Economic indicators
What is Non-Farm Payrolls, and why does it move the dollar?
Non-Farm Payrolls, almost always shortened to NFP, is the monthly count of how many jobs the US economy added or lost outside of farming. It is released by the Bureau of Labor Statistics, usually on the first Friday of the month at 13:30 UK time, and it is the single most watched scheduled release in currency markets.
Farming is excluded because agricultural employment swings wildly with the seasons and would drown out the signal. What is left is the part of the labour market that actually tells you something about the economy.
The reason it matters is the same reason inflation matters: employment is one of the two things a central bank is mandated to care about. A jobs number changes what the Federal Reserve is expected to do with interest rates, and rate expectations are what price a currency.
It is not one number, it is three
Almost every headline reports the payroll count alone. That is the least informative of the three figures released at the same moment, and it is why so many first reactions get reversed within minutes.
| Figure | What it is | Why it matters |
|---|---|---|
| Payroll count | Net jobs added, in thousands | The headline. Volatile and heavily revised. |
| Unemployment rate | Share of the labour force without work and looking | Slower moving, harder to dismiss, from a separate survey. |
| Average hourly earnings | Wage growth, month over month and year over year | The inflation link. Often the one that actually moves the dollar. |
They can and do contradict each other, and when they do the market has to decide which story it believes. A strong payroll count with falling wage growth is a very different report from a strong count with accelerating wages, even though the headline is identical.
Two surveys, one release
The payroll count comes from a survey of employers. The unemployment rate comes from a survey of households. They measure different things in different ways, which is why the two can point in opposite directions in the same month.
Someone holding two part time jobs counts twice in the employer survey and once in the household survey. Someone self employed may not appear in the employer survey at all. Neither figure is wrong; they are answering different questions.
Why wages often matter more than jobs
This is the part that separates people who read NFP from people who watch it.
A central bank raising rates to fight inflation is watching for wage growth, because wages are what turn a temporary price shock into persistent inflation. If prices rise and wages follow, households can keep paying the higher prices and inflation embeds. If prices rise and wages do not, demand falls away and inflation burns itself out.
So a jobs report showing plenty of hiring but cooling wage growth can be read as good news on inflation, and can weigh on the dollar despite a strong headline. The count says the economy is fine; the wages say the central bank has less reason to keep policy tight. Markets weight the second.
Revisions, and why last month keeps changing
Every NFP release revises the previous two months. These revisions are routinely large enough to change the meaning of the original print, and they arrive with almost no coverage.
A report showing a strong month alongside a heavy downward revision to the month before may represent no improvement at all, just a number moved from one month to another. Reading the headline in isolation, you would conclude the labour market strengthened. Reading it with the revision, you would conclude nothing much changed.
The practical habit is to look at the revision line before reacting to the headline. It takes a few seconds and it is the most common thing people skip.
Reading a report in practice
- Count against forecast. Not against zero, and not against last month. Markets price the consensus in advance, so only the gap is new information.
- Revisions to the prior two months. Add them to the headline mentally. A beat of 30k alongside a downward revision of 40k is a miss.
- Average hourly earnings. Year over year is the number to watch. This is the inflation channel and often the real driver of the move.
- Unemployment rate, and why it moved. A rate that fell because people found work is different from one that fell because people stopped looking. The participation rate tells you which.
- Did yields agree? If two year Treasury yields moved with the dollar, the market believes the policy path has changed. If not, expect the move to fade.
The forex part: it only prices against something
NFP is a US release, so it moves the dollar leg of every pair at once. That does not mean it moves every pair equally.
A strong report lifts the dollar most against a currency whose own central bank is on hold or cutting, because the expected rate gap widens on both ends. Against a currency whose central bank is also tightening, the same report may move the pair very little, because both expected paths shifted together and the differential is roughly where it started.
This is why traders watching only the US side are regularly surprised by how small a move a dramatic headline produced. The pair prices the difference between two economies, which is the argument made at more length in what market context means in trading.
Why the first move is so often wrong
The first seconds after the release are dominated by automated reactions to the headline count. The composition, the revisions and the wage detail take longer for humans to read, and the market frequently reverses once they have.
This is not a strategy recommendation. It is a reason to understand that the initial spike is a reaction to one number out of a report containing several, and it has not yet been read.
Related terms
ADP employment is a private payroll estimate released two days earlier. It is widely used as a preview of NFP and has a poor record of predicting it. Treat it as its own data point rather than a forecast.
Jobless claims are published weekly and measure new filings for unemployment benefits. Far more frequent than NFP and useful for catching a turn between monthly reports.
JOLTS measures job openings. It says something about labour demand rather than hiring, and central banks have watched it closely as a gauge of how tight the labour market is.
Participation rate is the share of working age people either working or looking. It is what tells you whether a falling unemployment rate is good news or not.
Keep reading
- What is market context in trading?
- What is CPI, and why does it move currencies?
- What is the COT report?
See the labour market on both sides of your pair
MarketContext puts the two economies behind a currency pair side by side: every release against its forecast, what it does to the policy path, and where bond yields sit against it. You read the data and reach your own conclusion.