By the time you read this, the Federal Open Market Committee will be in the room.
Futures on 15 September 2026 put an 85% chance on a 25-basis-point hike, about 8% on 50 basis points, and about 7% on a pause. Money markets were nearer 93% on a hike. That is not a debate about direction. It is a debate about the dots, the language, and whether Chair Kevin Warsh treats this as a one-and-done or the start of a new string.
The last hike in this cycle was a long time ago. The last six weeks were not. Warsh’s Jackson Hole line was tighter than other FOMC speakers. He put the employment side of the mandate behind price stability. He talked about higher-than-expected inflation. Core PCE printed 3.7% in August against 2.7% in June. Two-year yields sat near 4.10%. The 10-year tagged 5.04% on 15 September. That is the rate that competes with gold.
The Market Ear’s 16 September preview — the long FOMC note that desks passed around overnight — is the map this piece follows. It is not a crystal ball. It is a list of what the committee can break.
What the Committee Can Say
The funds-rate target is widely expected to move to 3.75%–4.00%. After August, the median 2026 dot was already a 3.4% year-end funds rate. That implied two more quarter-point moves if the first one lands this week. Reuters had 85% of economists expecting a hike. About 53% expected the next hike in December. About 15% still wanted a 50-basis-point step. That last camp is small. It is not empty. Warsh has sounded closer to it than the rest of the table.
Oxford Economics, in that same preview, argued the hawkish case should not be dismissed. It called a third consecutive moderate inflation print a live risk that financial conditions are not tight enough. Officials, Oxford said, do not typically react to every market wiggle when inflation is already too high.
Goldman Sachs went the other way on choreography. The bank expects a hike and a careful statement. It does not expect a signal of more hikes this week. It thinks the communication risk is that the chair sounds more hawkish than the committee. That gap is how you get a “policy error of all errors” tape: hike into a labour market that is cooling while inflation is still sticky, then talk as if the job is unfinished.
Stephen Miran, who used to sit near this debate from the White House side, put four tests in public: core PCE at the low end of recent prints; a labour market that is not rolling over; the idea that June–July hawkishness was a mid-cycle pause, not a peak; and a warning that “control the long end with words” is a weak theory. Whether you buy those tests or not, the bond market already moved. The long end is the referee.
How a Hike Hits Gold
Gold does not vote at the FOMC. It lives with the real yield that comes out of the room.
Spot has been working near $4,270–$4,300 after a run above $4,600. Tony Kim, running commodities at a large U.S. house in the same preview pack, said he did not know if gold would reprint its old high before year-end or a year later. He did say $4,000 looked like a floor. Central-bank buying and emerging-market official demand were the long bid. The Strait of Hormuz was the short fuse. A mine in the water lane does not care about a dot plot. A 5% 10-year does.
The usual textbook says a hike and a hawkish statement lift real yields and knock bullion. That textbook has been late for two years because official buyers kept absorbing metal while Western funds chopped. It can still be right for a week. A 5.04% 10-year is already a test. A Warsh press conference that sounds like two more hikes is a harder test. A statement that hikes and then talks about data-dependence is a milder one.
Gold mining stocks feel the same hour twice. Once when the metal drops. Once when the equity tape treats them like the S&P. August’s miner index bounce of about 33% was a debasement month. This week is a yield week. J.P. Morgan’s 15 September miner dashboard still called valuations a “room to run” case at spot. That case assumes the metal holds. It does not assume a 50-basis-point surprise and a 5.3% 10-year.
Physical gold and a miner are not the same asset. One sits in a vault. The other sits on a diesel invoice. Do not use one to hedge the other on FOMC day.
Copper and Oil Are in the Same Room
Copper held near $14,000 a tonne after a record near $14,736 on 8 September. LME warehouse stocks jumped about 20% in 30 days. The curve flipped toward contango. That is the inventory story. The Fed story is demand. A hike that firms the dollar makes a tonne more expensive for every buyer outside the United States. China just printed stronger industrial output and weak consumption. That split does not love a stronger greenback.
Crude is the other fuse. The same preview pack had Saudi east-west pipeline damage still in the tape, Iranian-linked risk in the Strait, and prices that had been living in a $105–$108 world. Diesel at that setting is an AISC problem for miners even if gold holds $4,200. A hike that cools growth can ease oil. A hike that does not ease the Strait cannot.
The Scenarios That Matter for a Metals Book
J.P. Morgan’s rates desk laid out a grid that metals desks should steal.
Hike 25, no guidance: the consensus. Yields can drift. The dollar can fade. Gold can breathe.
Hike 25, hawkish dots: two more 2026 dots stay on the page. The 10-year stays bid. Gold tests the $4,000 handle Kim called a floor. Miners gap first.
Hike 50: a minority. Equities and metals both eat glass. That is the 8% tail, not the base.
No hike: the 7% tail. If inflation data keep coming in hot, the market will not treat a skip as a gift. It will treat it as a delay. Gold can rally the hour and sell the next statement.
None of those paths is a recommendation. They are the ways a metals portfolio gets marked.
What Canadian Readers Should Do With This
Do not size a new gold or copper equity on the eve of the dots. The names you already research — Agnico, Barrick, Wheaton, Teck, Hudbay, the rest of the liquid Canadian book — will trade the headline, then the press conference, then the SEP table. That is three prints. Liquidity is not a shield. It is a faster drop.
If you own metal for title, the vault does not need a forecast. If you own miners for torque, you need a view on AISC and on whether $4,000 gold is a floor or a trapdoor. This page will not pick that for you.
Conclusion
The Fed is likely to hike. The dots will tell you if the committee thinks the job is starting or finishing. Core PCE at 3.7% and a 5% 10-year are the facts gold has to live with this week. A $4,000 floor is one desk’s line, not a law. Copper’s warehouse relief and oil’s pipeline scar are the industrial side of the same hour. Read the statement. Then read the dots. Then decide if you wanted a bar or a stock.
Disclaimer
This article draws on The Market Ear / ZeroHedge FOMC preview dated 16 September 2026 (Tyler Durden compilation), including cited views from Oxford Economics, Goldman Sachs, J.P. Morgan scenario work, and commodity comments attributed to Tony Kim and others in that pack. Odds, yields, and prices move by the session. Company names are examples for research, not recommendations. This is not investment advice. Speak with a licensed adviser. The author and publisher accept no liability for actions taken on this article.