Two Markets, One Fed: Bonds Have Fully Priced Warsh's Hike; Equities Say It Won't Happen

One Fed, Two Consensuses

There are two markets looking at the same central bank right now, and they cannot both be right.

The first is the rates market. It thinks the Federal Reserve is about to raise interest rates. The Atlanta Fed's own Market Probability Tracker, built from CME three-month SOFR options, puts the probability that the funds rate sits above the current target range in the September window at 76.06 percent. [8] Fed funds futures give only a 19.6% chance that rates are still at 3.50-3.75% by the December 9 meeting. [9] That is roughly 80% odds of a higher rate by year-end. [9] The two-year Treasury already yields 4.26 percent — fifty-one basis points above the top of the target range. [10]

The second is the equity market, or more precisely the multi-asset community that owns the equity risk. Bank of America's Global Fund Manager Survey, fielded July 2-9, 2026 — after the dot plot, after the repricing — found that 83% of respondents do not expect the Fed to raise rates before the November midterm elections. [15] That is a 17% probability on the same event to which the futures strip assigns roughly two-thirds. [15] Not a difference of degree. A difference of kind.

The hypothesis we set out to test was that a discount-rate regime change is under way and nobody is positioned for it. Half of that is right and half of it is backwards. The regime signal is real, datable and enormous: on January 2 the same Atlanta Fed tracker showed a hike probability of 5.74 percent against a cut probability of 34.58 percent. [8] Inside six months the market has gone from pricing cuts to pricing a hike. But the claim that nobody is positioned is false in the instrument that prices it. The strip has priced it, the cash two-year has already paid for it, and leveraged funds are net short 2,872,406 SOFR three-month contracts. [7] The people who are not positioned are the ones holding the stocks.

So the dislocation is not between the market and the Fed. It is between two markets — and it is dangerous in both directions, because there is a second correction the hawks have not absorbed. Every hawkish dot in the June projections was cast on June 16-17. This morning, July 14, June CPI printed core inflation at negative 0.02% month-on-month. [12]

What Actually Happened on June 17 — and Why the Chair Won't Own It

Kevin Warsh took the oath of office as Chairman and a member of the Board of Governors on Friday, May 22, 2026, and the FOMC unanimously selected him as its chairman the same day. [18] He has chaired exactly one meeting.

At that meeting, the policy rate did not move. "The Committee decided to maintain the target range for the federal funds rate at 3-1/2 to 3-3/4 percent," the June 17 statement read. [1] That is the same range Jerome Powell's final meeting set on April 29. [6] What changed was not the level but the stated direction of travel — and the change is stark. In April the Fed still carried an easing bias, and a governor dissented in favour of a cut: Stephen I. Miran "preferred to lower the target range for the federal funds rate by 1/4 percentage point at this meeting," while Beth M. Hammack, Neel Kashkari and Lorie K. Logan dissented because they did not support inclusion of an easing bias at all. Four dissents. [6] Seven weeks later the June statement was approved by a 12 – 0 vote, with no dissents and — breaking with decades of practice — no voter names listed. [1]

The statement itself was cut to the bone. It drops April's "extent and timing of additional adjustments" language, carries no forward guidance whatsoever, and closes with a flat, unprecedented pledge: "The Committee will deliver price stability." [1] [6] Forward guidance is not merely absent; it has been formally abolished — "guidance—which we agreed was not well suited to the current policy conjuncture" — alongside the announcement of task forces in five areas: Fed communications, the balance sheet, data sources, productivity and jobs, and the Fed's inflation frameworks. [5]

The dot plot is the story

What moved the market was the Summary of Economic Projections. The median 2026 federal funds projection came in at 3.8 percent, up from 3.4 in March, against a current target-range midpoint of 3.625 percent. [3] Warsh confirmed it at the press conference: "The median participant judges that the appropriate federal funds rate to be at 3.8 percent at the end of this year and 3.6 at the end of next." [5] The Fed's own median path implies higher rates by year-end.

The distribution is where the teeth are. Eighteen participants submitted: one at 4.375, five at 4.125, three at 3.875, eight at 3.625, one at 3.375. [3] Nine of eighteen sit above the 3.625 midpoint; eight sit at it; one sits below. Six of them — the five at 4.125 and the one at 4.375 — are projecting a year-end midpoint two or more quarter-points above today's, which is to say two hikes or more. [3] The dispersion moved with the median: the 2026 central tendency is now 3.6–4.1 percent, against a March central tendency of 3.1–3.6. [3] In March, the top of the entire range of participant views was roughly where the midpoint sits today. In June, the bottom of the central tendency is at it.

The market read this exactly as it was written. CME-implied odds of at least one hike in 2026 jumped to 84% on FOMC day itself, up from 59.5% a day earlier — a 24.5-percentage-point move in a single session, driven by the dot plot and not by the statement, which changed nothing about the rate. [16] The S&P 500 erased a 280-point Dow gain to close down 91.25 points, or 1.2%, at 7,420.10. [16] Bloomberg's Michael McKee put the reading on the record, to Warsh's face: "with the dot plot, nine members suggested that they want a rate increase by the end of the year, and the markets have taken that as forward guidance." [5]

And that repricing was not a straight line. The nearest-window hike probability went 5.74 in January, down to 2.97 in early March, up to 40.17 by March 20, back down to 6.38 on May 1, then 18.83, 44.91, 63.42 and finally 76.06 on July 10. [8] The hike trade was on and off twice before the June dots landed. It took the dots to make it stick — and the cut side of the same window is now effectively extinguished, at 1.07 percent. [8]

The chairman is not one of the hawks

Here is the twist that almost nobody has priced. The median of 3.8 is not any participant's dot. With eighteen submissions, the median is the average of the ninth and tenth ordered dots — which is to say the arithmetic midpoint between a hold at 3.625 and a hike to 3.875. The Committee is split exactly nine-nine. One defector kills the median. [3]

And the missing nineteenth submission is the chairman's. Only eighteen participants submitted projections in June, against nineteen in March. [3] Warsh, asked about it: "I did not submit a, a dot. For me, it's not helpful in the conduct of policy." [5] In his opening remarks: "I, however, have refrained from offering any projections of my own, consistent with my long-held views on the SEP, at least as currently structured." [5]

He then actively deprecated his own committee's dots on the day they landed:

"when I saw the submissions, I noted that all the submissions were coming in with pencils, you know those kinds with the big erasers... they didn't feel bound by them six weeks from now or six days from now... So I didn't hear tons of conviction. What I heard was the kind of humility that I think we should have." [5]

And when the FT's Claire Jones asked him point-blank why, given the inflation risks he had just described, he had not raised rates that day, he refused to answer: "my answer to your first question was to be very curt. I've got nothing more to say than the statement itself." [5] Asked whether the two-year yield was signalling more tightening was needed, he declined again: "I'm not going to offer any commentary on market reaction over the last 30 or 60 minutes." [5]

The minutes, released July 8, confirm the hike was live in the room but not on the table: "A few participants commented that, in light of these developments, there was a case for raising the target range for the federal funds rate, but those participants indicated that they supported maintaining the current target range at this meeting." [4] The split is in the prose as well as the dots — "many participants" saw the year-end rate within or slightly below the current range; "Many other participants, however, assessed that the appropriate level of the federal funds rate would be above the current target range at the end of this year." [4]

The market is pricing a hike off nine anonymous forecasts that the chairman explicitly refused to join, from a committee that voted 12-0 to do nothing. It is pricing a hike off the chairman's silence.

The Hike Is Already Paid For

The load-bearing claim of the bullish-on-hawkishness thesis is that this is an unpriced discount-rate shock. There are exactly three places a shock like that could be hiding: the futures strip, the cash curve, and speculative positioning. It is in all three. It is not hiding anywhere.

One: the strip

Take the meetings in order, as fed funds futures priced them this morning, after the CPI. July 29: a hold at 3.50-3.75 is 85.6% and a hike to 3.75-4.00 just 14.4% — although that hike bucket stood at 41.0% the previous day and 27.7% a week ago, so this morning's soft print gutted the July hike specifically. [9] September 16 is where the hike lives: 3.75-4.00 at 50.5% and 4.00-4.25 at 7.3%, against 42.2% for a hold. [9] Sum the buckets above the current range and September's cumulative hike odds are 57.8%. The hike, not the hold, is still the base case for September even after the soft CPI.

By October 28 — the last meeting before the midterms — only 31.5% of the distribution has rates still at 3.50-3.75; 48.4% sits at 3.75-4.00, 18.2% at 4.00-4.25 and 1.8% at 4.25-4.50. [9] And then the decisive number: by December 9, the market prices only a 19.6% chance that rates are still at 3.50-3.75. The rest is 42.0% at 3.75-4.00, 29.6% at 4.00-4.25, 8.0% at 4.25-4.50 and 0.7% at 4.50-4.75. [9] A cut is priced at essentially zero.

Note the direction of drift, because it matters. The previous-day column for December showed a 10.0% hold probability. [9] Before this morning's CPI the market was more hawkish than it is now, not less. Even after the print, the curve moved from "July or September" to "September or later" — not from "hike" to "hold."

The options market says the same thing from a different angle. For the December window, SOFR options imply a 76.16 percent hike probability against 6.87 percent for a cut, with a mean expected rate of 406.47 basis points — nearly two quarter-point hikes above the 3.625 percent midpoint where the funds rate sits today. [8] Nor is it priced as a one-off: the March-2027 window still carries a 75.35 percent hike probability against only 12.99 percent for a cut. [8] The market expects the higher rate to still be there nine months out. There is even a real two-hike tail: an 18.91 percent probability that the December window lands in the 425bps–450bps bucket, precisely the outcome the six most hawkish dots describe. [8] [3]

This is a base case, not a blind spot.

Two: the cash market has already done the hiking

The two-year Treasury yields 4.26 percent as of 07/13/2026 — above the 3-3/4 percent top of the target range. [10] That is fifty-one basis points of hiking already embedded in the instrument whose entire job is to price the Fed. And it has been a steady climb: the two-year bottomed at 3.38 percent on 02/27/2026, coincident with the trough in SOFR-implied hike odds, and has risen eighty-eight basis points since. [10] [8] The FOMC-day move is visible in the cash curve itself: the two-year went from 4.05 on 06/16 to 4.20 on 06/17 — fifteen basis points on the dots alone. [10] The wire reported it on close-to-close pricing: "The two-year Treasury yield, which more closely tracks expectations for Fed action, jumped to 4.21% from 4.05%." [16]

Meanwhile the actual policy rate has not moved at all. EFFR printed 3.62 percent on 2026-07-13, sitting in the lower half of its range; SOFR is 3.6 percent. [22] All of the movement is in expectations. None of it is in the rate. A reader who acts today on "the hike is coming" is not front-running a discount-rate shift. They are buying something that already moved, on June 18.

And it is not just the front end. The long end has moved further: the 30-year is at 5.10 and the 20-year at 5.11 on 07/13/2026, against 4.86 and 4.81 on 01/02/2026; the 10-year is 4.62 versus 4.19. [10] But — and this is the tell — what has not been priced is a demand-driven inflation problem. The 5-year real (TIPS) yield is 2.06 percent and the 10-year real yield 2.36 percent, against nominals of 4.37 and 4.62. [11] [10] That puts the 5-year breakeven at roughly 2.31pp — marginally above the 2 percent target. The bond market is pricing hikes and anchored inflation simultaneously. That is a policy reaction function being priced, not an inflation scare. And the real yield — not the funds rate — is the discount rate that prices equities. It has moved a long way already.

Three: positioning

Per CFTC Commitments of Traders as of July 7, 2026, Leveraged Funds in CME three-month SOFR futures — open interest 13,110,035 — hold 1,046,636 long against 3,919,042 short. [7] That is 29.9 percent of open interest short against 8.0 percent long, a skew of 21.9 percentage points, and a net short of 2,872,406 contracts. Short SOFR futures is the hike trade, and it is enormous.

The same picture holds in the cleanest instrument available. CBOT 30-day fed funds futures, open interest 2,325,148: Leveraged Funds hold 400,383 long against 673,553 short — 17.2 percent of open interest long against 29.0 percent short, a net short of 273,170 contracts. [7] Fed funds futures carry no cash-futures basis trade. There is nothing to hedge with them. That is a directional view on the Fed, and it is short.

And the fast money is trimming, not adding: leveraged funds reduced SOFR-3M shorts by 87,568 contracts in the week to July 7. [7] This is a crowded, mature trade being modestly covered — the definition of a position that is fully on, not one waiting to be put on.

The adversarial check on our own evidence

There is a fourth number that looks like it corroborates all of this, and we are throwing it out. In UST 2-year note futures, Leveraged Funds are 339,905 long against 2,096,433 short. [7] On its face, devastating confirmation. In fact, nothing of the kind: Asset Managers in the same contract are 2,475,937 long against 583,944 short. [7] That mirror-image structure is the signature of the cash-futures basis trade — hedge funds short futures against long cash notes financed in repo. It is a financing trade, not a bet on the Fed. Anyone citing the levered two-year short as proof the market is "positioned for hikes" is misreading a plumbing position, and we would rather say so out loud than bank a number that flatters us.

Note what the discarded evidence leaves behind, though. Real money — asset managers — is net long the two-year, the instrument that has already done the hiking at 4.26 percent. If this morning's CPI marks the turn, they win, and the levered SOFR shorts are the ones offside.

The Other Market: 83% Say It Won't Happen

Now hold everything above in your head and read the survey.

BofA's Global Fund Manager Survey was fielded July 2-9, 2026 — a field window and a set of findings a second outlet reports independently. [15] [21] That is after the dot plot, after the June 17 repricing, after the two-year cleared the top of the target range, with the futures strip already pricing roughly 68% odds of a hike by the October 28 meeting — the last one before the midterms. [9] And still: "83% do not expect the Fed to raise interest rates before the U.S. midterm elections in November." [15]

Seventeen percent, against roughly sixty-eight. On the same event, in the same fortnight. This is not two groups disagreeing about magnitude or timing. They hold incompatible views of a binary.

And this is not a survey of the uninformed. These are the people who own the equity risk — and they are not merely unpositioned for a hike. They are positioned against one, with no cushion whatsoever. Cash allocations fell to 3.6% from 4.1%, triggering BofA's own contrarian sell signal. [15] Fifty-four percent expect a "no landing." [15] U.S. equity allocations are at their highest overweight since December 2024. [15] Eighty-two percent call long semiconductors the most crowded trade, and 45% name an AI bubble the top tail risk. [15] Maximum risk. Minimum dry powder. Into a discount-rate event the same investors say will not happen.

They have also already marked down the thing that would justify the hike: BofA cut its end-2026 oil price forecast to $71/barrel from $86. [15] Which is coherent, actually — it is the same read the CPI just validated. It is simply not the read the rates market is trading.

Set the number of hikes and find out who pays

The temptation is to declare one side of this stupid. Resist it. The productive question is arithmetic: at any given number of hikes actually delivered, which market is offside — and by how much?

The rates market has already paid. The two-year at 4.26 percent is 51bp above the top of the range; against the 3.625 percent midpoint it has priced roughly two and a half quarter-point moves. [10] That means the market's own modal outcome — a single hike in September — is, to the front end, a mild dovish surprise. The strip barely moves. The people who get hurt are the 83%. Deliver nothing at all, and the 2,872,406-contract leveraged SOFR short gets squeezed hard, while equities receive a rate they never priced as a risk in the first place. Deliver BofA's three, and the SOFR shorts finally get paid and the 83% get run over.

There is no setting at which both markets are right. That is the entire finding. The hike is fully priced in the instrument that prices it and entirely unpriced in the instrument that would suffer from it — and once you see that, the "unpriced discount-rate shock" framing collapses and something more interesting takes its place. The risk has not gone away. It has moved.

Every Hawkish Dot Was Cast Before This Morning's CPI

Now the second correction, which is the one that makes this dangerous in the other direction.

The nine hawkish dots were submitted on June 16-17. [3] June CPI landed this morning, July 14. [12] Between those two dates, the inflation case fell apart.

Headline CPI (seasonally adjusted) fell 0.42% month-on-month in June, taking the year-on-year rate to 3.46% from 4.17% in May — a 0.71-percentage-point deceleration in a single month, the fastest of the cycle. [12] Core CPI was negative on the month: -0.02% m/m, with year-on-year falling to 2.57% from 2.82%. [12] On the CPI measure, core has never been the problem. It never got above 2.82%.

This was a broad downside surprise, not a rounding error. Bloomberg consensus was for headline to fall just 0.1% m/m and rise 3.8% y/y, and for core to rise 0.2% m/m and 2.8% y/y. [13] All four prints came in below consensus, and the monthly headline decline was "the largest single-month decline since April 2020." Energy fell 5.7% and gasoline 9.7% on the month. [13]

It was energy, and only energy

The CPI energy index fell 4.94% m/m in June, after rising 11.94% in March, 6.33% in April and 4.89% in May. Energy year-on-year peaked at 23.54% in May and is already down to 15.70%. [12] The entire inflation spike was energy, and energy has turned. Base effects from the April peak mean it decays mechanically from here unless the shock re-runs.

And the shock has round-tripped. Brent started 2026 at 61.98, peaked at 138.21 on 2026-04-07 during the Iran war, was 80.33 on FOMC day, and was back to 69.56 by 2026-07-06. [14] That is roughly half the April peak, and only about $7.58 above where the year began. The FOMC's 3.6 percent PCE projection and the nine hawkish dots were built on an energy shock that was already 42% off its high on the day they were submitted, and has since fallen another 13.4% from FOMC day to the latest EIA print. [14] [3]

The Fed knows this, incidentally — it says so itself. The June statement attributes the overshoot to supply, not demand: inflation "remains elevated relative to the Committee's 2 percent goal, in part reflecting supply shocks that have driven price increases in certain sectors, including energy." [1] The minutes go further, attributing the increases to "the lingering effects of tariffs, supply chain disruptions related to the closure of the Strait of Hormuz, and strength in demand for some goods and services stemming from robust AI-related investment." [4] A supply shock is precisely the thing a central bank is supposed to look through — and the June SEP's inflation revision was enormous: 2026 PCE median lifted to 3.6 percent from 2.7 in March, up 0.9pp in one quarter, with core PCE to 3.3 from 2.7, up 0.6pp. [3] Those are projections built on a barrel of oil that no longer exists at that price.

There is no Summary of Economic Projections at the July meeting. The next dot plot is September 15-16. [23] The hawkish June dots therefore stand unrevised through the July decision, with no official mechanism to walk them back short of the statement or a Warsh press conference. And today's CPI is effectively the last major inflation datum the July FOMC will see before it votes on July 29 — fourteen days from today. [23] [12] The July meeting is a data vacuum with a stale hawkish signal frozen inside it.

The market responded rationally, and only partly: the July hike bucket collapsed 26.6 percentage points in a session, from 41.0% to 14.4%, while September's cumulative hike odds still sit at 57.8%. [9] It moved the hike out, not away.

The hole in our evidence, stated plainly

Here is where we owe the reader something most commentary will not give them. The hawkish case does not rest on CPI. It rests on PCE, which is the index the Fed actually targets — and we hold no primary release from the Bureau of Economic Analysis. None.

The only PCE numbers we can verify are two, and neither is a measurement. The first is a forecast: BofA's June note asserted that "the Fed's inflation problem has gotten unambiguously worse. Core PCE could reach 3.5% in May, nearly 70bp higher than it was a year ago." [17] Could reach. That is a projection by a sell-side house, and it is labelled as one here. The second is the FOMC's own SEP projections of 3.6% headline and 3.3% core for 2026 — also a forecast, made by the same people whose dots we are interrogating. [3]

We therefore print no actual core-PCE figure anywhere in this report, because we cannot verify one. Neither, if they were honest, can most of the people quoting one at you. What we can measure has turned decisively lower, and it turned after every hawkish dot was cast. That cuts against us as much as for us, and we will come back to it in the falsifiers.

Tightened Words, Loosened Plumbing

Grant the hawks everything. Grant them the September hike. It still would not be what the discount-rate story requires.

Look at what the Fed's own median path actually does after 2026: 3.8 this year, then 3.6 for 2027, 3.4 for 2028, and 3.1 in the longer run. [3] It hikes once and then retraces. That is not a repricing of the structural discount rate. It is an insurance hike against a supply shock, and the Committee's own numbers say the shock is temporary. For anyone valuing long-duration equity cash flows, the difference between "the terminal rate is higher forever" and "the funds rate goes up 25bp and comes back down over three years" is the difference between a repricing and a headline.

The contradiction in the fine print

And then there is what the supposedly hawkish Fed actually did. The June implementation note directs the Desk to "increase the System Open Market Account holdings of securities through purchases of Treasury bills and, if needed, other Treasury securities with remaining maturities of 3 years or less to maintain an ample level of reserves," and the statement reaffirms "its policy of maintaining ample reserves in the banking system." [2] [1] Warsh's first executed act as chairman was to keep buying.

The H.4.1 for the week ending July 8, 2026 confirms it. Total assets: 6,735,609 in millions of dollars, up 73,697 over the year. [20] U.S. Treasury securities held outright: 4,499,699, up 291,234 on the year — more than offsetting MBS runoff of 190,098 to 1,948,398. [20] The balance sheet is growing.

Set that against the man's own documented doctrine. The Council on Foreign Relations' account of his confirmation hearing records that Warsh "has indicated he would seek to reduce the central bank's reliance on quantitative easing and shrink its balance sheet," favours reverting to a strict 2 percent target, and "wants to abandon the practice of forward guidance." [19] He has executed the guidance half — comprehensively — and inverted the balance-sheet half. He has tightened the words and loosened the plumbing.

The administered rates tell the same story of a Fed holding a floor rather than pressing down: interest on reserve balances maintained at 3.65 percent effective June 18; standing overnight repo at 3.75 percent; ON RRP at 3.5 percent with a $160 billion per-counterparty daily limit; primary credit approved at the existing 3.75 percent. [2] Nothing here is tightening. And the Desk itself flagged the reason for the bill purchases in the minutes: repo rates dropped to 15 basis points below the interest rate on reserve balances in mid-May, and the effective federal funds rate declined 2 basis points — the first such change since November. [4] Money-market softness, not tightness, is what the ample-reserves purchases are answering.

What Warsh actually said about his own stance

The crux of the whole reaction function is a single answer he gave. Policy, he said, is "somewhat restrictive" — but:

"I would have a hard time managing to say those words if I were to see what's happening in financial markets. So I'd say it's uneven. That's perhaps a function of different transmission mechanisms of monetary policy, whether monetary policy is coming from our interest rate tool or our balance sheet tool." [17]

Read that carefully. In his own framework, the instrument that is too loose is the balance sheet. Which implies that the tightening he wants is balance-sheet tightening — not a rate hike. That is precisely what 76.06 percent of the SOFR options market is paying for. [8]

The minutes corroborate the ambiguity: "Several participants remarked that they did not see the current policy stance as restrictive, while a few other participants commented that they saw the current policy stance as slightly restrictive." [4] Against core CPI at 2.57% y/y, the effective funds rate of 3.62% gives a real policy rate of about +1.05pp — mildly restrictive. Against headline CPI at 3.46%, it is +0.16pp — not restrictive at all. [22] [12] Which is the arithmetic underneath that sentence in the minutes. Against the Fed's own target index, we have already said we cannot compute it.

CFR named the tension at the heart of this man before he was even confirmed: the "instinctively hawkish" label "sits in some tension with his view that productivity gains driven by artificial intelligence (AI) could justify lower interest rates than would otherwise be warranted," and Trump "made clear his nominee would be expected to lower rates." [19] Their closing line is this report in one sentence: "if inflation reaccelerates, he may find himself compelled to raise interest rates—doing precisely the opposite of what the president who nominated him had in mind." [19]

And the political economy underneath it all is not a demand problem that a hike can fix. "After five consecutive years of inflation above the Fed's 2 percent target, with cumulative price increases approaching 25 percent, many Americans continue to feel the strain of an affordability crisis." [19] Warsh's price-stability pledge is attached to a five-year price level grievance, and a price level cannot be hiked away. Meanwhile the household economy is nowhere near the "solid pace" the statement describes: the University of Michigan Index of Consumer Sentiment was 49.5 in June 2026 — up from 44.8 in May, but down from 60.7 a year earlier, with Current Conditions at 47.7 and Expectations at 50.7. [26] Sub-50 is deeply depressed territory. That is not an economy screaming for restriction.

The Case Against This Report

Take the other side seriously, because it is professionally held by people who are not fools.

Bank of America — not a fringe house — forecasts three quarter-point hikes in 2026, in September, October and December, lifting the benchmark to 4.25%–4.5% from 3.5%–3.75%, reversing the last cut of 0.25 percentage points made on Dec. 10, 2025. [17] Their reasoning deserves to be quoted rather than paraphrased: "the Fed's inflation problem has gotten unambiguously worse... The Fed was willing to look through the tariffs, but it is losing patience after the latest round of supply shocks. Also, housing-driven disinflation has now mostly run its course, while other core services remain very sticky." [17] And they note, correctly, that Warsh conspicuously declined to push back on the idea of hikes.

Against them, Chen Zhao, chief global strategist at Alpine Macro: "while half of the Fed's voting members may be signaling their intention to raise rates, the odds of actual tightening remain very low." [17] He argues the end of the Iran war could send oil to $50–$60, wage growth is weakening, and AI is already improving productivity. BofA itself concedes that Warsh could be "strategically hawkish" to gain credibility while biding his time to cut later. [17]

The professional consensus is openly split. That fact alone destroys any claim that the hawkish view is a hidden edge. It is not. It is a coin-flip between two large houses, and it is sitting in the price.

The ways this report is wrong

Our framing has specific, dated ways to break, and we would rather name them than have them found.

The biggest is the deflator. If June core PCE prints at or above 3.4% y/y, the hawks have their justification, today's CPI was noise, and the nine dots were right. If it prints below 3.0%, the Fed's own target index has confirmed the CPI's turn. We hold no BEA source, so we cannot even tell you the release date without guessing, and we will not guess. This is the single largest hole in the evidence base and it is not on our side of the argument.

Hormuz. The disinflation is energy and only energy, and energy is hostage to a strait that keeps closing. Renewed US-Iran hostilities brought Hormuz shipping to a near-standstill again in July despite the June 17 truce; three merchant vessels were struck while transiting the Strait; roughly 6,000 seafarers are stranded; and oil recovered to approximately $77 per barrel on July 9. [24] The waterway carries roughly one-fifth of global oil and gas supplies. [24] On the other side of the same coin, Morgan Stanley reported that "35 oil and gas tankers had exited the Strait of Hormuz on Thursday, marking the first time the level had returned to within the range typical of pre-war levels," with Brent at $70.78 and WTI $67.74 on July 2 and analysts warning of a temporary surplus. [25] But note the clock: Iran's free-transit concession runs 60 days, and US-Iran sanctions relief expires August 21 — before the September FOMC where the hike is currently priced. [25] Re-close the strait and the entire energy shock re-runs and the hawks get their case back. This hike is, in substance, a levered bet on a waterway.

The two-year. If it retraces toward 3.90-4.00%, the market is unwinding the hike in real time and our "fully priced, you are late" framing flips from "too late" to "wrong direction."

Warsh's own mouth. His entire documented reaction function is one press conference in which he pointedly refused to give one. If he hedges on hikes at his semiannual testimony, or leans on transitory energy effects, or endorses this morning's disinflation, the hawkish premise unwinds inside 48 hours — because it rests on his silence, not on anything he has actually said. Conversely, if he explicitly validates the nine dots, the equity market's 83% is in immediate trouble. [5]

Positioning we cannot see. The COT we have is dated July 7 — a week before the CPI. The leveraged funds were already trimming their SOFR short by 87,568 contracts in that week, before the print. [7] Any claim about post-CPI speculative positioning is speculation, ours included.

And the dots themselves could harden. Six of eighteen participants are already at 4.125 or 4.375 — two hikes or more. [3] The 2026 central tendency has already shifted to 3.6–4.1 from March's 3.1–3.6. [3] The options market gives an 18.91 percent chance of the December window landing at 425–450bps. [8] If the September SEP's 2027 and 2028 medians rise above the 2026 median, we are wrong about the character of this move — it stops being insurance and becomes a plateau, and the equity valuation argument gets far worse.

What we will not do

We will not tell you whether the Fed hikes. Both calls are live, both come from credible houses, and the honest position is that the outcome is genuinely uncertain. What we will adjudicate is the asymmetry, and it is not symmetric in the way the consensus story implies.

The hike is priced — roughly 80% by December. [9] The equity market is not positioned for it — 83% expect none. [15] The data that has arrived since the dots were cast has run dovish, hard. [12] That configuration is dangerous in both directions. A delivered hike hurts the 83%, who are carrying their highest US equity overweight since December 2024 on 3.6% cash. [15] A cancelled hike squeezes 2,872,406 contracts of leveraged SOFR shorts into a front end that has already priced two and a half moves. [7] [10]

And underneath both scenarios sits a nine-nine committee whose median is not any member's actual view, chaired by a man who did not submit a dot, would not say why he had not hiked, and described his colleagues' projections as pencils — the kind with the big erasers. [5] Anyone selling you certainty here is selling you the June dot plot. The chairman has already told you what he thinks of it.