Update, October 9: The CME slide this post flagged has now gone through the line. October hike odds fell to 19.4% (hold 80.6%), below the 30% threshold, so the tracker's CME component flips from bullish to neutral and the DXY Bias Score slips to +4 from +5. The minutes themselves still did not move the score. The rest of this post is left as originally published, so its +5 references were accurate when written.
The September FOMC minutes landed on October 7, and the headline takeaway is easy to state: most Fed officials expect another rate hike by year end. The harder part is knowing how much weight to give that, because the minutes describe a meeting held three weeks ago — before the August PCE print, before CME hike odds slid, and before September's jobs report missed by a wide margin.
What the minutes confirm
If the September 16 statement read as hawkish, the discussion behind it was more so. Every participant backed the 25bps hike to 3.75–4.00%, and the committee vote was unanimous at 12–0. Beyond the vote itself, a few details stand out:
- The risk balance shifted. Almost all participants saw inflation risks tilted to the upside, while risks to the labor market had diminished and were now seen as broadly balanced.
- More is expected. Most participants said another increase would likely be appropriate by year end, while stressing that every meeting stays open and depends on incoming data. That matches the 4.1% median dot in the September projections.
- Policy isn't seen as tight. Several participants described the current rate as not restrictive, or only mildly so, and a couple said they had raised their estimate of the neutral rate.
- Expectations are a worry. Some participants raised the concern that after more than five years of inflation above 2%, elevated inflation could start to affect expectations and wage- and price-setting.
Fed staff also nudged their own inflation forecasts higher for 2026 through 2028 compared with July, and don't see inflation back at 2% until 2029.
The PCE detail that connects to our earlier correction
The minutes also put a number on something we flagged last week. The BEA's methodology change to PCE was already on the committee's radar: staff estimated that under the new approach, August headline PCE inflation would be 3.6% instead of 3.8%, and core 3.2% instead of 3.4% — roughly 0.2 points off each.
The actual August print then came in at 3.4% headline and 3.0% core, about 0.2 points below even that methodology-adjusted estimate. So the cooling in the PCE data was partly mechanical and partly a genuine downside surprise — a fairer read than either "it's all methodology" or "inflation is falling fast." (We corrected our own PCE write-up on the tracker to reflect exactly this.)
A hike doesn't automatically mean a strong dollar
One passage that gets less attention: the minutes note the trade-weighted dollar actually depreciated over the weeks before the meeting. The reasons given were narrowing interest-rate differentials and improving foreign growth, plus the joint US–Japan intervention to support the yen in late July, which pushed the dollar lower directly because of the yen's weight in currency indexes. The staff also noted that other central banks were turning hawkish, with the ECB raising rates.
That is a useful reminder for anyone trading the dollar off Fed headlines alone: what matters is the Fed's path relative to everyone else's, not the Fed's path in isolation.
Why the minutes are already partly stale
Minutes are a lagging document by design, and this set is a clear example. Here is what has happened since the meeting they describe:
| Development | Date | Effect |
|---|---|---|
| Williams speech: another hike "late this year," but no urgency | Sep 29 | CME October hike odds 71.4% → 47.1% |
| August PCE cools (3.4% headline, 3.0% core) | Sep 30 | CME hike odds → 37.1% |
| September NFP: +29K vs 84K forecast, prior months revised down 60K | Oct 2 | DXY Bias Score +8 → +5 |
| FOMC minutes released | Oct 7 | Describes a meeting before all of the above |
Read together, the committee's own view (one more hike by year end) and the market's view (October increasingly unlikely) are not necessarily in conflict. They may simply be pointing at different dates. A hike by year end can still happen in December even if October is off the table, which is why we think the live debate has shifted from whether to when.
Does this change the DXY Bias Score?
No. The score is built only from voted decisions and released data; commentary — speeches, minutes, press conferences — is deliberately kept out so sentiment can't distort it. The minutes are commentary, so the score holds at +5, Bullish DXY.
One mechanical heads-up for the next cycle: the dot-plot component of the score is designed to go neutral at meetings that don't publish new projections. October 27–28 is one of those meetings, so that component is built to drop out automatically, all else equal, regardless of what the committee decides.
What to watch next
- October 14 — CPI and October 15 — PPI. Two chances for the inflation data to either support or undercut the committee's year-end hike signal.
- CME FedWatch vs. the 30% line. October hike odds are already down to 37.1%; the tracker's CME driver turns neutral below 30%.
- October 27–28 — FOMC. The first meeting since the minutes' hike-by-year-end signal, and the first test of whether the weak jobs data changed anyone's mind.
The takeaway
Treat minutes as a window into the committee's reaction function, not as fresh news. They tell you how officials were thinking when conditions looked a certain way. Since then, conditions have moved — and the useful question isn't "did the minutes say hike?" but "what would it take for the committee to still want that hike once the newer data is in?" That kind of reading, data to score to chart, is exactly what the October YouTube series is built to walk through as each release lands.