Gold pays no interest and no dividend. The main cost of holding it is the return you give up by not holding something that does pay — specifically, inflation-protected U.S. Treasury bonds (TIPS), whose yield is called the "real yield" because it already strips out expected inflation. When real yields rise, holding zero-yielding gold becomes more expensive in opportunity-cost terms; when they fall, that cost shrinks.
As of October 14, 2026, the 10-year real yield (ticker DFII10) sits at 2.84%, up 111 basis points (1.11 percentage points) over the past year and up 24 basis points in just the last month. That is the 95th percentile of its own one-year range — an unusually high level historically. MacroGold's regime classification currently labels this "restrictive." Over the same period, gold has fallen sharply from its highs. That is consistent with the textbook mechanism, but it is a pattern, not a proof of causation, and the correlation data below shows why it is easy to misread day to day.
If you check the daily correlation between gold and the 10-year real yield over the past 21 trading days, it reads -0.26 — a fairly loose relationship. Measured over 63 days it is -0.25. Taken at face value, that looks like the "real yield channel" barely exists. It doesn't mean that.
MacroGold's own data flags this explicitly: when you instead compare the size of moves over rolling 60-day windows, the relationship strengthens to -0.51. The explanation is a mismatch of scale — gold can swing several percent on a single day for reasons that have nothing to do with rates (a geopolitical headline, a dollar move, silver and mining-stock flows, which show daily correlations of 0.80 and 0.83 respectively with gold), while real yields usually grind by a few basis points per day. A loose daily correlation does not contradict the real-yield channel; it simply isn't built to detect it. This is also a direct answer to "why didn't gold react to the Fed today" — a single day's real-yield move is often too small to register against the day's noise, even when the multi-week trend is doing exactly what the mechanism predicts.
An FOMC (Federal Open Market Committee) decision is only market-moving to the extent it differs from what was already expected. Bond markets price in a rate path weeks in advance; by the meeting date, that path is mostly reflected in yields already. What moves real yields — and therefore gold's opportunity-cost calculus — is the gap between what the Fed delivers (and signals for future meetings) and what was priced in.
That gap can run in either direction. Right now, the effective federal funds rate (DFF) stands at 3.88% as of October 14, 2026, having actually risen 25 basis points over the past month and 25 basis points over the past three months — even as the unemployment rate has eased to 4.2% and the 10-year/2-year yield curve has flattened toward inversion (0.33%, down from 0.55% a year ago), both of which are usually read as late-cycle signals that would argue for eventual easing. That tension between a rising policy rate and softening labor data is itself a form of "expectations gap" that markets have to resolve — and it is precisely the kind of divergence that can make a single Fed meeting more consequential than the headline rate move suggests.
| What happens at the meeting | What it signals about the future path | Typical real-yield reaction |
|---|---|---|
| Rate decision matches consensus, forward guidance unchanged | No revision to the expected path | Limited — yields were already pricing this in |
| Dot plot (the Fed's own rate projections) shows fewer future cuts than markets expected | Path tightens relative to expectations | Real yields tend to rise |
| Dot plot shows more cuts, or faster, than expected | Path eases relative to expectations | Real yields tend to fall |
| Chair's press conference commentary shifts the tone without changing the dots | Can still move expectations on its own | Often the biggest single move of the day |
Read this table as a mechanism map, not a forecast: it describes the direction real yields have tended to move given a given type of surprise, based on the opportunity-cost relationship described above — it says nothing about what any specific upcoming meeting will deliver, and MacroGold's own backtests have found no measurable, repeatable trading edge in positioning around FOMC dates specifically. The practical takeaway for a reader following a meeting live is to watch the dot plot and the press conference tone, not just the rate number, and to watch where the 10-year real yield and the dollar index move in the following hour — those are the transmission variables, not the headline.
For readers who want to track upcoming FOMC dates and other scheduled macro releases alongside gold's price action, MacroGold maintains an economic calendar at macrogold.ca/calendar.
Fed decisions affect gold mainly by resetting the market's expected path for real interest rates; a meeting that simply confirms expectations tends to produce a muted reaction regardless of whether the rate itself was raised, cut, or held. The clearest current data point supporting this channel is the inverse relationship between gold and the 10-year real yield, which strengthens from -0.26 to -0.51 once you measure it over 60-day moves rather than single days — itself a reminder that short-term price action is a poor way to judge a multi-week mechanism.
What the data does not show is any reliable way to predict the size or direction of gold's reaction to a specific future meeting, nor any backtested edge in trading around FOMC dates. Correlation figures describe historical co-movement over the windows measured (21, 63, and 252 trading days as of October 16, 2026); they are not stable constants, and other drivers — the dollar index (-0.41 over 63 days), risk sentiment as measured by the VIX (-0.46 over 63 days), and flows into silver and mining equities — are often moving at the same time, which makes isolating the Fed's specific contribution to any single day's move difficult even after the fact. Nothing here describes a trade, a target, or an expected return; it describes a mechanism investors can watch for themselves using the data available at each meeting.
Real rates, the dollar, flows and positioning — a full analysis every Sunday, with the objection that attacks it.
See pricing