Geopolitics & macro: If Warsh kills the dot plot tomorrow, is the market right to read it as hawkish?

2026-06-16 · long-form

Executive summary

Kevin Warsh chairs his first Federal Reserve meeting today and tomorrow, and the rate decision is the least interesting thing about it. Markets price a 98.4% chance the Fed holds at 3.50%–3.75% on Wednesday T3. The live event is the document, not the rate. Warsh has spent a decade attacking forward guidance as a straitjacket, and the reporting into the meeting is that he may withhold his own rate projection, strip the easing bias from the statement, or decline to publish the dot plot at all T3.

The question this report answers: if Warsh removes the dot plot tomorrow, is the market right to read that as a hawkish signal? Our answer is no. Removing the dots changes the instrument the market uses to read the path, not the path itself. The path is an extended hold — the kit's standing call, and one the curve has spent the past week converging toward. What dot removal does is mechanical. It pulls out the anchor that pins the long end of the curve to the expected short-rate path, at the exact moment a positive term premium is back and doing the work in the long end T3.

So a back-end selloff on Wednesday is the obvious risk, and it will be read as "hawkish Fed." That read confuses a term-premium shock with a reaction-function shift. The clean tell is the shape of the move: a selloff concentrated in the 10-year and 30-year with the 2-year and the funds-rate futures roughly stable is a term-premium event and the hold call is intact. A 2-year selloff with funds futures repricing toward an actual 2026 hike is a reaction-function event and would mean we are wrong.

House view reconciliation

The current house view, last reviewed 2026-06-15 PM, holds the US rate path at roughly 98% priced for a hold at 3.50%–3.75% and carries the kit's extended-hold variant — registered in the 2026-06-09 Tuesday long-form — at "2-1 up." The 2026-06-14 AM note already refined the discriminating event: resolution was reframed from "the dot plot and the presser" to "the statement bias language and the presser, with the dot plot's existence itself in question," after reporting that Warsh may eliminate the dots and strip the bias language T3.

This report confirms and extends that line. It does not change the path call. It adds the layer the house view has named but not worked through: what the removal of the dot plot does to the market's ability to read the path, and how to score the variant when the instrument used to score it may not be published.

Two falsifiers from the 2026-06-09 variant came up since the last long-form, and both went the kit's way. May core CPI printed at 2.9% year-over-year, at consensus rather than above it, so the "core surprise above consensus" falsifier did not trigger T1. University of Michigan five-year inflation expectations fell to 3.4% in the June preliminary from 3.9%, the opposite of the unanchoring the second falsifier watches for T1. The path side of the variant is being vindicated. The update this run is to register that the live debate has moved from direction to communication regime, and to add the curve-decomposition tell as the Wednesday scoring rule.

The setup

Three things changed in the week since the last Tuesday long-form, and together they move the question.

The hike-pricing drained. On 2026-06-05 the May jobs report flipped the tape toward a hike, and prediction markets briefly priced roughly 70% odds of a Fed hike by year-end T3. That has unwound. CME FedWatch now shows the December meeting as close to a coin flip between a hold and a single 25 basis-point cut — 42.7% for rates staying at 3.50%–3.75% through December, slightly ahead of one cut T3. The 2026 distribution the kit pushed back against last week — hold-versus-hike — has become hold-versus-cut. The market converged toward the extended-hold call from the hawkish side.

The energy shock is reversing. The US-Iran agreement is now electronically signed, with Trump authorizing removal of the naval blockade and a June 19 signing ceremony in Switzerland T3. Oil is pricing the reopening. Brent fell about 5% on Tuesday toward $79, the lowest since early March and a fourth consecutive losing session, as the prospect of restored Hormuz flows pulled the war premium out of the curve T3. The same energy spike that drove May headline CPI to 4.2% is now in retreat, which removes the one input that gave the hawks their number.

The rest of the world is tightening anyway. The Bank of Japan raised its policy rate 25 basis points to 1.00% this morning, its first hike since 1995, on a 7-1 vote T3. The European Central Bank raised its deposit rate 25 basis points to 2.25% on June 11, effective June 17, its first increase in nearly three years, unanimously, with Lagarde saying a further hike in July is very likely T1. The Fed sits inside a synchronized tightening cluster — but, as the analysis below argues, it sits at a different point in that cluster, and the dot plot is about to make its position harder to read.

The analysis

What the dot plot actually does, and what removing it does not

The dot plot is the chart of where each of the 19 Fed policymakers thinks the funds rate should sit at the end of the next few years. It was introduced in 2012 as a transparency tool. Warsh's objection is that it became a commitment device — markets read a set of conditional, individual forecasts as a firm collective promise, and policymakers then feel locked into a path the incoming data may not support T3. He flagged the Summary of Economic Projections as part of a broader overcommunication problem at his April confirmation hearing T3.

The key point for an investor is what removal does and does not touch. It does not change the reaction function. A chair who looks through an energy shock looks through it whether or not he prints a dot. It does not change the policy rate, which is held tomorrow regardless. What it changes is the market's model of the path. The dot plot is the bridge between the funds rate the Fed sets today and the 10-year yield investors hold. Remove the bridge and the long end has to be priced some other way.

That other way is the term premium — the extra yield investors demand to hold a long bond rather than roll short ones — and the term premium is no longer near zero.

Why this lands in a positive-term-premium regime, and why that matters

For most of the 2010s the term premium was negative or near zero. In that world the 10-year yield was mechanically pulled toward the expected average of future short rates, so the dot plot was a genuinely useful forecasting input: read the dots, get the path, get the 10-year. That regime is over. The term premium is back, driven by higher inflation uncertainty, a Fed less willing to deploy its balance sheet to suppress long yields, and the general uncertainty the Middle East conflict injected T3.

In a positive-term-premium regime the two components of a long yield — the expected rate path and the premium — can move in opposite directions. A 10-year yield can rise while the expected path is flat, because the premium is widening. This is the part the market is set up to misread. If Warsh removes the dots and the 10-year backs up, the reflex is to call it hawkish, as if the Fed signaled a higher path. But the path was held. What moved is the compensation investors demand for path uncertainty that the Fed just deliberately increased by removing its own guidance.

The precedent is instructive. In the 2013 taper tantrum the term premium rose roughly 100 basis points in four months even though the Fed had not actually tightened — the move was about the removal of an assumption, not a change in the policy rate T3. Warsh removing forward guidance is a similar setup: a repricing of uncertainty, not of the rate. Goldman's economists make the cost explicit in reverse — clear Fed communication is associated with lower borrowing costs and fewer market shocks, so withdrawing it raises both T3.

The reaction function still says hold, and the data since last week confirms it

Strip the communication question and the policy call is unchanged from the 2026-06-09 long-form. The energy shock that drove headline CPI to 4.2% is a relative-price supply shock, and standard doctrine is to look through it unless it unanchors expectations or feeds broad second-round inflation T2. May confirmed the look-through case. Core CPI was 2.9% year-over-year, and core commodity prices actually fell 0.1% on the month — the inflation was energy and little else, with the 1.3-point gap between the 4.2% headline and the 2.9% core measuring the oil spike directly T1. Expectations are anchored: five-year breakeven inflation sits at 2.53% and the University of Michigan five-year survey reading fell to 3.4% T1. And the shock is now mean-reverting in real time as the Iran deal reopens Hormuz and Brent falls toward $79 T3.

A chair who discounts the oil component, facing anchored expectations and a falling oil price, does not hike. The hold is the path. The dot plot's removal does not change that — it changes how loudly the market can hear it.

Why the synchronized tightening cluster is not a read-across

The Bank of Japan and the European Central Bank both hiked into the same energy shock this month, and the easy inference is that the Fed is the laggard about to follow. The inference is wrong on starting points. The ECB is normalizing from an accommodative 2.00% base toward 2.25%, catching up to tightening it deferred in April; a move to 2.25% is still below neutral, not restriction into a slowdown T1. The Bank of Japan is exiting a multi-decade emergency stance — a move from 0.75% to 1.00% is the unwinding of extraordinary accommodation, not a tightening cycle in the Western sense T3. The Fed alone is already restrictive, at 3.50%–3.75% with a firm dollar. Same shock, three different positions in the cycle, three different correct responses. The cluster confirms that nobody is cutting into an energy shock. It does not imply the restrictive central bank in the group hikes further.

Variant perception

Consensus framing, as it will be applied Wednesday afternoon: if Warsh removes the dot plot or strips the easing bias, that is a hawkish signal — the Fed is taking the implied cut off the table and refusing to commit to easing. A back-end selloff would be read as the bond market agreeing.

Our variant: dot removal is path-neutral and term-premium-positive. It does not move the reaction function, which still says extended hold. It raises the term premium by withdrawing the Fed's own anchor on the path at a moment when the premium is already the swing factor in long yields. The correct reading of a Wednesday back-end selloff is therefore a repricing of communication uncertainty, not a hawkish shift in policy. The market is positioned to misattribute a mechanical shock to a policy shock, and to mark the 10-year as a Fed signal when the Fed just removed the part of its signal the 10-year was reading.

What supports the variant: the curve has already drained the hike-pricing toward a December coin flip T3; core inflation and expectations confirm the look-through path; the term-premium regime is independently documented T3; and the 2013 analog shows long yields can move 100 basis points on an assumption change with no rate change.

What would falsify it — and the test is unusually clean because it is a curve-shape test available within minutes of the 2:00 p.m. release. If the selloff is concentrated in the 10-year and 30-year while the 2-year and the funds-rate futures hold roughly steady, the move is term premium and the variant holds. If instead the 2-year sells off hard and funds futures reprice toward an actual hike by year-end, the market is reading a genuine reaction-function shift, and the variant is wrong. A second falsifier is verbal: if Warsh in the press conference frames the removal as a response to inflation risk — "we will not pre-commit to cuts while inflation runs at 4%" — rather than as a structural communication reform, then he is using the change as a hawkish instrument and the consensus read is correct. The discriminating signal is whether the chair talks about the dots as a forecasting philosophy or as an inflation stance.

The asymmetry favors the variant. If we are right, any Wednesday back-end spike is a term-premium event that mean-reverts as the market re-learns to price the path without dots, and the front end stays anchored to the hold. If we are wrong, the 2-year and funds futures will show it the same afternoon, leaving the position re-checkable on a known schedule.

Implications for AlphaSteve

The top-down implication is that tomorrow's risk is a discount-rate signal-to-noise problem, not a policy-direction problem. For an equity book that discounts cash flows off the long end, the relevant change is that the 10-year is about to become a worse forecast of policy and policy a worse forecast of the 10-year, because the bridge between them is being removed. The kit should treat a Wednesday back-end move as term premium first and hawkishness second, and should resist marking long-duration positions off a 10-year spike that is mechanically driven rather than path-driven. The extended-hold base case is unchanged and, after the May core print and the oil reversal, better supported than it was a week ago.

• Portfolio: No position change triggered. The read argues against fading long-duration software or rate-sensitive financials on a Wednesday back-end selloff that is term-premium-driven rather than a real hike signal — wait for the 2-year to confirm before treating it as policy.
• Watchlist: Long-duration names whose multiples key off the 10-year should be screened against a "term premium up, path flat" scenario, not just a "Fed hawkish" scenario; the two have different half-lives.
• Theses on the workbench: Any thesis using the dot plot as a forward-rate input needs a new method. The dots may not be published, and even if they are, the positive-term-premium regime breaks the old "dots imply the 10-year" mapping. Switch the rate input to the funds-futures path for the front end and a separate term-premium assumption for the long end.
• Sectors: No sector-view shift this run. Continue tracking Brent and the Hormuz reopening as the live tell on whether the energy base effect rolls off into the back half.
• House view updates: Extend the Geopolitics & macro "US rate path" section to register the communication-regime layer — dot removal is path-neutral and term-premium-positive, not hawkish — and add the curve-decomposition scoring rule for Wednesday.
• Daily-scan adjustments: Add the 2-year/10-year decomposition of any Wednesday FOMC reaction as the score for the extended-hold variant, and add Warsh's framing of the dot change (philosophy vs. inflation stance) as the verbal falsifier.

Charts / data

Table 1 — The week's data drained the hike, not the hold

The hawkish case rested on the labor heat and the 4.2% headline. Over the week the curve drained the hike, oil reversed, and core stayed at consensus. The path debate moved the kit's way; the open question is now the communication regime.

Table 2 — How to score Wednesday's curve reaction

Sources

• T1 — headline +0.5% m/m and 4.2% y/y (highest since April 2023); core +0.2% m/m and 2.9% y/y; energy +3.9% m/m, +23.5% y/y; core commodities −0.1% m/m. https://www.bls.gov
• T1 — deposit rate +25bps to 2.25% effective 2026-06-17; first hike in nearly three years; unanimous; July hike "very likely." https://www.ecb.europa.eu
• T1 — 2.53%. https://fred.stlouisfed.org/series/T5YIE
• T1 — five-year inflation expectations 3.4%, down from 3.9%.
• T2 — look-through doctrine for relative-price supply shocks.
• T3 — Warsh may withhold his own projection; dot plot as commitment-device objection. https://www.cnbc.com
• T3 — BoJ +25bps to 1.00%, first since 1995, 7-1 vote. https://www.cnbc.com
• T3 — ~70% year-end hike odds at the post-jobs peak. https://www.cnbc.com
• T3 — 98.4% hold June 17; 42.7% hold-through-December, slightly ahead of one cut; JPMorgan (Feroli) remove-easing-bias view; EY-Parthenon (Daco) on SEP/dot symbolism; Goldman (Hatzius/Mericle) no major near-term change. https://www.indexbox.io
• T3 — positive-term-premium regime; dot-based 10-year forecasting model broke; components can move in opposite directions. https://ferrantecapitaladvisers.com
• T3 — front-end volatility rises as guidance is withdrawn; 2013 taper tantrum term premium +~100bps in four months with no tightening. https://soliswealth.com
• T3 — clear Fed communication linked to lower borrowing costs and fewer shocks. https://247wallst.com
• T3 — Brent ~$79, −5% on the day, lowest since early March, fourth consecutive losing session on Hormuz-reopening expectations. https://tradingeconomics.com
• T3 — US-Iran agreement signed; blockade removal authorized; June 19 Switzerland ceremony.
• T3 — reporting that Warsh may eliminate the dots and strip bias language.
• AS-cal
• See sources-policy for the citation discipline applied.

House view changes this run

• Geopolitics & macro → US rate path: Extend (no weight change; ~98% hold at 3.50%–3.75% priced for Wednesday). Register the communication-regime layer on top of the existing extended-hold variant: removal of the dot plot is path-neutral and term-premium-positive, not hawkish. A Wednesday back-end selloff should be read as a term-premium repricing on withdrawn guidance, not a reaction-function shift, unless the 2-year and funds-rate futures confirm a genuine hike repricing. The May core print (2.9%, at consensus) and the UMich five-year reading (3.4%, falling) both went the variant's way; the path side of the extended-hold call is better supported than a week ago.
• Daily-scan: Add the 2-year/10-year decomposition of any FOMC-day reaction as the variant's scoring rule, and add Warsh's framing of the dot change — forecasting philosophy versus inflation stance — as the verbal falsifier.
• No other house-view section changed this run.

Linked

• _house-view
• sources-policy
• voice-and-style
• _long-form-template
• 2026-06-09-warsh-reaction-function-hike-mispricing
• 05-Environment/rates-and-discount-rates
• 05-Environment/geopolitics-country-risk