UMich Sentiment 47.8 (vs 51.0 Est) — Inflation Expectations Jumped 0.6pp to 4.6%

Fundamentals · 2026-09-11

Consumer sentiment 47.8 against a 51.0 consensus and 51.7 in August — a 3.2-point miss, down 7.5% on the month and 13.2% on the year, and the second-lowest reading of 677 months since November 1952; only May 2026's 44.8 is lower. The split is the story: current conditions fell just 1.9% to 50.9 while expectations collapsed 11.1% to 45.8. Year-ahead inflation expectations jumped 0.6 points to 4.6%, the highest since June, and long-run expectations rose to 3.4%, ending three straight months at 3.3%. Survey director Joanne Hsu names the cause: "a resurgence in fuel prices and trade tensions." But she also reports that independents were little changed, and her own research says the national estimate tracks independents.

What It Changes

• It is the second hawkish print of the morning, and the one the Fed weights more. Core CPI at 8:30 was one month of data. Long-run inflation expectations breaking a three-month run at 3.3% is the anchoring question, and it is the argument a hawk reaches for on 16 September.
• Year-ahead expectations jumped 0.6 points in a single month, to 4.6%. That is not a drift, and Hsu attributes it directly to fuel prices — the same fuel prices that put 36.8% of this morning's CPI headline into one line item.
• The sentiment collapse itself is growth news, not inflation news. At 47.8 this is the second-weakest reading in seventy-four years, and the weakness is entirely forward-looking: expectations fell 11.1% while current conditions fell 1.9%.
• One caveat does real work here. Hsu reports Democrats and Republicans both fell sizably while independents barely moved — and the survey's own published research finds national estimates align with independents. The headline drop may be substantially partisan; the inflation-expectations jump is harder to explain that way.

Impact

• USD — Mixed, lean bullish — rising inflation expectations five days before a live meeting outweigh a soft sentiment print, because the committee's stated test is about inflation credibility rather than growth.
• Against it: 47.8 is a demand signal, and a consumer this pessimistic eventually shows up in the activity data the other side of the mandate reads.
• The dollar had already absorbed a hot core at 8:30. Whether it extends on this is the same asymmetry question — the hike was largely paid for at roughly 69% before either number landed.
• US Indices (ES / NQ / YM / RTY) — Bearish — consumption is roughly two-thirds of the economy and the forward-looking half of this survey just fell 11.1%.
• Discretionary retail takes the sharpest read. This follows yesterday's wholesale trade report, where seven consumer-facing categories were shipping less in real terms than a year ago, and an existing-home market at a fourteen-month low.
• RTY is the most exposed of the four to a domestic-consumption shock, and it also carries the most rate sensitivity into a hawkish expectations print.
• Gold (GC) — Mixed, lean bullish — this is the first datapoint in some time that argues the Fed may not contain this.
• Gold has spent the week falling into rising yields, which is gold trading as a rates instrument. Long-run expectations breaking a three-month anchor is the kind of thing that flips it back to trading as an inflation hedge.
• The regime marker is unchanged: gold rising alongside yields would say the market has stopped believing containment. One survey does not establish that.
• Crude (CL) — no directional read — causation runs the other way. Fuel prices are the stated input to this survey, not its output.

Inside The Number

The preliminary September Index of Consumer Sentiment came in at 47.8, against a consensus of 51.0 and August's 51.7. That is a 3.2-point miss, a 7.5% fall on the month and 13.2% on the year.

Put it in the series and the number gets harder to dismiss. There have been 677 monthly readings since November 1952, and only one has ever been lower than this — May 2026, at 44.8. August's 51.7 was already extreme by historical standards. September is the second-worst print in seventy-four years.

The composition is where this becomes a specific claim rather than a mood. Of the two sub-indexes, current economic conditions fell 1.9% to 50.9 while the index of consumer expectations fell 11.1% to 45.8. Consumers are saying their present situation is roughly what it was a month ago and that the next twelve months look materially worse. Hsu puts it plainly: "Year-ahead expectations for both personal finances and business conditions plunged."

On inflation, the survey moved in both series at once. Year-ahead expectations jumped from 4.0% to 4.6%, which Hsu calls the highest reading since June, and which she notes "substantially exceeds the 3.4% seen in February before the Iran conflict began, along with all 2024 readings." Long-run expectations ticked up to 3.4%, ending three consecutive months at 3.3%, and remain above their 2024 range of 2.8% to 3.2%.

The 0.6-point move in the year-ahead series is the larger number, but the 0.1 in the long-run series is the one that matters to a central bank. Year-ahead expectations track the pump; five-to-ten-year expectations are the closest thing a survey gets to a measure of whether people still believe the target. Three months at 3.3% was a plateau. One month at 3.4% is not a break. It is, however, the wrong direction in the week of a live meeting.

Hsu names the cause, and this morning's data corroborates it. "With a resurgence in fuel prices and trade tensions, consumers anticipate greater pressures on their pocketbooks to come." The consumer price report published ninety minutes before this one put gasoline up 3.9% on the month and 27.4% over twelve months, with gasoline alone accounting for 36.8% of the entire headline. Energy is up 16.3% on the year and fuel oil 52.0%. The survey period closes before the CPI release, so these consumers were not reacting to a statistic — they were reacting to the price on the forecourt that the statistic measures.

Now the caveat, which is substantial and comes from the survey itself. Hsu reports that "Democrats and Republicans alike posted sizable declines, while independents were little changed from August." The Surveys of Consumers has published repeatedly on exactly this problem, including a May 2026 report titled "National Estimates Continue to Align With Views of Independents" and an April 2025 report on partisan perceptions in sentiment measurement. If the national estimate tracks independents, and independents did not move, then a meaningful share of a 7.5% headline decline is partisan rather than economic.

That caveat cuts unevenly, which is the useful part. It weakens the sentiment collapse considerably. It does not obviously weaken the inflation-expectations jump, because a 0.6-point move in year-ahead expectations against a documented 27.4% rise in gasoline prices has a straightforward non-partisan explanation sitting right next to it.

One more reason to hold this loosely: it is preliminary, and last month's preliminary was wrong. August's initial reading was 51.0 and was revised up to 51.7 in the final. That is a 0.7-point upward revision, and it happens to be the exact figure forecasters used as this month's consensus. The final September number lands on 25 September.

Hsu also flags what did not deteriorate: "Five-year expected business conditions remained stable at readings well below their historical average, suggesting that consumers believe that emerging risks this month may not have further worsened the long-run outlook." Consumers think this year is bad. They have not concluded that the next five are worse than they already thought.

The longer arc is the bleakest part. Hsu puts sentiment 16% below February, before the Iran conflict began, and 13% below a year ago. Checked against the series: February was 56.6, so 47.8 is 15.5% below it, and September 2025 was 55.1, so 13.2% below that. The 2026 average is now 51.7, against 57.6 in 2025 and 72.5 in 2024. The consumer has lost roughly a third of their 2024 confidence in under two years.

The mechanism behind that is in the data this desk has been assembling all fortnight: consumer prices up 3.4% over twelve months against average hourly earnings up 3.1%. Wages have been losing to prices, and the family home is now losing too — yesterday's existing-home median rose 1.6% against that same 3.4% inflation rate.

The Internals

The three indexes, and how far each fell:

Index · September 2026 · August 2026 · September 2025 · Month change · Year change

Index of Consumer Sentiment · 47.8 · 51.7 · 55.1 · Down 7.5% · Down 13.2%

Current Economic Conditions · 50.9 · 51.9 · 60.4 · Down 1.9% · Down 15.7%

Index of Consumer Expectations · 45.8 · 51.5 · 51.7 · Down 11.1% · Down 11.4%

Against consensus, and against the history:

Measure · Reading · Comparison · Result

Headline sentiment · 47.8 · 51.0 consensus · Missed by 3.2 points

Headline sentiment · 47.8 · 51.7 in August · A second consecutive monthly fall

Rank in the series · 47.8 · 677 readings since November 1952 · Second-lowest ever recorded

Only lower reading · 44.8 · May 2026 · Four months ago

2026 average · 51.7 · 57.6 in 2025, 72.5 in 2024 · Down a third from 2024

Inflation expectations, which is the half of this release the Fed reads:

Horizon · September 2026 · August 2026 · Change · Context

Year ahead · 4.6% · 4.0% · Up 0.6 points · Highest since June; 3.4% in February pre-conflict

Long run, five to ten years · 3.4% · 3.3% · Up 0.1 points · Ends three straight months at 3.3%

Long run versus 2024 range · 3.4% · 2.8% to 3.2% in 2024 · Above the whole range · The anchoring question

The 2026 path, which is the shape of a bad year rather than a bad month:

Month · Sentiment · Note

January 2026 · 56.4 · The year's high point

February 2026 · 56.6 · Last reading before the Iran conflict

March 2026 · 53.3 · First conflict month

April 2026 · 49.8 · Below 50

May 2026 · 44.8 · The series low

June 2026 · 49.5 · Partial recovery

July 2026 · 55.2 · Recovery holds

August 2026 · 51.7 · First of two declines

September 2026 · 47.8 · Preliminary, and the second-lowest ever

Who Is Actually Gloomy

The survey director's own framing, which is more careful than the headline:

What Hsu reported · Why it matters

Democrats and Republicans both posted sizable declines · The aggregate fall is being driven by partisan respondents

Independents were little changed from August · The group the survey's research says tracks the national estimate did not move

Year-ahead expectations for personal finances and business conditions plunged · The deterioration is forward-looking, not about present circumstances

Five-year expected business conditions remained stable · Consumers did not extend this month's shock into the long-run outlook

Fuel prices and trade tensions named as the cause · Both are verifiable against this week's price data

Why the partisan caveat has teeth, and where it stops:

Claim · Does the partisan caveat weaken it?

Sentiment fell 7.5% to a near-record low · Yes, substantially — independents were flat

Expectations fell 11.1% against current conditions at 1.9% · Partly, for the same reason

Year-ahead inflation expectations rose to 4.6% · Not obviously — gasoline is up 27.4% over twelve months

Long-run expectations rose to 3.4% · Not obviously, and this is the series the Fed watches

What this morning's consumer price report says the survey was reacting to:

Price · Monthly change · 12-month change

Gasoline, all types · Up 3.9% · Up 27.4%

Energy, all · Up 2.1% · Up 16.3%

Fuel oil · Up 10.1% · Up 52.0%

All items · Up 0.4% · Up 3.4%

Average hourly earnings, for comparison · Not published monthly here · Up 3.1%

Against This Morning's Open

• The Open called what would matter in this release, and it was right. At 06:30 it said UMich "matters mainly for its inflation-expectations components — one-year expectations eased to 4.0% in August, and the long-run series has held 3.3% for three months running" — The Open. Year-ahead went to 4.6% and the long-run series broke its three-month run at 3.3%. Both of the things it pointed at moved.
• It listed the release as hitting USD and ES, with prior 51.7 and no consensus published. The consensus was 51.0 and the print was 47.8.
• The sentiment headline is the one thing The Open did not anticipate, and neither did the consensus — a 3.2-point miss to the second-lowest reading on record was not on anyone's card this morning.
• Taken with the 8:30 release, the morning delivered the hawkish resolution of the core-CPI fulcrum and a jump in inflation expectations. The Open framed the session as a one-decimal question. It got the decimal, and then a second inflation datapoint on top of it.

What This Sets Up

• Final September sentiment — Friday, 25 September, 10:00am ET. The revision matters more than usual: August's preliminary was 51.0 and the final came in at 51.7, a 0.7-point upward revision. A similar revision would take this reading off the second-lowest slot.
• Whether long-run expectations hold 3.4% or fall back to 3.3%. One month above a three-month plateau is not a break in the anchor. A second month is the beginning of one, and it is the single number in this release with the most direct line to a rate decision.
• Whether independents start moving. If the next reading shows independents falling too, the partisan explanation for this month's collapse disappears and the sentiment signal becomes real.
• Whether the year-ahead series tracks fuel back down if fuel falls. Hsu attributes the 0.6-point jump to a resurgence in fuel prices. Crude gave back more than $5 overnight. If expectations do not ease when prices do, the attribution was incomplete.

What Is This?

• What it is: The University of Michigan Surveys of Consumers, running since 1946 and published as the Index of Consumer Sentiment with a preliminary reading around mid-month and a final reading at month end, both at 10:00am ET. It is a survey of US households, directed by Joanne Hsu. The headline index is built from five core questions and splits into two sub-indexes: current economic conditions, which asks about the household's situation now, and consumer expectations, which asks about the next twelve months and the next five years. Alongside the index the survey publishes inflation expectations at one-year and five-to-ten-year horizons, which are watched far more closely than the sentiment number itself.
• Why it matters: Inflation expectations are the mechanism by which inflation becomes self-sustaining, so a central bank that has lost the anchor has lost the argument. This release landed five days before an FOMC with three standing dissents for a hike, and ninety minutes after a consumer price report whose core component ran hot.
• How to read it: Read the two sub-indexes separately — a fall driven by expectations means something different from one driven by current conditions, and this month is emphatically the former. Read the long-run inflation number ahead of the year-ahead number, because the year-ahead series tracks fuel prices almost mechanically while the five-to-ten-year series is the one that measures credibility. And treat the level with care: sentiment has been depressed relative to actual consumer spending for several years, the survey's own research documents a large partisan component, and the preliminary reading revises — last month's moved 0.7 points between the two publications.

</parameter>

Want the full explainer? Economic News Events, Explained breaks down this release and every other one we cover, in plain English.

_For informational purposes only. Not investment advice._