Hawkish FOMC Minutes Point To Another Rate Hike

I kept refreshing the release page longer than I care to admit. Not because I expected a surprise cut, and not because anyone still thinks the committee is in a hurry to ease. I wanted to see whether the room was still unified after a hike that plenty of traders had already tried to talk themselves out of. It was. Every official backed the move. Most of them still think another increase before the calendar turns is the sensible call. That combination is rarer than the headlines make it sound, and it changes how I read the next six weeks.

The September record landed with a distinctly firm tone. The target range for the federal funds rate was lifted a quarter point to 3.75 to 4.00 percent, and the discussion around that decision did not sound like a reluctant compromise. Officials offered different reasons, yet they landed in the same place. Inflation is still too high. Growth has not cracked. And the risk, in their view, still leans toward prices staying stubborn rather than cooling on schedule.

What The September Record Actually Settled

Start with the vote, because the vote is the cleanest part. All 19 participants supported the increase. That is not a split decision dressed up as consensus. It is a committee that, whatever private doubts exist about the path after this, agreed the last step was warranted. A separate cluster of officials went further and said higher rates were required by their own economic outlook, not merely as insurance against a bad inflation surprise.

The line that markets circled first is simple enough to quote without dressing it up.

Most participants assessed that another increase in the target range for the federal funds rate would likely be appropriate by year end.

Most is not all. That gap matters, and I will come back to it. Still, “most” inside a room that just hiked in unison is a stronger signal than a vague promise to stay data dependent. Several participants also said the underlying momentum in the economy looked like it had picked up. When growth feels firmer and inflation refuses to behave, the bias inside that room tilts toward doing a bit more, not standing still and hoping.

I’ve found that investors often treat minutes as a replay of the statement. They are not. The statement is the compromise sentence. The minutes are the argument that produced it. This time the argument was about upside inflation risk, a labor market still described as close to full employment, and a policy rate that some officials do not even view as clearly restrictive.

Why Every Official Could Agree On The Hike

Unanimity does not mean identical reasoning. The record is explicit on that. Participants offered a range of views for why they supported the increase. Some framed it as protection against inflation pressures that could intensify. Others said the move followed directly from how they see growth, spending, and prices evolving even in a baseline case. Both camps could sign the same decision.

That distinction is easy to skip and expensive to ignore. An insurance hike can be reversed if the scare fades. A hike justified by the central outlook is stickier. If a decent share of the room believes the economy itself calls for a higher setting, then a soft patch in one data print is less likely to unwind the bias. Perhaps the most interesting aspect of this release is how little daylight there was between those two stories on the day of the vote.

Chair comments after the decision described the move as removing a dose of accommodation while inflation stayed stubbornly high. That phrasing is deliberate. Accommodation is what you take away when policy is still doing some of the work of supporting demand. It is not the language of a committee that believes it is already deep into restrictive territory.

The Inflation Picture Officials Would Not Soften

August headline PCE inflation was estimated around 3.8 percent, with core PCE near 3.4 percent. Those are not crisis numbers. They are also not “job done” numbers for a target that still sits at 2 percent. Higher energy costs, geopolitical tension, and the investment wave tied to artificial intelligence all showed up in the discussion as contributors to price pressure.

Participants generally viewed inflation risks as skewed to the upside. Almost all of them saw that tilt. A smaller group thought the skew had become more pronounced in recent months. Job-market risks, by contrast, were described as broadly balanced. When one side of the mandate looks lopsided and the other looks roughly even, the policy conversation stops being symmetrical. You do not get poetic language about a soft landing in a record like that. You get a preference for staying a step tighter than markets would prefer.

• Headline inflation still well above the long-run goal, with energy doing real work in the latest readings
• Core inflation cooler than the headline, yet not cool enough to declare the disinflation trend secure
• Upside risks cited by nearly the entire room, not a vocal minority
• Labor risks judged broadly balanced, with unemployment near 4.1 percent
• Several officials arguing the current setting is not restrictive, or only mildly so

Core goods prices also remained elevated, according to several participants, as effects linked to the AI buildout appeared to increase while earlier tariff effects faded. That handoff matters. If one temporary source of goods inflation is easing and another demand-heavy source is taking its place, the committee cannot simply wait for last year’s shock to roll off the year-ago comparisons.

Growth That Refused To Cooperate With A Pause

The staff outlook prepared for this meeting was stronger than the one prepared for July. That single comparison explains a lot of the tone. GDP growth was described as solid. Consumer spending held up. Business investment kept getting a lift from the AI buildout. Labor conditions were viewed as close to full employment.

Resilience is a compliment until you are the person setting interest rates against an inflation overshoot. A household sector that keeps spending, and a corporate sector that keeps ordering equipment, chips, power gear, and data-center capacity, does not hand the committee an easy excuse to stop. Several participants said the scale and pace of that buildout had continued to surprise to the upside. Surprises of that kind are not neutral for policy. They raise the odds that demand runs ahead of what the supply side can comfortably deliver.

In my experience, committees talk themselves into patience when the real economy is clearly losing altitude. This record does not read that way. It reads like a group that looked at momentum, looked at prices, and decided the patient approach had already been tried.

How AI Showed Up As An Inflation Story, Not Just A Growth Story

The buildout discussion is the part of these minutes that feels newest, and it is worth sitting with. Officials did not treat the investment boom as a free productivity gift that would quietly solve the inflation problem. Some of them did note the productivity prospect. Several also flagged the demand side.

Some participants said the buildout could cause aggregate demand to outpace aggregate supply over the medium term, putting upward pressure on inflation. Others pointed to cost pressures faced by businesses, including higher costs for transportation and input materials, with energy prices and the buildout both in the mix. Strong demand for skilled workers in related sectors was cited as a driver of strong wage gains for those workers. Market commentary, relayed in the discussion of yields, also pointed to heavy private debt issuance financing AI infrastructure as competition for capital.

So the same phenomenon is doing three jobs at once. It lifts investment. It can lift productivity later. And right now it soaks up power, materials, specialized labor, and financing. A central bank that only celebrates the first two effects is going to be late. This committee, at least in September, was not willing to be late.

Several participants observed that the rate of price increases in the core goods category also remained elevated, as effects of the AI buildout appeared to increase while the effects of tariff increases waned.

I would not treat that sentence as a forecast that chips and servers will drive the entire CPI. I would treat it as evidence that officials are watching a live demand impulse inside the goods and investment complex, not a fading one. If you are marking a portfolio to a “hikes are done” story, that is an awkward sentence to explain away.

Yields, Term Premium, And A Market That Still Looked Easy

Longer-term Treasury yields had climbed into the meeting. Nominal yields were up around 35 basis points across the two-year to ten-year stretch of the curve. Part of that move reflected a higher expected policy path and stronger data. Another part, in the market commentary officials heard, reflected geopolitics, uncertainty around the Treasury buyback program’s announcement and implementation, and competition for capital from private issuance tied to AI infrastructure.

Changes in real rates accounted for most of the net increase in longer-maturity yields. That is the kind of decomposition that sounds technical and is actually plain. Investors were demanding more real compensation, not merely pricing a burst of near-term inflation. A few participants noted that the Treasury market had been functioning smoothly, while still flagging the need to plan for stress. Smooth is good. It is not a promise.

Here is the twist that keeps the minutes hawkish even after a backup in yields. Many participants said that despite the climb in longer-term rates, financial conditions still looked supportive of growth. Equity prices had risen substantially this year. Corporate bond spreads had stayed narrow. If the market is doing some of the tightening for you, you can sometimes afford to wait. If equities are firm and credit is easy, a higher ten-year yield is not doing the whole job.

A useful way to read that table is as a checklist rather than a forecast. None of the rows, on their own, forces a hike in October. Together they make a pause harder to justify unless the data break.

The Currency Desk Note That Markets Should Not Ignore

Buried near the operational discussion was a point about the joint U.S.–Japan operation to support the yen in late July. Officials noted that the intervention contributed directly to dollar depreciation, given the yen’s weight in currency indexes. The Desk acted purely as fiscal agent for the Treasury, using Treasury funds. The System Open Market Account portfolio was not involved.

Why mention it in a rates piece? Because currency operations and rate decisions get mashed together in market chatter whenever the dollar swings. This record draws a line. Support for the yen was a Treasury operation, executed by the Desk in its fiscal-agent role. It was not a monetary-policy easing smuggled through the balance sheet. Anyone treating that episode as a quiet signal about the funds rate is reading a different document than the one that was published.

Politics At The Edge, Policy Still In The Middle

The hike drew public criticism from the White House, with the president blaming colleagues around the chair and calling the decision very political. I am not going to litigate that claim here. What the minutes show is a discussion organized around inflation estimates, growth momentum, financial conditions, and the stance of policy. You can dislike the outcome. The internal reasoning on offer is economic, not electoral.

Since the meeting, a run of comments from senior officials has suggested the committee may be in no rush to move again immediately. Vice Chair Philip Jefferson and New York Fed President John Williams, in separate remarks, indicated there is time to assess the economy before considering another increase. Investors noticed. Pricing for a quarter-point move at the October 27–28 meeting fell from roughly 70 percent in the days after the September decision to something closer to 20 percent. Two-year yields, the segment most sensitive to the near-term policy path, dropped more than 10 basis points over the past week, toward 4.8 percent.

Those speeches do not erase the minutes. They sit beside them. A record that says most participants thought another increase by year end would likely be appropriate can coexist with a preference to see the next inflation report before locking in October. Year end is not October. The window includes the December meeting. Traders who collapsed “another hike” into “hike in four weeks” were always making an extra assumption.

October Versus December Is The Real Argument Now

Some officials, including three who voted against holding steady in July, could dissent again in favor of an increase if the majority chooses to leave the rate unchanged in October. That is the friction point. The minutes already recorded a view, from several participants, that the current policy rate is not restrictive or only mildly restrictive. If that group still holds the view after the next data cluster, a hold will not be quiet.

Will they get the votes? I do not know, and anyone who speaks with certainty is selling something. What I do know is the burden of proof has shifted. After a unanimous hike and a “most participants” line on another move by year end, the case for skipping October has to be made with evidence, not with fatigue. The consumer price report due October 14 is the obvious candidate. A soft print cools the room. A firm print puts the dissent risk back on the front page.

1. Watch whether incoming inflation data confirm the upside skew officials already see, or chip away at it.
2. Separate October odds from year-end odds. The minutes speak more clearly to the second.
3. Track real yields and credit spreads together. Higher bond yields with still-narrow spreads are not full tightening.
4. Listen for whether “mildly restrictive” language spreads or fades in public remarks.
5. Treat a possible October dissent as information about 2026’s last meeting, not as theater.

Short version: the market has talked itself into patience faster than the minutes justify. That can be rational if the data cooperate. It is not free.

A Closer Look At The Restrictive-Or-Not Debate

Policy debates often stall on a single adjective. Restrictive is that adjective. If the funds rate is clearly above a neutral setting, holding steady can still slow the economy. If it is only mildly restrictive, or not restrictive at all, holding steady is closer to standing aside while demand keeps rolling.

Several participants placed the current rate in that milder camp. I read that as a warning against assuming 4 percent is automatically tight just because it is higher than it was two years ago. Neutral moves. It moves with potential growth, with fiscal impulse, with the appetite for risk, and with an investment boom that raises the return on capital in whole sectors. A rate that felt tight against a sluggish pre-buildout economy can feel ordinary against an economy that is ordering power plants and server halls at a pace that keeps surprising the staff.

There is a household version of the same point. Mortgage rates and card rates bite. They always do. But a committee that looks at aggregate demand, unemployment at 4.1 percent, and inflation still north of 3 percent is not setting policy for the most rate-sensitive borrower alone. It is setting policy for an economy that, on the evidence in this record, still has momentum.

Simple stance check from the minutes:
  Inflation risks ........ skewed higher
  Job risks .............. broadly balanced
  Growth ................. solid, momentum up
  Conditions ............. still supportive
  Year-end bias .......... another hike, most say

None of that is a formula. It is a snapshot of how the room was leaning when it voted. Snapshots age. They do not become irrelevant the week a couple of officials say they can wait.

Energy, Geopolitics, And The Part Models Keep Missing

Higher energy costs were not a footnote. They showed up in the inflation estimates and in the cost pressures businesses are facing. Geopolitical tension sat in the same paragraph. When oil and freight move for reasons that have nothing to do with domestic demand, a central bank cannot fine-tune them away with a quarter point. It can decide whether to accommodate the shock or lean against the second-round effects.

This record leans against accommodation. That is a judgment call, and reasonable people disagree with it. The risk of leaning against an energy spike is that you slow an economy that was only passing a price through. The risk of looking through it is that households and firms reset their expectations and the spike stops being a spike. Officials here sound more worried about the second risk. Given where core inflation already sits, I understand the instinct even when I do not love the growth trade-off.

Geopolitics also showed up in the yield discussion, as one reason term premium might have risen. A world that keeps handing markets new supply shocks is a world where longer-term rates can gap without the committee changing a word. Planning for Treasury-market stress, which a few participants flagged, belongs in that same file. Smooth functioning today is not a strategy.

What “Most” Leaves Open For The Dissenters And The Doubters

Words in minutes are negotiated. “Most” is one of the negotiated ones. It tells you a majority leaned toward another increase by year end. It also tells you a minority did not, or at least would not sign that sentence. The release does not name them. It does not need to. The existence of the minority is the option value in a hold.

If October data soften the upside skew, that minority can become a majority without anyone having to recant the September vote. The hike already happened. The next one is a forecast, not a promise. Forecasts inside this institution get revised when the staff outlook gets revised. The staff outlook was stronger than July’s. It can be weaker than September’s by December. That is how this process works, and pretending otherwise is how people get hurt on two-year notes.

Still, I would not fade the majority lightly. A room that just voted 19 to nothing, with most of its members on record favoring another step, does not flip because a futures contract repriced. It flips because inflation, jobs, or financial conditions change the story it is telling itself.

How A Portfolio Might Take This Without Getting Cute

I am not going to pretend a blog note is an allocation. The practical read is narrower. Front-end rates had gotten ahead of the minutes in the days after the decision, then gave some of that back when senior officials emphasized time to assess. Both moves can be true. The level that matters is whether two-year yields near 4.8 percent adequately pay you for a committee whose base case, as of September, still included another hike before year end.

Equities are a different problem. The minutes themselves say prices have risen substantially and that conditions look supportive. Supportive conditions are good for risk assets until they become the reason policy stays tight. If the buildout is both the earnings story and the inflation story, the same theme can help the index and hurt the multiple. That tension does not resolve in one afternoon.

Credit spreads remaining narrow tell a similar tale. The corporate market is not pricing a break. Officials noticed. A central bank that believes conditions are easy will be less impressed by a backup in government yields than equity bulls hope. If you need a single sentence for a risk meeting, use that one.

Near-term path check: October odds fell hard, year-end bias in the minutes did not.

Cash and short bills still have a role while that gap is open. So does humility about the exact meeting. The expensive error, in my view, is building a book that only works if the committee has secretly abandoned the assessment it just published.

The Labor Market Is Not Offering A Rescue

Unemployment at 4.1 percent, described alongside a job market near full employment, is not a distress signal. Risks on that side were broadly balanced. Some wage pressure was tied specifically to skilled roles in the buildout, which is a composition story as much as a cycle story. Composition stories can still leak. When a hot sector pulls pay higher, neighboring sectors notice, and service inflation is where those notices show up.

Could the labor picture cool without a formal break? Sure. Participation, hours, and quits can do quiet work. The minutes do not describe that cooling as the thing already underway at a pace that would let the committee declare victory. They describe a market that is tight enough to remove one common excuse for patience.

I keep coming back to the balance of risks because it is the actual decision rule, even when officials avoid saying so in public. Almost all saw inflation risks tilted up. Job risks were broadly balanced. You do not need a speechwriter to know which way that points while both are true.

What Would Actually Change The Room

Three developments would make the “most participants” line look dated. A clear downshift in core inflation that is broader than one category. A labor market that stops looking balanced and starts looking soft, with unemployment rising for the wrong reasons. Or financial conditions that tighten for real, meaning equities and credit join the bond market instead of offsetting it.

Anything short of that leaves the September assessment intact. A single cool CPI print helps. It does not retire the upside skew officials said had, if anything, become more pronounced. A soaring stock market after a hawkish record does the opposite of the committee’s work. People forget that. Conditions are an input, not a trophy.

There is also the awkward middle case, which is the one I think is most plausible. Inflation edges down but stays above a pace the room will bless. Growth stays decent. The buildout keeps surprising. In that world the committee can skip October, talk about lags, and still deliver the increase most participants already judged appropriate before the year is out. Markets that price only the skip are pricing the press conference, not the minutes.

Reading The Staff Upgrade Without Overreading It

Staff forecasts are not votes. They do shape the questions officials ask. A stronger outlook than the one prepared for July means the baseline walking into September already had more growth in it, and the committee still chose to hike. That sequence is the opposite of a reluctant tightening into a deteriorating forecast. It is tightening into an upgraded one.

Upgrades get revised. I have watched enough cycles to know a staff can take growth down as quickly as it put it up, especially if energy prices reverse or the investment wave pauses to digest. The point is narrower. At the moment of the decision, the institutional forecast was not flashing a stall. Officials who wanted to wait for weakness did not have weakness in front of them.

Several participants commenting that economic momentum appeared to have increased fits the same picture. Momentum is a slippery word. In a minutes context it usually means the recent run of spending and investment no longer looks like a fade. If that judgment holds through the autumn data, the year-end hike most of them already lean toward becomes the path of least internal resistance.

A Plain-Language Map Of The Next Few Weeks

Strip the jargon and the map is short. The committee hiked. It did so together. Most members think one more step before year end still fits the outlook. Inflation risks point up. The job market is not crying out for relief. Markets have cooled their October bets after a pair of cautious speeches, and the next inflation report can heat those bets up again. Yields did some tightening. Equities and credit did not finish the job.

If you want a metaphor that is less tired than “higher for longer,” try this one. The committee took a step up a staircase and most of the people on it said the next step still looks warranted before they reach the landing. A couple of them have since said they are willing to pause on the stair and look around. Pausing on a stair is not the same as deciding the staircase was a mistake.

That is the whole release, once you stop treating it as a script for the next press conference. The script can change. The staircase comment is already in the record.

Questions Worth Asking Before The October Meeting

Is 3.75 to 4.00 percent doing enough if several officials say it is barely restrictive? Does a 35 basis point backup in intermediate yields count as tightness when spreads are narrow and equities are up on the year? Can the buildout raise productivity fast enough to offset the demand it is creating, or is that a 2027 story being used to excuse a 2026 problem? And if most of the room already answered the year-end question in September, what exactly has to break for that answer to be withdrawn?

I do not have clean answers. I have a preference for not inventing them. The honest position is that the minutes raised the bar for a full stop and lowered the bar for one more increase sometime before year end. October is optional inside that frame. The increase is not, unless the data move.

One last personal note, because these releases reward it. I went in expecting a hawkish tint and a few careful caveats. The caveats are there. The tint is darker than the post-meeting speeches imply. If you only trade the speeches, you will keep getting surprised by the documents. If you only trade the documents, you will miss the timing. The useful habit is to hold both, and to let the inflation print on the 14th tell you which one the room is about to trust.

Stronger growth, elevated energy prices, and persistent inflation were judged to warrant a more restrictive stance, with most members expecting another rate increase before year end.

Synthesis of the September policy discussion

That is the line I would keep on the desk. Not as a promise. As the hurdle the next six weeks of data have to clear if the committee is going to walk away from it.