The U.S. economy creates more jobs, yet gold falls. In another month, hiring again beats expectations, but both gold and the euro rise. These apparently contradictory reactions turn on the same question: what did the report change about the expected path of interest rates?
The short answer
The market impact of a jobs report depends on its surprise relative to expectations and the risk the Fed is focused on at the time. A stronger-than-expected report, supported by wages, can reinforce expectations that interest rates will stay higher for longer. That can support the dollar while weighing on dollar-denominated gold and EUR/USD. A weak report can bring expected rate cuts forward and work in the opposite direction.
This is not a trading rule that produces the same result every month. Rising unemployment can overshadow a strong payroll headline; safe-haven demand can support the dollar; another development can move gold. The first minute after publication and the price later that evening may also tell different stories.
The examples I selected from the ten calendar years from 2016 through 2025 illustrate this variation. Critical Fed decisions during Jerome Powell’s chairmanship point to the same conclusion: employment is an important input, but no single payroll number determines the policy rate.
Start by reading the number correctly
Nonfarm payrolls (NFP) measures the monthly change in payroll employment at U.S. nonfarm establishments. The unemployment rate comes from a separate household survey. One counts jobs, while the other measures people’s employment status, so payrolls and unemployment can rise in the same report. They are not directly interchangeable series. The BLS explanation describes this distinction and subsequent revisions.
The market’s first comparison is usually the release against a survey taken before publication. A gain of 300,000 jobs might sound strong. If the expectation was 400,000, however, it is a downside surprise of 100,000. If the expectation was 150,000, the same result becomes an upside surprise. These are illustrative numbers, not a historical month.
Next comes the report’s detail: were earlier months revised down, why did unemployment change, did wage growth accelerate, and what happened to hours worked? Using a subsequently revised payroll series as if it were the first number investors saw misreads the historical information available to the market.
Composition also matters for wages. In April 2020, heavy job losses among lower-paid workers pushed up average hourly earnings among those still employed. That did not mean the same workers received a pay raise of that size in one month. BLS explicitly identified the effect in the original report.
Two different routes to gold and the euro
An employment surprise can first change expectations about future monetary policy. Bond and currency markets do not need to wait for the next Fed meeting. Federal Reserve high-frequency FX research finds that economic announcement surprises can reach exchange rates such as EUR/USD very quickly. However, that study predates the decade examined here; I am not presenting its findings as a newly estimated coefficient for 2016–2025.
One important channel for gold is the expected real interest rate: the nominal rate less expected inflation. Gold pays no interest, so a higher expected real return on safe interest-bearing assets can raise the opportunity cost of holding it. A stronger dollar can also make gold more expensive for buyers using other currencies. Inflation concerns, uncertainty and safe-haven demand can work in the other direction. Chicago Fed research on gold emphasizes that the real-rate relationship varies over time and that an observed correlation does not establish causation.
For EUR/USD, the comparison is relative. The exchange rate states how many U.S. dollars one euro buys; a rise means euro appreciation against the dollar. If expected U.S. rates rise more than expected euro-area rates, that can favor the dollar. But the European Central Bank’s outlook and European data also matter. A rise in U.S. yields does not necessarily mean the expected rate differential widens by the same amount. The Fed’s analysis of relative policy expectations supports this distinction.
Gold and EUR/USD may therefore move together without behaving like the same asset. Gold has safe-haven demand and its own buyers; the euro reflects the relative economic and monetary outlook in the United States and Europe.
Employment may not be the only explanation
The biggest mistake is to assign every price change on a jobs day to NFP. Another central bank announcement, geopolitical development or wave of position unwinding can move prices that day. Even an in-line report can accompany a sharp move as earlier positions are unwound.
June 7, 2024 offers a concrete example. Strong payrolls and wages weighed on gold, but news that China had paused its gold purchases was another influence that day. Attributing the entire decline to U.S. employment would collapse two separate developments into one explanation. On November 4, 2022, payrolls beat expectations, yet higher unemployment and slower annual wage growth offered a different reading; gold and the euro rose.
A rise in nominal Treasury yields around publication does not, by itself, prove a real-rate channel either. The change in expected inflation, the maturity and other risks also matter. A same-day price move may be consistent with a mechanism without proving that mechanism was its only cause.
When should the interpretation change?
I would not keep predicting pressure on gold simply because the payroll headline was strong if expected U.S. real rates were falling. For the euro, an explanation based solely on U.S. data is also incomplete if U.S. and European rate expectations move together.
A reversal of the first move, large downward revisions to earlier months, or conflicting signals from wages and unemployment should prompt a fresh reading. If gold and the euro diverge, gold-specific demand or European developments deserve closer attention. These are alternatives to check alongside the report, not explanations invented afterward to accommodate any outcome.
Ten examples across ten years
The table selects one release from each calendar year. The date is the publication date; the month in parentheses is the period measured. Payroll figures are initial release / pre-release expectation, in thousands of jobs. The 2016 row uses a MarketWatch survey; the others use Reuters surveys.
Prices are spot observations reported at the time. Percentages are the changes reported in those stories, not returns measured over identical windows before and after publication. GMT times appear only when the source states them explicitly. An untimed observation should not be read as a closing price or a first-minute quote.
| Release and reference month | Initial NFP / expected | Spot gold and EUR/USD observation | Detail that changed the reading |
|---|---|---|---|
| January 8, 2016 (December 2015) | 292 / 215 | Gold was down 0.8% at $1,100.65 at 13:56 GMT. Late in the day, the euro was around $1.0922 versus $1.0928 previously. | Monthly hourly earnings fell despite the strong headline; the dollar’s initial bounce did not last. Data, gold, FX, survey. |
| February 3, 2017 (January) | 227 / 175 | Gold was little changed at $1,215.80 at 14:25 GMT after recovering earlier losses. The dollar fell against the euro. | Wage growth was modest and unemployment rose to 4.8%. The payroll headline was not decisive alone. Data, markets. |
| June 1, 2018 (May) | 223 / 188 | Gold was down 0.3% at $1,294.06 at 13:22 GMT; the dollar rose against the euro. | Unemployment fell to 3.8%. Payrolls and low unemployment sent a reinforcing signal of strength. Data, markets. |
| January 4, 2019 (December 2018) | 312 / 177 | Gold was down 0.8% at $1,283.37 at 16:24 GMT. A separate FX report had the euro down 0.34%. | Unemployment rose while payrolls and wages were strong; policy and risk developments also mattered that day. Data, gold, FX. |
| May 8, 2020 (April) | −20,500 / −22,000 | Gold was little changed at $1,718.31 at 14:34 GMT. An FX report had the euro slipping to $1.0832. | Losing 20.5 million jobs was disastrous, but fewer losses had occurred than expected. Wage composition was distorted. Data, gold, FX. |
| May 7, 2021 (April) | 266 / 978 | Gold rose 0.9% to $1,832.07 at 16:09 GMT. A separate market report had the euro up 0.85% at $1.2167. | A large downside surprise challenged the expected recovery pace and prospects for early tightening. Data, gold, FX. |
| November 4, 2022 (October) | 261 / 200 | The day’s Reuters report had gold up 3.1% at $1,680.33 and the euro up 2.1% at $0.9956. No exact quote time was stated. | Unemployment rose to 3.7% and annual wage growth slowed. Both assets rose despite the payroll beat. Data, markets. |
| June 2, 2023 (May) | 339 / 190 | The market report had gold down 1.5% at $1,948.11 and the euro down 0.5% at $1.0707. | Unemployment rose to 3.7%, but payrolls and upward revisions were strong. A debt-ceiling agreement was also in focus. Data, markets. |
| June 7, 2024 (May) | 272 / 185 | Gold was down about 3% at $2,304.54 at 17:57 GMT. An early FX report had the euro falling to $1.0828 from about $1.0897 before the data. | Wages were strong; China’s pause in gold buying added a separate headwind for bullion. Data, gold, FX. |
| September 5, 2025 (August) | 22 / 75 | Kitco reported spot gold up nearly 1% at $3,578.22. A separate Reuters report had the euro up 0.79% at $1.174425. Exact quote times were not exposed. | Unemployment reached 4.3% and the preceding two months were revised down in total. Rate-cut expectations strengthened. Data, gold, FX. |
I do not calculate a percentage of strong reports that made gold fall from this table. The cases are not representative of all releases, and their observation times differ. They do establish three useful distinctions: economically bad data can beat expectations, a strong headline can conceal weak details, and the first price reaction can change later in the day.
What balance did the Fed strike under Powell?
Jerome Powell chaired the Fed and the Federal Open Market Committee from February 5, 2018 through May 22, 2026. The dates below are selected Committee decisions under his chairmanship, not his personal decisions or the direct consequences of individual NFP releases. His chairmanship is also distinct from his continuing Board membership. The official Fed biography confirms the boundary. Kevin Warsh took over on May 22, 2026.
The target range is for the federal funds rate. A 25-basis-point move is 0.25 percentage point. The table uses decision dates; implementation can occur on a different day.
| Decision date | Target range: before → after | Employment and other considerations |
|---|---|---|
| March 21, 2018 | 1.25–1.50% → 1.50–1.75% | Strong hiring and low unemployment supported gradual tightening in an improving outlook; inflation was still below 2%. |
| December 19, 2018 | 2.00–2.25% → 2.25–2.50% | Employment was strong and inflation near target. The Fed continued monitoring global and financial developments. |
| July 31, 2019 | 2.25–2.50% → 2.00–2.25% | Despite a strong labor market, low inflation and risks to the global outlook prompted a cut. |
| March 3, 2020 | 1.50–1.75% → 1.00–1.25% | Pandemic risks prompted an emergency cut before the worst employment reports had arrived. |
| March 15, 2020 | 1.00–1.25% → 0.00–0.25% | A second emergency step addressed deteriorating activity and financial conditions. February employment still appeared strong. |
| November 3, 2021 | 0.00–0.25% → unchanged | As employment recovered, asset purchases were reduced. This was not a rate hike; inflation was elevated. |
| March 16, 2022 | 0.00–0.25% → 0.25–0.50% | A strong labor market accompanied the start of rate increases amid high inflation and energy pressure. |
| June 15, 2022 | 0.75–1.00% → 1.50–1.75% | High inflation was central to the 75bp hike; strong employment alone does not explain it. |
| July 26, 2023 | 5.00–5.25% → 5.25–5.50% | The cycle’s final hike: resilient employment, elevated inflation, and attention to cumulative tightening and credit conditions. |
| September 18, 2024 | 5.25–5.50% → 4.75–5.00% | Slower hiring and greater confidence in disinflation brought the risks to the two goals into better balance. |
| September 17, 2025 | 4.25–4.50% → 4.00–4.25% | Downside employment risks had increased. The Committee made a risk-management cut despite still-elevated inflation. |
| April 29, 2026 | 3.50–3.75% → unchanged | Job gains were low, but energy-driven inflation and Middle East uncertainty persisted. Rates stayed unchanged at Powell’s final chair press conference. |
The 2019 and 2020 decisions show that the Fed did not simply follow backward-looking employment numbers. Inflation dominated in 2022, while labor-market weakening received greater weight in 2024 and 2025. April 2026 is a reminder that low job growth does not bring cuts under every set of conditions.
The BIS September 2024 review also found that Treasury yields had become more sensitive to payroll surprises from 2023 onward. The U.S. two-year yield fell 26bp on the August 2, 2024 report day. This is a bond-market finding, not evidence of a fixed equivalent response in gold or the euro. A changing Fed priority is one possible explanation, not an established sole cause.
Read the next report with four questions
The four-question employment framework connects the headline to the price response:
- Where is the surprise? Compare the first release with the pre-release expectation and record revisions separately. Identify which survey supplied the expectation.
- Do the details support the headline? Read wages, unemployment, participation and hours together. Do not equate rising unemployment automatically with an identical amount of job destruction.
- How did policy expectations change? Follow the expected U.S. rate path and its difference from the euro area. For gold, preserve the distinction between nominal and real rates.
- Over what interval did the price move? Separate the initial response, subsequent hours and other news. If the move reverses, reassess the interpretation rather than force the first explanation to fit.
I explore the broader transmission from interest rates to asset prices in the bond yields and Nasdaq Guide. The lesson I take here is to follow the expectation that changed instead of predicting direction from one number. That is why NFP matters: it provides a snapshot of the economy today while potentially reshaping tomorrow’s expected interest rates.
Methodology and limitations
This article examines the historical mechanism through one selected release per year between January 1, 2016 and December 31, 2025. It is not an event study covering every monthly report or a trading backtest. Large surprises, different policy regimes, contrary responses and source quality influenced selection, so selection bias remains. The Powell timeline separately extends through his final chaired meeting in April 2026 and does not list every intervening meeting.
Initial payroll figures come from dated BLS archives; expectations and market observations come from contemporary reporting. Gold means the XAU/USD spot price per troy ounce and the euro means EUR/USD. Gold futures, funds and a dollar index are not substitutes. I did not calculate an average return by pooling observations with different clocks, and I preserved missing quote times. Precisely estimating how much a release caused a price to move would require prices from the same source over consistent windows and further work to separate other news.
Evidence cutoff: 2026-09-05T13:00:00+03:00
Not investment advice; for research and educational purposes.
Sources and further reading
- BLS: The employment report and its two surveys. Dated initial releases are linked in the ten-year table.
- Federal Reserve: Macroeconomic announcement surprises and high-frequency FX responses.
- Chicago Fed: What drives gold prices?.
- Federal Reserve: Relative policy expectations and the dollar.
- BIS: Market volatility in 2024 and greater sensitivity to employment news.
- Federal Reserve: Powell’s tenure, his final chair press conference on April 29, 2026. Original decisions are linked in the timeline.
