WHAT THE NUMBERS FEEL
WHAT THE NUMBERS FEEL
A dispatch from the bottom of the Michigan survey
44.8.
Sit with that for a moment. The University of Michigan's final consumer sentiment reading for May came in at 44.8 — weaker than every single economist projection in the Bloomberg survey, and lower than the preliminary read of 48.2 published two weeks ago. It fell through the floor that April had just set as the previous all-time low. It is now below the June 2022 nadir, when CPI was running at 9.1%, Russia had just invaded Ukraine, and the global supply chain was still a crime scene.
Americans feel worse about the economy right now than they did then. Absorb that.
The sentiment series goes back to 1952. We have never been here. Not during the oil shocks of the 1970s that economists still use as their bogeyman. Not during the dot-com wreckage. Not when Lehman failed and the whole apparatus shuddered. Not during a pandemic that closed the world for two years. The University of Michigan has been asking Americans how they feel about their economic lives for 74 years, and the answer, in May 2026, is: worse than ever.
You can dismiss sentiment data if you want. Plenty of strategists do. They'll show you the jobs numbers, point at S&P profit margins — which hit a record 13.4% in Q1, a number that would have seemed delusional five years ago — and argue that vibes are not fundamentals. They're not wrong that the divergence is real. They are wrong that it doesn't matter.
Inflation expectations over the next year jumped to 4.8%. In February, before the US-Israeli war with Iran started and before oil crossed $100 a barrel, that number was 3.4%. One and a half percentage points of inflation expectation added in three months, entirely on the back of a geopolitical event that most Americans couldn't have located on a map before it started. The national average for a gallon of gasoline crossed $4 — the level at which AAA, in a 2022 survey, found a majority of Americans begin actually changing their behavior. Not grumbling. Changing.
GasBuddy's daily active user base nearly doubled in March. People are driving to find cheaper gas. McDonald's is warning analysts about pressure on customer spending. Whirlpool used the phrase "recession-level" to describe the drop in appliance demand. These are not vibes. These are companies watching transactions in real time and reading the data.
"No one cared about inflation until it became a problem," a PNC economist said recently. There's something almost plaintive about that observation. The Fed spent a decade worrying about deflationary spirals. It spent 2021 calling the price surge transitory. Now inflation is so culturally embedded that Google searches for the term hit all-time highs earlier this year. The economy didn't change people's minds. It burned a new category into them.
Here is what makes this moment structurally different from previous sentiment troughs, and why the strategist dismissal misses the point.
In 2022, when sentiment last hit bottom, the Federal Reserve was behind the curve but it could act. Rates were at zero. The hiking cycle had barely started. There was runway. You could tell a story about central bank credibility being restored. Painful, but coherent.
In 2026, that runway is gone. The Fed funds rate is sitting where it is, the new chair was sworn in yesterday morning, the 30-year Treasury closed last week at 5.064%, and one-year inflation expectations are at 4.8%. The policy response space has contracted dramatically. You can't hike into a war-driven oil shock without crushing the consumer who is already, by their own account, more miserable than at any point in recorded survey history. You can't cut without the long end of the curve treating it as an inflation signal and repricing upward again. Warsh is in a box that Powell didn't build and can't be blamed for, but that doesn't make the box less real.
The S&P 500 closed at record highs this week. The Dow is at 50,579. Consumer sentiment is at 44.8. That gap — between what financial assets are pricing and what the people who actually make up the economy are experiencing — is one of the more unsettling disconnects in recent memory. It's not new; the divergence has widened steadily since 2020. But it keeps getting wider, and at some point the question stops being whether a convergence happens and starts being which direction it converges from.
Bitcoin, for what it's worth, is trading around $77,500. Down from $80,000 a week ago, and roughly $34,000 below where it was a year ago. The strategic reserve thesis — Trump's government as perpetual bid — has so far not materialized into a floor. The crypto-friendly Warsh narrative that juiced the space during his confirmation hearing has been thoroughly overtaken by the reality that nobody, not even the most ideologically malleable central banker, can conjure easy money when year-ahead inflation expectations are approaching 5%.
There's an almost classical irony in BTC sitting $34,000 below its year-ago price while inflation fears drive the Michigan survey to its worst reading in history. The asset that was supposed to protect against currency debasement is doing nothing of the sort while the people who most need a hedge are paying $4.50 for gas and choosing between groceries.
The long-term inflation expectations data is the one thing in this report that isn't flashing red. Five-year expectations rose only modestly, suggesting that most Americans still believe, on some level, that the current pain is temporary. That's the last structural support for the idea that the Fed doesn't need to slam rates higher. If that reading breaks — if the five-year starts migrating toward 4% — the calculus changes entirely and the conversation shifts from "when do rates fall" to "how high can they go before something breaks."
Watch that number. It's doing a lot of load-bearing right now.
The rest of the Michigan data is telling a story about a country that is three months into a war it didn't choose, with gas above $4, a new central bank chair who was sworn in at the White House with the president watching, record equity valuations overhead, and a feeling — diffuse, stubborn, historically unprecedented — that something is wrong in a way that the headline numbers have not yet fully priced.
The headline numbers, as they usually do, will catch up eventually.
They always do.