Nearly 40% is the number that matters, and it isn't a forecast.
That's the implied probability of a rate hike that futures reportedly carried into the August 4 meeting — a meeting where new Fed Chairman Kevin Warsh, per the Reuters account, "declined to provide any guidance" about what he thinks should happen with rates and "has been very light on the details of how he thinks about setting rate policy."
Sit with that pairing for a second. A 40% hike probability, and a chairman who won't tell you how he decides.
That gap is the whole story. The 40% isn't the market pricing an economic outcome. It's the market trying to guess an internal head-count it can no longer see.
The 40% isn't the market pricing an economic outcome. It's the market trying to guess an internal head-count it can no longer see.
Last week, we covered the hawkish hold — a live FOMC, a slim but real hike risk, held at 3.5% to 3.75%. This week the hold reportedly happened as expected. The news is what Warsh withheld: forward guidance, and soon a shorter calendar.
Start with what's actually knowable, because it's almost nothing. According to Reuters, three officials reportedly voted for a hike at last week's meeting. The committee is split, with — by the reporting going into the meeting — perhaps three or four of a dozen voting members ready to move now.
Under the old regime, you'd have a Summary of Economic Projections to render that divide into a number: a dot plot, a median, a distribution. You'd know the shape of the disagreement.
Now you don't.
So consider what a duration investor is left to price. Not "will inflation run hot," but "which faction wins the next internal vote, and I have no reaction function to constrain the guess."
The rate is set by a committee. The committee is split down the middle. And the one person who could publish a framework to anchor expectations has chosen not to.
This is the part to be clear about: it's a design choice, not an oversight.
The case for Warsh holding is coherent on its own terms. He doesn't personally buy the hike arguments; a hike would undercut the task forces he stood up; and raising rates puts him crosswise with the Trump administration. Fine — those are reasons to hold.
But holding is a decision. Refusing to explain the reaction function is a communication strategy, and a specific one.
The tell is in how his tenure has been read: Warsh's "tough talk" about inflation has already raised borrowing costs and built credibility, and the open question is whether he'll follow through with action. He is, in effect, trying to make rhetoric do the work of policy.
Opacity is load-bearing for that strategy. If you publish a reaction function, markets price the function and your words stop moving anything. If you stay vague, every utterance carries weight — including utterances that aren't yours.
In a speech, Jeff Schmid, the Kansas City Fed president — who does not have a vote this year — said the economy "appears to be performing well with the notable exception of inflation," that it "does not look like monetary policy is restrictive," and that "bringing inflation down to the Fed's 2 percent objective will require tighter policy."
A non-voter, talking about farm-country issues, calling for tighter policy. In a normal week, filed under regional-president color.
In this regime, it's a data point about the committee's true center of gravity. With no dots and no chair framework, a non-voter's speech becomes a probe into where the median actually sits — because the median is the only thing that matters and it's the one thing that's hidden.
FXStreet's Speechtracker scored Schmid at 7.3 out of 10 on hawkishness, above the 7.0 historical average, and flagged his framing of AI investment as an inflation driver. That's the machinery of a market reverse-engineering a vote count from tone, because the institution stopped publishing the count.
Note what Schmid did not say: when, or by how much. So even the probe is noisy. You get direction without magnitude, from a person who can't vote, filling a vacuum the chair created on purpose.
Multiply that across a dozen speakers and you have the mechanism. Every regional president becomes a market-moving Kremlinological signal, and the aggregate of those signals is what now sets the term premium — the extra yield investors demand for holding duration through uncertainty they can't resolve.
That premium isn't a view on inflation. It's compensation for being forced to guess a hidden vote over an unknown horizon. It is, structurally, noise priced as risk.
Now add the second change, the one that turns a communication problem into a structural one.
Warsh is reportedly weighing cutting regular FOMC meetings from eight to six a year — the New York Times' account, via four people familiar — with Bloomberg reporting the specific shape: six rate-setting meetings plus two focused on "broader economic issues." Eight has been the standard since the 1980s. Next year's eight dates are already posted on the Fed's website.
Fewer meetings sounds like less noise. It's the opposite.
Think about the interval. Eight meetings means roughly six-and-a-half weeks between decision points. Six means stretching that toward nine or ten.
Each interval is a window over which expectations have to drift with nothing official to re-anchor them. In the old world, that drift got reset every six weeks by a statement, a presser, and — quarterly — a fresh dot plot.
Take away the guidance and the dots, then lengthen the interval, and you've done two compounding things at once. You removed the anchors, and you widened the sea the ship drifts across between the times you're allowed to look.
A committee that can't agree and won't explain itself now leaves duration investors alone with that vacuum for longer stretches. And the only thing that fills a longer vacuum is more premium.
This is the precise bind the NYT framing gestures at — Warsh wanted regime change, but a reset might have to come first. His latest appearance set off a bout of volatility and left him, in their telling, under pressure to clear up mixed signals about his commitment to getting inflation down.
The volatility is the point. It's not a communications hiccup to be smoothed with a follow-up speech. It's the mechanical output of removing the instruments markets used to price a split committee.
You can hold rates and still destabilize the curve. That's the counterintuitive result here: the funds rate didn't move on August 4, but the amount of uncertainty markets must carry between decisions went up.
It becomes term-premium noise when the institution hides the split and then meets less often to reveal it.
Warsh's bet is that discretion buys him flexibility — that by refusing to commit to a reaction function, he keeps his options open and lets his words steer. What the 40% actually shows is the cost of that bet, borne by everyone pricing duration.
Flexibility for the chair. Guesswork for the market. And a calendar redesign that lengthens the exact windows in which that guesswork runs unchecked.