Which News Really Moves Gold? 204 NFP Releases and Fed Decisions Measured, 2015–2026
- We measured how much more gold moves in the hour of 112 US jobs reports (NFP) and 92 Fed rate decisions (FOMC) than in the same hour on normal days.
- FOMC is bigger and more reliable: a median 5.0× the normal hour, and 97.8% of meetings moved gold more than 2×.
- NFP is strong but unreliable: a median 2.4×, and in 16% of releases gold barely reacted (less than 1.5×).
- Chasing the first post-news candle: over the next 2 hours price continued in the same direction only 45% (NFP) and 52% (FOMC) of the time. A coin flip.
Why measure instead of trusting the news calendar
Every economic calendar marks dozens of releases a month as "high impact", but none tells you how much gold actually reacts, or whether it reacts consistently. For manual traders and EA users, the useful questions are:
- How many times more does price move in the release hour?
- Does it move every time, or only sometimes?
- After the first candle, does price keep going the same way?
This article answers those three questions for the two releases gold traders watch most: US Non-Farm Payrolls (NFP) and the Federal Open Market Committee (FOMC) rate decision.
Method
| Prices | Hourly XAUUSD in UTC, 2015 – Sep 2026 |
| FOMC | Decision dates taken directly from the Federal Reserve website, 92 meetings · announced 14:00 New York (the 14:30 press conference falls in the same hour) |
| NFP | Release dates calculated from the official Bureau of Labor Statistics (BLS) rule, 112 releases · published 08:30 New York |
| Baseline | Same weekday and hour on days with neither release, e.g. NFP vs an ordinary Friday at the same hour |
| Measure | High–low range of the 1-hour candle as a share of price, so years with very different gold prices can be compared |
We confirmed the price data is in UTC: on NFP days the most volatile hour is 12:00 UTC in summer and 13:00 UTC in winter, exactly 08:30 New York.
For NFP we excluded January, days next to US holidays, and days when the gold market was closed (such as Good Friday), because release dates in those periods often don't follow the normal rule. We deliberately did not use "days when price moved a lot" to guess release dates, because that would exaggerate the result.
Result 1: FOMC moves gold almost twice as much as NFP
| NFP (112) | FOMC (92) | |
|---|---|---|
| Median range vs normal hour | 2.4× | 5.0× |
| Releases with more than 2× | 63.4% | 97.8% |
| Releases with more than 3× | 35.7% | 85.9% |
| Gold barely reacted (under 1.5×) | 16.1% | 0% |
This runs against the common belief that NFP is the biggest event of the month. In the data, Fed decisions move gold more, and more reliably. Not a single meeting moved gold less than 1.5× normal, while in about 1 in 6 NFP releases gold moved barely more than on an ordinary Friday.
The biggest FOMC hour in the data was 17 June 2026, with gold moving $110.6 in one hour (12.7× normal), followed by 16 March 2016 (12.5×) and 21 September 2022 (10×). The biggest NFP was 2 October 2015 (8.1×).
Result 2: what that means in dollars today
Converted to dollars for 2025–2026, when gold is both expensive and volatile:
| Time slot (2025–2026) | Normal hour | Release hour (median) |
|---|---|---|
| Friday 08:30 New York (NFP) | $18.9 | $39.5 |
| Wednesday 14:00 New York (FOMC) | $10.7 | $38.0 |
Holding 0.10 lot through the release, a $38–40 swing in one hour is $380–400 per trade, and that is the median: half the releases were bigger. (See What Lot Size for a $500 Gold Account? for position sizing.)
Result 3: chasing the first candle is a coin flip
A popular approach is "see which way price jumps on the news, then follow it". So we measured how often, after the first release-hour candle closed, the next 2 hours went the same way:
NFP continued 45% of the time and FOMC 52%, both indistinguishable from a coin flip. The direction of the first candle says nothing about what comes next. Often price spikes one way and then reverses hard, catching chasers on both sides.
Also note that spreads usually widen sharply around releases and orders may fill away from the requested price (slippage), so the real cost of trading then is higher than the chart suggests.
How to use this
- Take FOMC more seriously than you might. If you avoid only one event, the data says it should be the Fed decision: 8 times a year, Wednesday afternoon New York time.
- NFP: brace for impact, but don't expect it every time. It is released on a Friday early in the month at 08:30 New York.
- Don't chase the first candle after the news. Its direction tells you nothing about the next two hours.
- Reduce size or avoid new trades before the release. A 3–5× normal swing in one hour easily takes out a stop set for a normal day, and without a stop the damage can be far beyond what you can afford.
- If you use an EA, check how it handles news, for example pausing new trades or closing positions before major releases. An EA that trades straight through news is taking risk that a normal backtest doesn't show. See How to Read an EA Backtest Report.
Limitations
- This covers NFP and FOMC only. Other releases such as inflation (CPI) can also move gold hard, but we don't yet have a fully verifiable history of release dates, so we have not published unverified numbers.
- We used hourly candles, so this measures the whole hour, not just the first minutes after the release.
- NFP dates are calculated from the BLS rule, not taken day by day from a historical calendar. We excluded months where the rule tends to be off (January, holiday periods), but a few months may still differ from the actual release date. That would make the NFP figures understate the reaction, not overstate it.
- 2025–2026 has few events (12 NFP, 13 FOMC), so figures for that period are noisier than the overall result.
This article is for educational purposes and is not investment advice. Trading gold and CFDs carries a high level of risk and you may lose all of your capital. Past performance does not guarantee future results.
