What a Fed rate hike means for gold signals (September 2026 edition)
By Emanuele Calcina · Published 12 September 2026 · Sources linked inline · Written ahead of the 16 September FOMC decision
About 90% is the market's number for a Fed hike on Wednesday 16 September, per the CME FedWatch reading quoted by GoldSilver after Friday's CPI print. Two weeks earlier it was a coin flip. If you follow gold signals, paid or free, that shift changes how the next few days should be traded, and it is worth being precise about what it changes and what it does not.
This is the September 2026 edition of a question that comes up every hiking cycle. The short answer: a hike hurts gold on the way in and often stops hurting it the moment it lands. The longer answer is below, with the two cycles worth remembering, the mechanism that actually drives the move, and the specific things that change in our signals on a Fed day.
Where the Fed is, in one paragraph
The target range is 3.50% to 3.75%. At the July meeting the committee voted 9 to 3 to hold, with Hammack, Kashkari and Logan dissenting in favour of a quarter-point hike, the first three-way dissent in one direction since 2016 (CNBC). The minutes of that meeting said officials saw a need to hike if inflation did not cool (CNBC). August CPI then printed 3.4% year on year with core up 0.3% on the month, above the 0.2% forecast (CNBC), and PPI the day before ran 5.4% year on year (GoldSilver). Chair Warsh used Jackson Hole to keep a hike on the table. The ECB has already hiked twice in three months (Euronews). So the Fed is not leading here. It is catching up.
The mechanism: real yields, not headlines
Gold pays no interest. The cost of holding it is whatever you could have earned in a safe asset instead, after inflation. That number is the real yield, and the cleanest version of it is the yield on 10-year inflation-protected Treasuries, which stood at 2.43% on 8 September with the nominal 10-year near 4.80% (GoldSilver). By Friday the nominal 10-year had touched 4.979%, its highest since late 2023 (Quartz).
A rate hike matters for gold only through that channel. If the Fed hikes and the real yield rises, gold gets more expensive to hold and tends to fall. If the Fed hikes and the market decides the hike is the last one, long yields can fall on the day, the real yield drops, and gold rises through a hike. That is not a paradox. It is the whole game. The decision is a headline; the yield curve is the trade.
Friday was a live demonstration. CPI landed, hike odds went to 90%, the 10-year touched 5%, and gold finished the New York session up on the day after a $100 round trip in the first hour (Yahoo Finance GC=F hourly: 4333.0 low to 4432.6 high in the 8 AM ET bar). A market that rallies on the news it was supposed to fear has already priced that news. We take that seriously going into Wednesday.
Two cycles worth remembering
| Cycle | What the Fed did | What gold did | Lesson |
|---|---|---|---|
| Dec 2015 to Dec 2018 | First hike in nearly a decade on 16 Dec 2015, then 0.25% to 2.50% over 36 months | Printed $1,046 straight after the first hike, then rose through the cycle (BullionVault; Goldmoney) | The low was the first hike. Fear of hiking hurt more than hiking. |
| Mar 2022 to Nov 2022 | 375 basis points in eight months, four of them 75 bp moves | From 2,052 on 8 March to 1,629 on 3 November, close to 20% (event study) | A fast sequence of hikes with rising real yields is the case where gold keeps falling. |
| Sep 2026 | One 25 bp hike ~90% priced from 3.50% to 3.75%; dot plot decides whether it is a sequence | Inside a descending channel since February, ranged 4333 to 4489 last week | Which of the two histories this resembles depends on the dots, not the hike. |
The 2015 case is the one people forget. The Fed had been threatening to hike all year, gold fell all year, and the actual hike on 16 December was followed by the low of the entire bear market within days. Sellers were exhausted; there was nobody left to react. The 2022 case is the opposite: the Fed kept surprising on the hawkish side, real yields went from negative to positive in months, and gold had no floor until the market believed the sequence was ending.
Wednesday will look like one of those two, and the statement will not tell you which. The dot plot will. A median dot that shows one more hike and then a long hold is 2015. A median dot that shows a run of hikes into 2027 is 2022. We do not know which we get, and we would rather wait for the answer than pretend to have it.
Free channel, no card
Watch how the free channel trades Fed week before you decide anything
Join the Free Telegram Channel →1-2 XAUUSD signals a day · entry, stop loss and take profit on every call · 4K+ members
Three outcomes and what each usually does to XAUUSD
A hike with hawkish dots. Real yields rise, the dollar firms, gold sells into the press conference and often into the London open the next day. Inside the channel we have been trading since February, that is the path to a retest of 4300, and below 4300 the July base near 4000 is back in play (our 7 to 11 September setups). This is the outcome where rallies get sold.
A hike with a "we are done for now" message. The hike is in the price, the guidance is softer than feared, long yields ease, and gold usually prints a fast dip on the statement and then recovers through the presser. This is the outcome where dips get bought, but only once the recovery is visible on the chart, because the dip can be $60 before it turns.
A hold. The roughly 10% outcome. Everything that traded on the 90% reprices in a line of text. It would probably be the largest candle of the week, and the direction would be up for gold and down for the dollar. We treat it as possible and it is one reason we are not short into the statement.
What changes in the signals on Fed day
The format does not. Every signal in the VIP channel, around five a day, ships with an entry, a stop loss and a take profit in the message. The free channel's one or two calls a day are the same shape. Since 9 September every VIP entry also carries a two or three sentence note on why it is a buy or a sell, written from the structure on the daily, 4-hour and 1-hour charts and the levels in play. On Fed day that note is the most useful line in the message, because it tells you whether a long is a range trade to a known top or a breakout attempt, and those get managed differently when the statement drops.
Timing changes. There are no new entries in the window from the 2:00 PM ET statement through the end of the 2:30 press conference. Open positions get managed, and that is all. The first candle after the statement is not traded; the first move is the algorithms reading the text and it reverses more often than it extends.
Stop distance and size change together. If the expected range for the hour is $100, a stop $15 from entry is not a stop. The stop moves outside the level that would prove the idea wrong, and the position gets smaller to keep the dollar risk the same. That is the only risk rule the channel has ever had. It just becomes visible on days like this.
Bias can change, and we say so when it does. A daily close above 4650 flips the plan to buying pullbacks. A daily close under 4300 confirms the sellers. Until one of those prints, the map is the same channel we have published every week since it formed, and Wednesday is the event most likely to push price to one of its edges.
What a hike does not mean
It does not mean gold signals stop working. It means the ranges get wider for a few sessions and the stops have to respect that. It does not mean shorting gold is free money. The 2015 cycle punished that idea badly, and gold rallying on Friday into a 5% ten-year is a warning. And it does not mean the long-run picture changes: the People's Bank of China bought 20.2 tonnes in August, its largest monthly purchase since October 2023, and has now added gold for 22 straight months (Kitco). That buyer does not read the dot plot.
What a hike does mean, for anyone following signals, is that the next week is a test of the provider's process, not their win rate. Do they post through the event or go quiet? Do they show the stopped-out calls afterwards? Do their stops make sense for a $100 hour? The channel history is unedited, and the record, losing weeks included, is on the results page with the method on the numbers page. The lifetime figure across the tracked record is 84%, and you can watch how it holds up on Wednesday from the free channel before you spend anything.
VIP · around 5 signals a day
Fed week, with the reasoning under every entry
See the VIP plans →$149 a month or $349 a year · current promo $99 a month or $249 a year · cancel anytime to stop the next payment
Background reading on sessions, stops and how a signal is structured is in the guides hub.
Questions people ask
Does gold go down when the Fed raises rates?
Usually into the hike and often not after it. Gold printed $1,046 straight after the December 2015 hike and then rose through the 2016 to 2018 cycle. In 2022 it fell from 2,052 in March to 1,629 in November across 375 basis points of hikes. The mechanism is the real yield: when the yield on inflation-protected Treasuries rises, holding gold costs more. The direction after the decision depends on whether the market thinks the Fed is finished.
What is priced for the 16 September 2026 Fed meeting?
Roughly a 90% chance of a 25 basis point hike from the 3.50% to 3.75% range, per the CME FedWatch reading quoted by GoldSilver after the August CPI print. It is a projections meeting, so the dot plot arrives with the statement at 2:00 PM ET.
Do gold signals change on Fed day?
The format does not: entry, stop loss and take profit on every call, plus a note under every VIP entry on why it is a buy or a sell. Timing and size do. No new entries in the window from the statement through the press conference, wider stops with smaller positions when the expected range is bigger, and no chasing the first candle.
Why did gold rise after CPI if a hike is coming?
Because the hike was already in the price. Gold rallied on Friday 11 September even as the 10-year yield touched 4.979% and hike odds went to 90%. When a market rises on bad news, the bad news was expected. That is why a hike on Wednesday is less dangerous for gold than a hawkish dot plot.
Where can I check the signal results?
The results page lists the record week by week, losing weeks included, and the numbers page explains how each figure is counted. The lifetime figure across the tracked record is 84%. The free Telegram channel shows the live calls before you pay anything.
Related reading
- Gold week ahead, 14 to 18 Sep 2026: FOMC, BoE, BoJ and how we trade it
- Gold signals during high-impact news: our rules for CPI, NFP and FOMC days
- What Myfxbook-verified gold signals look like
- How gold signals perform in volatile markets
- All gold trading guides
Risk disclaimer. Trading gold on margin carries a high risk of loss and is not suitable for everyone. Past performance, including any figure on this page, does not predict future results. Nothing here is financial advice. Market data and third-party figures were read on the dates stated and may have been revised since.