A plain-English guide to a CPI-plus-Michigan Friday.

Friday's analysis piece is the live example. August CPI printed headline +0.4 percent and core +0.3 percent at 8:30 AM ET; Michigan preliminary Consumer Sentiment printed 47.8 (weakest since May's record low) at 10:00 AM ET, with five-year inflation expectations at 3.4 percent (the anchoring problem). This reference piece is the framework for reading a Friday tape where both prints land ninety minutes apart, and specifically for how to price the interaction between the two.

Why the two prints together are more informative than either alone

Most CPI Fridays have CPI at 8:30 AM ET and no comparable Michigan print until either the following Friday (preliminary) or two Fridays out (final). When the CPI Friday and the Michigan preliminary land the same day, the market gets two orthogonal signals inside a ninety-minute window: a hard-data read on inflation itself (CPI) and a soft-data read on how consumers perceive both current inflation and future inflation (Michigan). The interaction between the two determines the FOMC's reaction function much more precisely than either read alone.

The Fed's reaction function does not treat "inflation is 3.4 percent" and "consumers expect inflation to be 3.4 percent in five years" as independent pieces of information. It weights them together, because the mechanism the Fed cares about is inflation persistence, and persistence is a function of both realised inflation and expected inflation. A hot CPI print with anchored expectations is easier to look through than a hot CPI print with unanchored expectations, even when the aggregate inflation number is the same in both cases.

The four combinations and what each means

The 2x2 matrix of hard-inflation-hot-or-cool by expectations-anchored-or-unmoored gives four setups, each with a different tape signature and FOMC read:

Hot CPI + anchored expectations. The setup is "supply-side impulse, look through". DXY moves 20-40 pips on the print; front-end moves 3-5bp; the FOMC's reaction function stays close to the current path. Historically this is roughly 30 percent of hot-CPI Fridays. Common in 2016-2019.

Hot CPI + unmoored expectations. The setup is "persistence signal, react". DXY moves 60-120 pips on the print; front-end moves 8-15bp; the FOMC's reaction function shifts materially. Historically this is roughly 15 percent of hot-CPI Fridays. This is the September 11, 2026 setup.

Cool CPI + anchored expectations. The setup is "target-consistent, path unchanged". DXY moves 15-30 pips on the print; front-end moves 4-8bp lower; the FOMC's path stays on the existing dot plot. Historically this is roughly 25 percent of CPI Fridays.

Cool CPI + unmoored expectations. The setup is "the market has already priced the persistence risk, take it out". DXY move is small and often two-directional; front-end moves 6-10bp lower on the CPI, then partially retraces on the Michigan number. Historically this is roughly 30 percent of cool-CPI Fridays.

Reading the ninety-minute window

The specific mechanic to watch is the position of DXY, gold, and the 2-year yield at 9:45 AM ET (fifteen minutes before Michigan). The market has spent 75 minutes digesting CPI, and the fresh Michigan print at 10:00 AM is the second orthogonal signal. Three signatures to look for:

1. Full extension on CPI, small residual on Michigan. DXY has moved most of its total daily range by 9:45 AM; the Michigan print produces a 5-10 pip residual and closes near the CPI-implied level. This is the "hard data dominates" signature and it usually maps to the "hot CPI + anchored expectations" case.
2. Half-extension on CPI, big second push on Michigan. DXY has moved half its range by 9:45 AM; the Michigan print at 10:00 AM drives a second impulse that is 60 to 100 percent as large as the CPI impulse. This is the "soft data confirms" signature and it usually maps to the "hot CPI + unmoored expectations" case (which is Friday September 11, 2026's read). It is the most dangerous setup for the Fed because both realised inflation and expected inflation are pushing the same direction and the FOMC's reaction function has to compound the two impulses.
3. Reversal on Michigan. DXY has moved on CPI in one direction and Michigan reverses it. This usually maps to a print that was already priced (large market move on CPI, but expectations printed opposite the direction of realised inflation) and it is the ambiguous signature where the FOMC can find flexibility.

The specific footprint on Friday September 11, 2026

Friday delivered the half-extension-plus-big-second-push signature. DXY moved from 99.20 pre-CPI to 99.28 at 9:45 AM (a modest 8 pip extension, roughly 50 percent of what the surprise implied under the framework's usual elasticity). Michigan at 47.8 with five-year expectations at 3.4 percent then drove a second push to 99.42 by 10:30 AM. The second impulse was 14 pips, nearly twice as large as the first, exactly the "hot CPI plus unmoored expectations" signature. The FOMC's reaction function is going to compound the two signals when it sets the September SEP dot plot on Wednesday.

What this framework says about the September 16 FOMC

The setup coming out of Friday puts the FOMC in a "hike plus hawkish dots" corner. Hike is required to demonstrate the reaction function is symmetric (up and down) rather than one-sided. Hawkish dots are required to signal that the persistence problem is being addressed with more than the immediate move. The dot plot median matters more than the hike itself: the market has already priced 25bp; the SEP median is where the surprise lives.

Under the framework's read, a September 16 SEP median for 2026 year-end at 3.90 or 4.00 percent is roughly in-line and the reaction is contained. A median at 4.10 percent or higher is hawkish and pushes DXY through 100 on the print. A median at 3.75 percent or lower (implying the hike is the only move) is dovish and produces the setup-inverting response DXY through 98.50.

Related reading

• Friday's analysis piece: the live example this framework describes.
• Inflation expectations surveys: the Michigan five-year methodology and how the FRBNY SCE differs.
• CPI components: what shelter, gasoline, and services line items each carry into the read.
• Pre-CPI positioning: the setup framework that describes the Thursday-Friday transition.
• SEP and the dot plot: how the Fed builds the September projection and where the surprise lives.