Soft data with one genuinely hard component that central banks take seriously.
intermediate · 4 min read · 12 XP
Confidence surveys ask people how they feel, and how people feel is a famously poor predictor of what they do. There is one exception inside these releases that matters enormously — the inflation expectations component, which central banks treat as a policy input in its own right.
Confidence surveys ask households about their view of current conditions and their expectations for the future — jobs, income, prices, and whether now is a good time to make a major purchase.
The main ones a currency trader sees:
| Survey | Region | Notes |
|---|---|---|
| University of Michigan Sentiment | US | Preliminary mid-month, final at month end. Contains the inflation expectations series |
| Conference Board Consumer Confidence | US | End of month. Weighted more towards labour-market perceptions |
| GfK / European Commission confidence | Europe | Monthly, region and country level |
| Ifo Business Climate | Germany | Business rather than consumer; a well-followed euro-area indicator |
| ZEW Economic Sentiment | Germany | Surveys financial analysts, not businesses or households |
Confidence indices have repeatedly diverged from actual spending for extended periods. Households have reported deeply pessimistic sentiment while continuing to spend at a healthy pace — a pattern strong enough that the relationship between confidence and consumption is best described as loose.
Reasons: survey answers are influenced by news coverage, fuel prices and political affiliation, none of which necessarily change behaviour. What people say about the economy is partly a statement about how they feel, and spending is constrained by income and credit rather than mood.
Do not trade confidence as a consumption forecast
The consumption forecast inside a confidence survey is weak. If you want to know what households will spend, watch income, credit availability and the labour market. Confidence is a mood reading, and mood is a noisy proxy for behaviour.
This is why the University of Michigan survey remains a genuine market event.
It publishes household inflation expectations at one-year and five-to-ten-year horizons, and central banks treat long-run expectations as close to a policy target in their own right.
The reasoning is central to modern monetary policy. If households and firms believe inflation will return to target, they behave in ways that help deliver it — moderate wage demands, moderate price increases. If they believe inflation will stay high, they bargain and price accordingly, and the expectation becomes self-fulfilling. This is what policymakers mean by expectations becoming "unanchored", and preventing it is treated as close to non-negotiable.
Why one line in a soft survey can outrank the survey
A jump in the five-to-ten-year inflation expectations series can move the dollar more than the entire headline confidence index, because it speaks directly to whether the central bank's credibility is intact. A bank that sees long-run expectations drifting will tighten, or hold tight, regardless of what growth is doing. Watch that line specifically on Michigan days.
For the euro, Germany's two headline sentiment surveys are frequently confused.
ZEW is released earlier in the month and gets a reaction; Ifo carries more information. When the two disagree, weight Ifo.
The spending. Confidence surveys measure how households feel, and feelings respond to news coverage, fuel prices and politics without necessarily changing behaviour. Actual expenditure is the behaviour. The divergence is genuinely informative in one way, though: it suggests spending is being supported by income or savings rather than optimism, which makes it more vulnerable if the labour market turns. Use the confidence reading as a fragility flag, not as a forecast.
| Component | Impact |
|---|---|
| Headline confidence index | Low — soft data, weak link to behaviour |
| Michigan long-run inflation expectations | Moderate to high — a direct policy input |
| Labour-differential in the Conference Board survey | Moderate — a useful early labour signal |
| Ifo expectations | Moderate for the euro |
| ZEW | Low — largely reflects recent market moves |
What to remember
Consumer confidence measures how households feel, which has repeatedly proved a loose predictor of how they spend — so the headline indices are soft data with limited trading value. The exception is the inflation expectations series inside the Michigan survey, which central banks treat as close to a policy target because expectations that become unanchored are self-fulfilling.