What are the consumer confidence indexes (Michigan and Conference Board)?

Consumer spending drives roughly two thirds of the American economy, so what households feel today is a rough sketch of what they will spend in six months. Two monthly surveys try to measure that feeling — and one of them contains a component the Federal Reserve genuinely acts on.

What do they actually measure?

Both are surveys, not hard data. They ask households two kinds of question: how things look right now — job security, whether this is a good time to make a big purchase — and what they expect over the next six to twelve months.

That split matters more than the headline. The current-conditions half mostly confirms what already happened. The expectations half is the part with any claim to being forward-looking, because an intention today becomes spending later.

Neither index measures what anyone did. It measures what people say they might do, which is a weaker thing — and the gap between cheerful surveys and flat retail sales opens up often enough that you should never treat sentiment as a spending forecast.

The two indexes are not the same survey

They are published by different institutions, on different dates, from different samples, and they are weighted differently.

• University of Michigan consumer sentiment — a preliminary reading in the middle of the month and a final one at the end, drawn from a small sample. It leans on expectations and reacts sharply to petrol prices and to inflation.
• Conference Board consumer confidence — released on the last Tuesday of the month, from a much larger sample, and weighted toward the labour market.

Because Michigan publishes twice, it gives you two tradeable moments: the preliminary print, and a final one that can revise it meaningfully.

Why do the two so often disagree?

Because they are sensitive to different things. Michigan responds to what people pay at the pump and in the supermarket. The Conference Board responds to whether jobs feel easy or hard to get — its survey includes a question comparing jobs "plentiful" against "hard to get", which tracks the unemployment rate reasonably well.

So a divergence is not noise, it is the message. If Michigan drops while the Conference Board holds, households are worried about prices, not employment. If the Conference Board drops while Michigan holds, the labour market is cracking first. Either reading tells you something a single headline number cannot.

The part the Fed actually watches

Michigan publishes inflation expectations: what households expect prices to do over the coming year, and over the next five to ten years.

This is the component that can move the dollar on its own, and it is worth understanding why. A central bank can live with inflation being temporarily high. What it cannot live with is people expecting high inflation indefinitely, because expectations feed wage demands and pricing decisions and become self-fulfilling. Economists call it expectations becoming unanchored, and preventing it is close to the whole job.

So when the long-run expectations line moves, the Fed notices — and a rate path that shifts is what actually reaches the currency market. The headline sentiment number rarely does that by itself.

How does it trade?

Treat it as a medium-impact release that shapes the narrative between the big ones rather than setting direction on its own.

• The surprise matters, not the level. A reading of 62 means nothing in isolation; 62 against a forecast of 70 means a lot. This is the same logic as any other calendar event — see forecast, previous and actual.
• Watch the revision. Michigan's final print can move away from the preliminary, and a market that has already settled on the first number reacts to the correction.
• Position sizing over prediction. These are not CPI or payrolls, but the spread widens anyway. The rules in trading news without getting your stop hunted apply unchanged.
• Expect it to be partly priced. Sentiment is correlated with things the market already saw that month — stock prices, fuel costs, headlines. Much of the move is already in the price before the release.

The takeaway

Consumer confidence is a survey of feelings, and feelings are a soft predictor of behaviour. Read the two indexes as a pair rather than one at a time, because the disagreement between them tells you whether households are worried about prices or about jobs. And if you only have time for one line in the whole release, make it Michigan's inflation expectations — that is the one the central bank is reading too.