What moves gold

How to Read the Jobs Report (NFP): Household vs Establishment Survey, Revisions and Wages

The US jobs report, often called NFP (non-farm payrolls), is one of the biggest monthly events for gold. But the headline number is only one line of a long report. This guide walks through the Bureau of Labor Statistics release the way a careful trader reads it: the two surveys behind it, the three numbers that matter most, how revisions work, and how gold has actually reacted in the release hour across 26 reports.

Gold spot$3,528.2−99.8 (−2.75%)NextUS Q3 GDP (second estimate) and October PCE inflationin 4d 20h · Wed 13:30 GMTMarket opencloses in 4h 34mRange$123.2
Gold, last 90 sessions, with US jobs-report daysGold, last 90 sessions, with US jobs-report daysDashed lines: non-farm payrolls release days · range $3,515–$4,755 · 20-day average daily range $94.5$4,000$4,500NFPNFPNFPNFP$3,536Aug ’26Sep ’26Oct ’26Nov ’26
COMEX gold, daily open/high/low/close, last 90 sessions · data to 20 Nov 2026 · xauusdglobal.com

The short answer

The jobs report (official name: The Employment Situation) is published by the US Bureau of Labor Statistics (BLS) at 8:30 am New York time, usually on the first Friday of the month. To read it like a trader, check these in order:

  1. Non-farm payrolls (jobs added), against the forecast.
  2. Revisions to the previous two months.
  3. Unemployment rate, against the forecast.
  4. Average hourly earnings (wage growth).

Then ask: is the whole picture stronger or weaker than expected? A stronger report usually lifts US yields and the dollar and pushes gold down; a weaker one usually helps gold. In our data, gold’s release-hour candle averaged a $42 high-to-low range across 26 reports, more than three times a normal hour before the release.

Two surveys, one report

The report combines two different surveys. Knowing which number comes from which helps you understand why they sometimes disagree.

Establishment survey (CES) Household survey (CPS)
Who is asked About 119,000 businesses and government agencies, about 622,000 worksites About 60,000 households
Run by BLS US Census Bureau, for BLS
Main numbers Non-farm payrolls, average hourly earnings, hours worked Unemployment rate, labour force participation
Reference period Pay period that includes the 12th of the month Calendar week that includes the 12th
In the report “B” tables “A” tables

Source: BLS Employment Situation technical note, checked 17 November 2026.

Why they disagree: the establishment survey counts jobs (a person with two jobs counts twice), and leaves out farm workers and the self-employed. The household survey counts people, including the self-employed. In any single month they can point different ways. Over many months they usually tell a similar story.

Number 1: non-farm payrolls

This is the headline: how many jobs the US economy added or lost in the month, after seasonal adjustment. Markets compare it with the consensus forecast (the average of economists’ estimates, shown on any economic calendar).

  • Above forecast: the economy looks stronger, so the Fed has less reason to cut rates. Yields and the dollar often rise; gold often falls.
  • Below forecast: the opposite. Gold often rises.

The size of the surprise matters more than the number itself. +150,000 against a forecast of +50,000 is a big beat. +150,000 against +250,000 is a miss.

Also look at where the jobs came from. Private payrolls usually matter more than government hiring. The report breaks jobs down by industry.

Number 2: revisions

Each report revises the previous two months. The BLS explains that the first estimates are based on the replies received in time, and each later estimate includes more replies and updated seasonal factors. A month only becomes final after two revisions.

How to read it: add the revisions to the headline. For example, a headline of +100,000 with the previous two months revised down by a total of 80,000 is much weaker than it looks.

There is also an annual benchmark revision. Once a year, the BLS lines up its payroll estimates with near-complete counts from unemployment insurance tax records. The BLS says that over the past 10 years the average absolute benchmark revision to total non-farm jobs was 0.2%. A preliminary estimate of that revision is published separately, and it can move markets too.

Number 3: the unemployment rate

This comes from the household survey. It is the share of people in the labour force who are out of work and looking for a job.

  • A rising unemployment rate (for example 4.1% to 4.3%) points to a weakening job market, which can bring rate cuts closer, usually good for gold.
  • A falling rate points the other way.

Watch the participation rate too. If unemployment falls only because people stopped looking for work, that is not real strength.

Number 4: average hourly earnings (wages)

Wages show how much pay is rising. Fast wage growth can keep inflation high, which keeps the Fed cautious about cutting rates. Traders look at:

  • The monthly change (often forecast around +0.2% to +0.3%).
  • The yearly change.

A big beat in wages can push gold down even when payrolls are soft.

Reading the whole report in 60 seconds

Check Stronger than expected (usually bearish gold) Weaker than expected (usually bullish gold)
Payrolls vs forecast Beat Miss
Revisions Up Down
Unemployment rate Lower Higher
Hourly earnings Hotter Cooler

If all four point the same way, the first move is often clear and can last. If they are mixed (strong payrolls, rising unemployment, soft wages), the first move often reverses as traders read past the headline. That is a common pattern on gold.

How gold has reacted: 26 reports measured

We measured the hourly candle that contains the release for 26 jobs reports from 5 July 2024 to 6 November 2026 (COMEX gold futures, GC=F, hourly via Yahoo Finance).

Measure Value
Average high-to-low range of the release hour $42.0
Median $31.0
Largest $104.5 (4 September 2026)
Average range of the hours before the release $12.7

Recent release hours:

Release date Release-hour range
5 Jun 2026 $72.7
2 Jul 2026 $80.8
7 Aug 2026 $71.3
4 Sep 2026 $104.5
2 Oct 2026 $54.6
6 Nov 2026 $41.6

Our own measurement, run 17 November 2026. Past moves do not predict future moves.

On 0.10 lot (usually 10 ounces), a $42 range is a $420 swing inside one hour. Spreads also widen around the release, and stop orders can fill at worse prices. See why gold spreads widen on news.

For the most recent report, payrolls rose 165,000 against about 65,000 expected, and gold spiked and then dropped within minutes; the full minute-by-minute story is in the October 2026 NFP gold reaction.

A step-by-step routine for NFP day

  1. The day before: note the forecasts for payrolls, unemployment and wages. Note the last two months’ numbers so you can see revisions.
  2. Check the time in your zone: 8:30 am New York is 13:30 GMT in winter. Our NFP time today page converts it.
  3. Decide your plan before the release: trade after the first 15–30 minutes, or not at all. Many traders close or reduce positions before 8:30.
  4. At release: read the four numbers in order. Do not trade the first spike on the headline alone.
  5. After 15–30 minutes: see whether the move held. If the reaction fits the whole report, look for a pullback entry in that direction with a normal-sized stop.
  6. Remember the weekend: most reports land on a Friday, so gold may gap on Sunday night.

Our full guide to trading gold on NFP day and the gold and NFP tool go further.

Common mistakes

  • Reading only the headline. Revisions and wages often decide the second move.
  • Forgetting the forecast. The surprise moves gold, not the raw number.
  • Trading the first second. Spreads are widest and slippage is worst right at 8:30.
  • Using full size. A normal stop can be jumped in a $40+ hour. Halve the size or stay out.
  • Ignoring the date. Some reports are not on the first Friday. Check the BLS schedule.

This page is education, not financial advice. Trading gold with leverage carries a high risk of loss.

Frequently asked questions

What is the difference between the household and establishment survey?

The establishment survey (CES) asks about 119,000 businesses and government agencies, covering about 622,000 worksites, how many people are on their payrolls. It gives the headline payrolls number and wages. The household survey (CPS) asks about 60,000 households about their work status. It gives the unemployment rate. The two can disagree in any month.

Why are NFP numbers revised?

The first estimate is based on the survey replies received in time. As more replies arrive, the Bureau of Labor Statistics revises the previous two months in each new report. Once a year, payrolls are also benchmarked to near-complete counts from unemployment insurance tax records.

What part of the jobs report matters most for gold?

The surprise matters most: payrolls, the unemployment rate and average hourly earnings compared with forecasts, plus revisions to earlier months. A stronger-than-expected report usually lifts yields and the dollar and weighs on gold; a weaker one usually helps gold. Wages matter because they feed inflation.

How much does gold move on NFP?

In our data for 26 jobs reports from July 2024 to November 2026, the hourly candle containing the release had an average high-to-low range of $42.0, with a median of $31.0 and a largest of $104.5 (4 September 2026). The hours before the release averaged $12.7.

What time is the NFP report released?

8:30 am New York time, usually on the first Friday of the month. In winter that is 13:30 GMT. Always check the official BLS schedule, because some months move, for example the December 2026 report is due on Thursday 8 January 2027.

Sources and further reading

External sites are independent of XAUUSD Global. Check dates and data yourself.

Comments

No comments yet.

Leave a comment

Comments are checked before they appear. Be respectful: no promotion, no trade signals. Rules · Privacy. No email needed.