, more than double a year ago, on revenue of about $11.60 billion, up roughly 101% year over year, against a disclosed order backlog above $60 billion (Blockonomi). CoreWeave consensus: revenue of about $2.55 to $2.56 billion, also more than double year over year, but an expected loss of roughly $1.22 to $1.40 per share as it keeps buying GPUs and building data-center capacity, with options pricing a swing of about 15.5% on the print (Yahoo Finance; Yahoo Finance).
Why it matters. Same demand, opposite economics. Super Micro sells the picks and shovels and earns a margin as it does. CoreWeave rents out the picks and shovels it had to finance up front, so revenue growth and cash burn arrive together. Doubling revenue tells you the demand is real in both cases. It tells you almost nothing about which business is fundable at what cost.
This is the funding split, made testable in one day. We have been tracking the market splitting AI names into those that fund their buildout from their own P&L and those that fund it with debt and equity. Today you get to watch both economics report inside a single session, on the same underlying theme. The question for CoreWeave is not whether revenue doubled, it is whether the loss per dollar of revenue is narrowing or widening.
So what for the allocator (ABC). Do not underwrite these two the same way just because they share a theme. The self-funder is an Alpha-sleeve bet on margin durability and backlog conversion. The capacity builder is a bet on cost of capital and dilution, which means the balance sheet and the financing terms matter as much as the revenue line. When two names in one theme trade on completely different risks, sizing them identically is the mistake.
Takeaway. Revenue growth is the easy part of the AI story. Who pays for it, and on what terms, is where the risk actually lives.
Analytics & education, not advice. DYOR.