The Fed decides Wednesday at 18:00 GMT and, for the first time in the Warsh era, the meeting is genuinely live. Markets price roughly a one-in-three chance of a hike, three or four voters are said to be ready to move now, and the decision lands 48 hours after a truce in the Strait of Hormuz knocked oil down 7% in a single session. There is no dot plot this time, so the entire signal lives in three places: the statement, the vote and the podium.
Six weeks ago the June meeting flipped the 2026 dot from a cut to a hike and the market spent the rest of the quarter arguing about when, not whether. Now the argument gets its first real test. The Fed arrives at its July meeting with the funds rate at 3.50% to 3.75% and a committee that is no longer pretending to agree: nine of eighteen June dots carried at least one 2026 hike, three or four of the twelve voters are reported to be ready to move immediately, and the market has priced the decision at roughly two-to-one for a hold, which is another way of saying one-in-three for the first hike of the cycle.
Then, two days before the vote, the most important input changed. On Monday the United States paused its strikes on Iran, Tehran said it had halted retaliation, and crude fell more than 7% in a single session, unwinding in one day a large piece of the oil-shock argument the hawks have been carrying since spring. This preview maps the whole setup: what actually gets decided on Wednesday, the hike case and the hold case, the June CPI print that argued for patience, the oil round trip that reshuffled the odds at the last minute, where the signal lives at a meeting with no dot plot, the press-conference tells, the earnings collision two hours after the podium, and the cross-asset scenarios for the dollar, gold, rates and risk.
The statement drops at 18:00 GMT (2:00 PM ET) on Wednesday, with Chair Warsh's press conference at 18:30 GMT. There is no Summary of Economic Projections at this meeting; the next dots come in September. Markets price roughly a 35–38% chance of a quarter-point hike, the first genuinely two-sided FOMC pricing of the cycle. Into the print the dollar sits near a four-week high with EUR/USD around 1.1370, gold has slipped to about $4,030 from its mid-July highs, the US 10-year yield is near 4.63% after touching 4.71% when crude traded above $100, and Brent is back near $86 after Monday's 7% truce-driven slide. The rate may not move; the pricing says the risk is real.
In June the hold was priced at 99% and the drama lived in the projections. July inverts that structure. If the Fed holds, it will be the fifth consecutive meeting with the target range at 3.50% to 3.75%, unchanged since the December 2025 cut. Economists overwhelmingly expect exactly that: the FactSet consensus is a hold, and no primary dealer has a hike as their base case. But the market is not treating it as a formality. Futures assign roughly a 35–38% probability to an immediate quarter-point increase, pricing that has rattled between one-in-five and two-in-five all month as oil and the strait headlines moved. A decision that carried no genuine two-way risk in June now carries it.
Because there are no projections at this meeting, the day cannot hide behind a dot plot. The decision itself, the vote count and any dissents, the wording of the statement and the press conference are the entire information set. That concentrates the event rather than diluting it: every one of those four elements is capable of moving the tape on its own, and they arrive in sequence, which is the classic setup for a whipsaw. The June meeting taught the market that lesson the hard way, with gold reversing $163 from its pre-decision high once the page and the podium had been reconciled.
The June meeting was the regime change. The median 2026 dot moved from 3.4% in March to 3.8%, converting the year's last projected cut into a projected hike, and the new Chair declined to submit his own dot while launching five task forces to rebuild the institution. The market read it correctly as a hawkish turn and repriced: the dollar reclaimed its 2026 highs, the front end priced out cuts entirely, and gold gave back its spring premium in stages.
July then supplied the ammunition for both camps at once. The hawks got the oil shock back: after a month of ceasefire calm, US strikes on Iran resumed on July 8, the confrontation escalated into an exchange over the Strait of Hormuz that Tehran declared closed, and Brent ran from the mid-$70s to above $100 at the worst of it, dragging the 10-year Treasury yield to 4.71%, its highest since January 2025. The doves got the data: the June CPI, released mid-month, undershot expectations across the board and pulled the headline rate down to 3.5%. Then, on the eve of the meeting, the truce drained the urgency from the oil argument. That is why the pricing is stuck near one-in-three: both sides of the committee spent July being proven right.
The hawks' argument is straightforward and has not changed since spring. Headline inflation has been running between the mid-3s and low 4s all year, roughly double the target, and the committee's own June projections imply the appropriate policy rate is higher than the current one: the 3.8% median dot is above the 3.50–3.75% range the Fed is actually sitting in. Nine of eighteen participants have already signed their names to at least one 2026 hike. For that camp, waiting is not patience, it is drift: every meeting that passes with inflation near 4% and policy on hold erodes the credibility of the 2% anchor, and the produce-price pipeline (PPI printed 6.0% earlier in the year) suggests more pressure is coming through. Three or four of the twelve voters are reported to be prepared to vote for an increase this week, not in September.
The strait gives the argument its urgency. An economy absorbing a supply shock with a closed Hormuz and $100 crude is exactly the environment in which inflation expectations can slip their moorings, and expectations are the one thing a central bank cannot cheaply recover. The University of Michigan inflation-expectations print on Friday is, not coincidentally, the data point the hawks cite most. Monday's truce blunts this case but does not retire it: a pause in strikes is not a reopened strait, and the July inflation data, when they arrive, will still carry the $100-crude weeks inside them.
Kevin Warsh has spent the inter-meeting period repeating one line — "prices are too high" — while conspicuously not endorsing the hike his own committee's dots imply. The reporting into the meeting converges on three reasons he is likely to hold, and they are worth taking seriously because they describe his reaction function, not just this week's vote.
Note what the three reasons have in common: none of them is "inflation is beaten." That is why the hold, if it comes, will almost certainly arrive wrapped in hawkish language. Warsh's cleanest play is the one he ran in June, sounding entirely willing to hike while not hiking, because it defends the anchor without committing the institution before his task forces report.
The single most important release of the inter-meeting period landed on July 14, and it broke the doves' way on every line. Headline CPI rose 3.5% year on year against a 3.8% consensus, and fell 0.4% on the month, the largest monthly decline since April 2020, as the June ceasefire lull dragged energy prices down hard. The core was the real story: flat on the month against expectations of +0.2%, putting the 12-month core rate at 2.6%. Services ex-energy were flat and shelter rose just 0.1%, the softest reads of the year in precisely the categories the hawks call sticky. The Conference Board's verdict — that the print "closes the door" on a July hike — is roughly where the economist consensus settled.
Run it through the trimmed-mean lens from the June preview and the picture gets more interesting, in both directions. In June we noted that the energy spike was a textbook upper-tail outlier, the kind a trimmed gauge discards, which let Warsh look through the 4%-handle headlines. The June print is the mirror image: the energy plunge is now the tail, this time on the downside, and the same trim discards it on the way down. The honest read of Warsh's own yardstick is that underlying inflation sits somewhere between the alarming headline of May and the reassuring monthly of June, closer to 3 than to 2, cooling but not conquered. That supports a hold; it does not support a victory lap, and it is why the statement's inflation language matters more than usual this week.
No input into this meeting moved more, or later, than crude. The sequence matters, because the committee's hawks and doves were each handed a different month to point at. Early July: strikes resume, Brent reclaims the high $70s. Mid-July: the confrontation moves to the strait itself, Tehran declares it closed, tankers are hit, and Brent trades above $100 while the 10-year runs to 4.71% and gold pushes back toward its highs. Then Monday, July 27: Washington pauses strikes, Tehran stands down its retaliation, and crude gives back more than 7% in one session, with Brent settling near $86 into the decision.
For the meeting, the truce cuts two ways. Mechanically it is disinflationary and it landed squarely on hike odds, which is a large part of why the dollar's four-week-high grind stalled and the euro caught a bid into Wednesday. But a truce is not a ceasefire, a ceasefire is not a reopened strait, and the July price data will still carry the $100 weeks inside them when they print in August. A committee that was burned in June for calling the first oil spike transitory will be wary of declaring this round finished because of a 48-hour-old pause. Expect Warsh to treat oil the way he treated it in June: as a tail his gauges are built to ignore, while keeping the hike option visibly alive in case the strait closes again.
July is a non-projection meeting, so the instrument that carried the June shock, the dot plot, is not on the table. That redistributes the entire signal into three carriers, and the desk reads them in order.
First, the vote. It is the least-watched line of the release, and potentially the most consequential. A unanimous hold would be a genuine dovish surprise given the reported three-to-four-member hike camp, and would say Warsh talked his hawks into patience. One or two dissents in favour of a hike would be the first hawkish dissents of the cycle and would function exactly as a hawkish dot: a dated, named, official record that the hike camp has moved from projections to votes. Three dissents would say the chair is holding the line against a third of his committee, and the market would immediately price September as live. The dissent count is this meeting's dot plot.
Second, the statement. The algorithms will diff it in milliseconds; the desk reads the direction of five things:
| What to watch | Hawkish read | Dovish read |
|---|---|---|
| Inflation language | "Remains elevated" survives the soft June CPI; oil risks flagged as two-sided or upside | An acknowledgement that inflation has "eased" or that recent readings were encouraging |
| Balance of risks | Risks still tilted to inflation; readiness to tighten "if progress stalls" | Risks "moving into better balance", employment back in the sentence |
| Policy bias | Any explicit tightening bias or a line that the committee "discussed" an increase | Language preserving optionality in both directions with no hike reference |
| The vote | One or more dissents in favour of a hike — the day's real dot plot | Unanimous hold |
| Balance sheet | No change; QT continues on autopilot | Any hint the runoff review from the task forces is being accelerated |
Third, the press conference, which is where the June meeting was actually decided, and to which we now turn.
The June presser established the Warsh style: fewer promises, no pre-commitment, an open disdain for forward guidance, and a willingness to let the committee's disagreement show rather than papering over it. It also established that the market moves more on his posture than on the statement. This time he arrives with a specific problem: he must explain why a committee whose own median projects a hike, a third of whose voters reportedly want one now, is not delivering it, without sounding like a chair who has lost his hawks or one who is easing into political pressure. These are the questions he is nearly certain to face, and the tell in each direction.
| The question | Hawkish tell | Dovish tell |
|---|---|---|
| Your median dot says the rate should be higher. Why isn't it? | "The question is when, not whether" — a hike framed as sequencing | Distances himself from the dots as a forecast, not a commitment |
| What would make you hike in September? | Names concrete triggers: reopened-strait oil, firm trimmed-mean prints, Friday's expectations data | Sets a high, vague bar and leans on lags |
| Was a hike discussed today? How close was it? | Confirms it was seriously debated; declines to characterise the margin | Minimises the debate, stresses consensus |
| Is the oil truce changing your inflation outlook? | "A pause is not a reopened strait" — refuses to bank the disinflation | Welcomes it as removing the main upside risk |
| The June CPI was soft. Is underlying inflation beaten? | Points to his trimmed gauges sitting near 3: cooling, not conquered | Emphasises the flat core month and soft shelter as the trend |
| How do you answer those who say the Fed is holding rates for the White House? | Explicit independence defence; repeats that the next move can be up | Any answer that leans on growth risks as the reason to hold |
As in June, watch the clustering rather than any single answer. A chair who repeatedly refuses to bank the truce, refuses to celebrate the CPI and confirms the hike debate was live is delivering a hawkish hold whatever the statement said. One who banks the disinflation and waves at the task-force timeline is quietly telling you the 2026 hike is migrating into 2027.
This is the part of the setup that has no June precedent. Roughly two hours after Warsh leaves the podium, Microsoft and Meta report second-quarter earnings, with Amazon and Apple following Thursday after the close and US Q2 GDP printing Thursday morning. The four hyperscalers are on a combined capital-spending run-rate of about $725 billion for 2026, up 77% year on year, and the market has spent two weeks openly worrying about whether that spend is deliverable and financeable at higher-for-longer rates. Wednesday evening is where both anxieties meet: the discount rate at 18:30 GMT, the cash flows at 20:00.
| When (GMT) | What | Why it matters for the Fed trade |
|---|---|---|
| Wed 18:00 / 18:30 | FOMC statement, vote / Warsh press conference | The decision, the dissent count and the September signal |
| Wed ~20:00 | Microsoft & Meta Q2 earnings | Capex guidance can amplify or reverse the FOMC equity move overnight |
| Thu 12:30 | US Q2 GDP, jobless claims | First growth read spanning the oil-shock quarter; feeds September pricing |
| Thu ~20:00 | Amazon & Apple Q2 earnings | Completes the hyperscaler capex picture |
| Fri 13:45 / 14:00 | Chicago PMI, UMich inflation expectations | The expectations print the hawks cite most, plus month-end flows |
The practical consequence is attribution risk. An equity move on Wednesday night will be a blend of the Fed and the capex guidance, and Thursday's open will misattribute it; the rates and FX markets, which do not trade Microsoft's guidance, are the cleaner read on what the Fed actually said. If you want the pure FOMC signal, watch the front end and the dollar into the close and treat the equity tape as contaminated until Friday. The news-feed workflow applies double this week: one headline, one bias, and no borrowing conviction from a move that belongs to a different catalyst.
With real two-way pricing for the first time in the cycle, the scenario tree has four branches rather than June's three, and the probabilities attached to each are the trade.
| Scenario | What lands | Dollar (DXY / EUR/USD) | Gold | Rates & risk |
|---|---|---|---|---|
| Surprise hike (~1-in-3 priced) | 25bp increase to 3.75–4.00%, framed as insurance against the oil shock | Dollar spikes; EUR/USD breaks the 1.1380 shelf and opens 1.1300 | Hit hard through $4,000 as real yields jump; the two-thirds who priced a hold reposition at once | Front end gaps higher, curve flattens; equities sell into their own earnings, tech worst |
| Hawkish hold (base case) | Hold with one or two hike dissents; "elevated" survives the statement; Warsh keeps September openly live | Dollar firm to stronger; EUR/USD pressed back toward 1.1380 after any knee-jerk pop | Heavy, pinned near $4,000–4,050; relief capped by the September threat | Front end holds the hike premium for September; equities hand the baton to Microsoft and Meta |
| Neutral hold | Hold, minimal statement changes, no dissents, Warsh non-committal on September | Dollar eases off the four-week high; EUR/USD works back into the 1.1450 zone | Relief bounce toward $4,150 as the one-in-three hike premium bleeds out | Front end rallies modestly; risk drifts with earnings in control |
| Dovish hold (tail) | Unanimous vote, inflation language softened toward the June CPI, truce welcomed | Dollar drops; EUR/USD through 1.1450 with the June trend in question | Squeeze toward $4,200+ as hike pricing unwinds entirely | Curve bull-steepens; equities and crypto rally into the earnings releases |
The dollar goes in near a four-week high, having front-run the hike odds all month, with EUR/USD compressed to about 1.1370 after printing the high 1.13s last week — a full figure and a half beyond the 1.1600 pivot that decided the June meeting. The asymmetry has therefore partially reversed: a hike is one-third priced, so delivering it is worth perhaps two-thirds of a normal surprise, while a genuinely neutral or dovish hold unwinds a month of accumulated hawkish premium. The zone that matters is 1.1380–1.1450: holding below it keeps the dollar trend intact into September; a decisive reclaim of 1.1450 on a soft outcome is the first evidence the July repricing overshot.
Gold enters at roughly $4,030, down about 1% on decision eve and some $300 off the mid-July, strait-driven highs near $4,340, squeezed from both sides as the war bid faded with the truce and the hike bid lifted real yields. That double loss of sponsorship is why it sits so close to the psychological $4,000 line. The map is direct: a surprise hike almost certainly forces a test and break of $4,000 with the June recap's $163 lesson as the template; a hawkish hold likely pins it in the low $4,000s; anything softer lets it retrace toward $4,150 first and the strait headlines take it from there. Positioning cuts both ways after a month of two-sided flushing — the Friday COT print into the meeting is worth the look before assuming the crowd is long.
The front end is carrying the event: it is where the one-in-three sits, and it will reprice within seconds of the statement in whichever direction the vote and the language point. Further out, the 10-year at about 4.63% tells the fuller story of the month, up to 4.71% when crude was above $100, then bleeding lower with the truce; it is trading the oil shock's inflation half-life as much as the Fed. A hike or hawkish-dissent hold flattens the curve from the front; a neutral or dovish outcome steepens it back as the premium unwinds. Watch the front end against Thursday's Q2 GDP as a pair: a hawkish Wednesday plus a soft GDP print on Thursday is the whipsaw combination that has repeatedly reversed FOMC-day moves this year.
Equities face an unusually dense 30 hours: the rate decision at 18:00, the two largest AI spenders' guidance at 20:00, the growth print Thursday morning and the other two hyperscalers Thursday night. The index-level read on the Fed alone is the usual one, a hike hits long-duration tech hardest and a relief hold helps it most, but any Wednesday move should be treated as provisional until the capex numbers land. Crypto keeps its role as the high-beta liquidity proxy: hit on a hike, bid on relief, with the added wrinkle that it is the only major risk asset trading while the earnings calls are live. It is the only overnight read available, and it should be weighted accordingly.
A live meeting with no projections is a sequencing problem. The 18:00 GMT release answers two questions at once, the rate and the vote, and the algorithms will trade the rate while the desk reads the dissents; the first move routinely belongs to the wrong one. Then the podium either ratifies or reverses it, and two hours later the earnings tape contaminates everything at the index level. So the workflow this week is stricter than usual: front end and EUR/USD for the true FOMC signal, gold for the real-yield confirmation, and no conclusions from equities until Amazon and Apple have reported and Friday's expectations print has closed the week. The first move fades, the second move decides, and this week there is a third move that belongs to a different story entirely.
A hold that is only two-thirds priced, a committee with a third of its voters reportedly ready to hike, a chair whose own yardstick argues for patience while his median dot argues for action, an oil shock that deflated 48 hours before the vote, and the two biggest AI spenders reporting two hours after the podium: that is Wednesday. The rate is probably staying at 3.50% to 3.75%; almost nothing else about the day is settled. The dissent count is the headline to trade, EUR/USD 1.1380–1.1450 is the zone that scores it, gold's $4,000 handle is the stress gauge, and Brent decides whether September inherits the debate. Trade the vote, listen for whether the hike was merely postponed, and do not let Microsoft's capex line tell you what the Fed said.
Figures are prior official releases, consensus and analyst estimates and market levels (sources include the Federal Reserve, the BLS and BEA, CME FedWatch, Trading Economics, and reporting on the July meeting and Chair Warsh's public remarks from CNBC, CBS News, Morningstar, Kiplinger, the Conference Board, LiteFinance, EBC Financial Group and Al Jazeera) as of publication on 28 July 2026, and are subject to revision. Market levels are drawn from TTerminal's own data as of the latest close into the meeting. This article is TTerminal's own market analysis and is not investment advice.
The macro calendar, the FOMC cross-asset reaction, the dollar, gold, yields and the risk map, all live across forex, indices, commodities and crypto.
Start 7-day free trial ↗