University of Michigan Consumer Sentiment Index
A thermometer for the consumer and an anchor-meter for inflation expectations: the final print's 4-star billing is half sentiment, half that expectations line.
Twice monthly (prelim mid-month, final month-end, 10 AM ET)
What it is
Compiled by the University of Michigan's Surveys of Consumers since 1978, the sentiment index draws on telephone interviews with roughly 600 households covering personal finances, buying conditions and the outlook. It is released twice a month: a preliminary (mid-month) and a final (month-end). The same release carries household inflation expectations — the 1-year and 5–10 year series — the latter being the Fed's core window on whether inflation expectations remain "anchored". On calendars like Jin10 the final carries 4 stars, the preliminary 3, and the expectations lines 3 — this page covers all three.
Release schedule
| Item | Details |
|---|---|
| • | Frequency: twice monthly — preliminary (mid-month) and final (last Friday of the month) |
| • | Release: 10:00 ET (23:00/00:00 Beijing time, DST-dependent) |
| • | Contents: sentiment index + current-conditions and expectations components + 1-year and 5–10 year inflation expectations |
| • | Sample: ~600 households by phone — smaller than the Conference Board (~3,000), hence larger month-to-month swings |
| • | Frame note: the preliminary covers a partial-month sample; the final completes it — the revision between them is itself information about intra-month sentiment shifts |
Why it matters
The unique value is inflation expectations: the 1-year series tracks households' felt price experience (highly sensitive to gasoline and groceries), while the 5–10 year series measures whether expectations remain anchored — a systematic un-anchoring forces the Fed to pay a higher policy price to prevent a self-fulfilling wage-price spiral. Sentiment itself leads consumption, which is about 70% of US GDP; deteriorating expectations components lead spending slowdowns. Together they give a 600-household survey outsized market influence.
Impact across assets
Typical impacts (using the "sentiment down + expectations up" combination):
| Asset | Typical impact |
|---|---|
| US equities | Falling sentiment pressures consumer/discretionary; with expectations rising, the "stagflation" narrative broadens the damage |
| US Dollar Index | Rising expectations → hawkish-Fed pricing → firmer; sentiment-driven declines → growth worries → softer |
| Gold | Rising expectations → medium-term hedge bid; but if it drives a more hawkish Fed → short-term pressure — separate the time scales |
| Crypto | A sentiment collapse → risk-off → pressured; in liquidity-driven regimes it follows tech |
| Treasuries | Rising expectations → higher nominal yields (inflation compensation); sentiment-led declines → yields down |
How to read it
The standard read:
| Dimension | How to read it |
|---|---|
| Sentiment level | Compare with the historical range: elevated = consumption momentum intact; a fast step-down = a consumer warning — direction and speed beat the absolute level |
| Expectations vs current conditions | The expectations component is the lead: falling expectations with stable current conditions = the "worry phase"; both falling = recession underway |
| 1-year inflation expectations | A "felt" gauge tracking gasoline and groceries — volatile; a single-month jump is not de-anchoring |
| 5–10 year expectations | The anchoring gauge: historically stable in a narrow band; a systematic drift upward is what the Fed fears |
The advanced frame: separate sentiment from expectations — falling sentiment with stable expectations = a demand-side worry (bond-friendly); un-anchoring expectations with stable sentiment = an inflation-pricing problem (more complicated for bonds and gold). Also track the preliminary-to-final revision: a large revision signals a second-stage shift in mood within the month.
Limitations & common mistakes
- Treating the 1-year expectation as de-anchoring: it is gasoline-sensitive and volatile — watch the 5–10 year for systematic drift; a one-month 1-year jump is "felt fuel prices".
- Over-reading a single month: ~600 households swing naturally — trend and revision direction beat any single print.
- Confusing it with the Conference Board: different question structures (U-Mich leans on personal finances and buying conditions; CB on labor markets) — divergence is normal and itself informative.
- Ignoring the prelim-final revision: the preliminary uses a partial sample — trading the final off the preliminary means accepting revision risk.
- Treating sentiment as a coincident consumption gauge: its value is the forward-looking expectations component; the current-conditions component overlaps more with hard retail data.
Related macro data
How it links to other macro data:
- With retail sales: sentiment is the forward questionnaire, retail the hard data — a deteriorating expectations component leads retail weakness by 1–2 quarters. retail-sales
- With CPI: household expectations are the "expectations side" of CPI — un-anchoring changes the speed of wage-price pass-through. cpi
- With the Fed decision: the anchoring gauge is a core FOMC input — a systematic rise in long-horizon expectations reprices the entire rate path. fed-rate
Symbols most sensitive to U-Mich Consumer Sentiment
Symbol pages that list this data as a factor to watch:
FAQ
Q When is the U-Mich sentiment released?
Twice a month at 10:00 ET: the preliminary around mid-month (after the second Friday) and the final on the last Friday. That is 23:00/00:00 Beijing time depending on DST. The final has the complete sample; both draws move markets.
Q Why do the inflation expectations matter more than sentiment itself?
Sentiment is partly covered by hard data (retail, jobs), but household inflation expectations are a unique gauge: the 5–10 year series is the direct anchoring read, and a systematic un-anchoring reprices everything via a more hawkish Fed. The 1-year is a felt-cost thermometer for fuel and groceries.
Q Preliminary vs final — what differs?
The preliminary lands mid-month on a partial sample; the final completes the month. The gap between them is second-stage information: a preliminary that weakens and a final that cuts further confirm a trend better than either alone.
Q Which is better, U-Mich or Conference Board?
Wrong question: U-Mich is ~600 phone households focused on personal finances and buying conditions; CB is ~3,000 respondents focused on labor markets. Different samples and questions — divergence is normal, and reading both beats picking one.
Q What is the direct trading use?
Three uses: a consumption lead (expectations component leads retail), the expectations side (the 5–10 year anchor gauge feeding rates and gold), and event vol (prelim/final prints move equities and bonds in the release hour).
Turn MT5 / MT4 into a Tray Tool
Check quotes, manage positions, and hide in one click.
Download Camovia TrayThis page is educational content about macroeconomic data. It is not investment advice. Macro impacts involve multiple interacting factors — always combine them with your own risk management.