May CPI lands Wednesday at 12:30 GMT with headline near 4% and an oil shock still running. But the deeper story is the ruler itself: new Chair Kevin Warsh wants to judge inflation by trimmed-mean and median gauges rather than core PCE, and Wednesday is the first big test of what that changes for the dollar, gold and the front end.
For two years the inflation trade was a one-way street: the only question was how fast disinflation would let the Federal Reserve cut. That street has closed. After an energy shock tied to the war in Iran pushed headline CPI from 3.3% in March to 3.8% in April, the May report due Wednesday is no longer a referendum on rate cuts. It is a referendum on rate hikes. That inversion is the single most important thing to understand before 12:30 GMT.
And there is a twist that makes this print different from any in years: the man now reading it wants to measure inflation by a different ruler. This is the full institutional read: where inflation actually sits, what the consensus expects, the components the desk decomposes line by line, the new trimmed-mean yardstick Warsh wants to judge it by, how that reshapes the Fed's reaction function, and the cross-asset scenarios for the dollar, gold, rates and risk into the June 16 to 17 FOMC.
The oil-shock regime this preview is built around has just started to crack. Crude fell about 4% on the day (WTI near $87.58) and is now roughly 20% off its 2026 peak after Israel and Iran agreed to halt attacks, the exact de-escalation we flag below as the one thing that flips the trade. Gold slid to about $4,297, its lowest since late March; the dollar eased to around 99.80 on the index; and bitcoin fell to around $61,600 as the hawkish rate repricing hit risk assets. Yet the hawkish story has not gone away: with the labour market resilient, markets now price more than a 70% chance of a Fed hike by December, up from roughly 14% a month ago. The energy premium is draining, which pushes the inflation risk from oil onto the domestic core, and makes tomorrow's composition, read through Warsh's trimmed lens, even more decisive.
The street looks for inflation to grind higher again. Headline CPI is seen rising roughly 0.3% on the month, lifting the annual rate to about 4.0% from 3.8%, with the hawkish tail of forecasters (including desks at JPMorgan) flagging risk toward 4.2% if energy passthrough is stronger. Core CPI, which strips out food and energy, is seen up 0.3% on the month and 2.9% year-on-year, a tick higher than April's 2.8%.
Two numbers, two different jobs. The headline is the political and market-psychology number, dominated by the energy spike, and it is what pushes the "is the Fed about to hike?" narrative. Core is the number the Fed actually sets policy on, because it filters out the volatile energy swing and reveals whether the inflation is becoming embedded. A 4% headline with a benign core is a very different report from a 4% headline with core re-accelerating, and the market will reprice within milliseconds on which one prints.
Disinflation was real until it wasn't. CPI had cooled toward the low 3s, then the conflict in Iran hit the oil market. Energy costs jumped sharply in the spring (the energy index rose on the order of 17% over twelve months in the April report), and that single category dragged the headline back up the hill: 3.3% in March, 3.8% in April. Crude staying bid on Strait of Hormuz risk means the energy contribution has not yet rolled off.
The danger for the Fed is not the first-round energy spike, which it would normally look through. The danger is second-round effects: higher fuel feeding into airfares, delivery, freight and ultimately core services, plus the risk that elevated headline prints unanchor inflation expectations. That is exactly the channel a hawkish central bank refuses to gamble on, and it is why a "clean" energy-only beat is not as comforting as it looks.
The algorithms trade the two headline numbers in the first second. The desk goes straight to the composition, because that is what tells you whether the move holds:
Under Chair Kevin Warsh the Fed has made price stability the overriding objective, and the market has repriced accordingly. The policy rate has been held in a 3.50% to 3.75% range, and for the June 16 to 17 meeting the probability of no change is effectively total: CME FedWatch has priced the hold at around 99%. The live debate is no longer about the timing of the first cut. Goldman Sachs sees rates on hold through 2026 with cuts pushed into 2027, and the market now prices more than a 70% chance of a 25 basis-point hike by December, up from around 14% a month ago.
That reframes Wednesday. CPI will not move the June decision, which is already a near-certain hold. What it moves is the 2026 path and the dot plot that lands a week later. A hot core gives the hawks the evidence to formalise a hiking bias in the June Summary of Economic Projections; an in-line-to-soft core lets the Fed keep optionality and stay on hold without escalating. Add a resilient labour market (May payrolls came in around +172K with unemployment at 4.3%, comfortably above the level that would pressure the Fed to ease) and the asymmetry is clear: the data are not building a case to cut, so every upside inflation surprise lands directly on hike odds.
An upside core surprise is amplified: it confirms the hawkish path and lifts hike odds, the dollar and front-end yields. A downside surprise is muted: it is a relief rally that most likely fades, because one soft print does not undo an oil shock or change a Fed whose bar to cut is inflation, not growth.
Here is the part most previews will miss, and it may matter more than the headline itself. The Fed is being handed this report by a chair who does not fully trust the way the number is built. Through his confirmation and his first weeks in the job, Kevin Warsh has been blunt that the central bank needs a different framework for reading inflation, dismissing the Fed's preferred core PCE gauge as only a rough approximation even after stripping out food and energy. His stated preference is to lean on trimmed-mean and median measures instead. Wednesday is the first major inflation print he will judge through that lens, and the choice of ruler could change the policy conclusion even if the headline lands exactly on consensus.
A standard core reading removes two categories by rule, food and energy, and keeps everything else. A trimmed-mean gauge does something more surgical: it ranks every component of the basket by how much its price moved that month, throws out the most extreme movers on both the high and the low tail, and averages what is left. The median measure is the limit case, taking the single middle component. The point is to expose the central tendency of inflation by refusing to let a handful of violent outliers, in either direction, dominate the print. Two official series already do exactly this: the Dallas Fed Trimmed Mean PCE and the Cleveland Fed Median CPI, alongside its 16% trimmed-mean CPI.
This is where it bites on Wednesday. The force dragging the headline back toward 4% is energy, and energy is the textbook outlier a trimmed gauge is built to discard. If May's strength is concentrated there, the trimmed-mean and median readings will sit materially below the headline and below core, and they have at times printed closer to target than core PCE: trimmed measures were running in the mid-2s against a core PCE near 3% earlier in the year. For a chair who wants a reason to stay on hold rather than escalate to hikes, a near-4% headline that dissolves into a benign 2-point-something trimmed core is precisely the cover he is looking for.
The trimmed lens is not automatically dovish, and this is the nuance the desk watches for. Stripping outliers only cools the read when the hot prints are genuinely isolated. If May shows broad strength, shelter firm, core services ex-housing sticky, goods re-accelerating on tariffs, then there is no narrow tail to trim away, the middle of the distribution is simply higher, and the trimmed-mean and median come in hot too. History is the warning: in 2019 and 2020 the trimmed median ran hotter than core PCE, which would have argued for a more hawkish stance, not a softer one. So the question on Wednesday is not only how high the headline is, but how wide the pressure is. A narrow, energy-led beat is a gift to a dovish hold; a broad-based beat turns Warsh's own preferred gauge against him.
There is a political and credibility layer the market will not ignore. Warsh arrives under visible pressure for lower rates, in the middle of an oil-driven inflation scare, and is proposing to elevate a measure that, right now, happens to read cooler than the official core. That invites the obvious critique that the Fed is changing the scorecard to fit the desired answer. Whether the market treats a soft trimmed print as genuine signal or as goalpost-moving will itself move the dollar and the front end. The cleanest way for Warsh to earn the benefit of the doubt is for the trimmed gauges and core to agree; the most destabilising outcome is for them to diverge sharply, leaving the Fed to argue about which number is real while markets pick their own.
We read Wednesday on both rulers at once. Alongside the CPI headline and core we track the implied trimmed-mean and median trend, we measure the breadth of the increase, how much of the basket is rising faster than roughly 4% annualised, rather than just its level, and we model the energy contribution explicitly so we can see what survives the trim. Under Warsh, a narrow 4% and a broad 4% are not the same report, and positioning that trades only the headline tick will be offside the moment the composition is digested.
We frame three outcomes around the core monthly figure, because core is what the policy path keys on. The headline sets the emotional tone; core decides whether the move sticks.
| Scenario | What prints | Dollar (DXY / EUR/USD) | Gold | Rates & risk |
|---|---|---|---|---|
| Hot | Core +0.4% MoM or higher, core YoY ≥ 3.0%, or hot supercore/shelter | Dollar firms off 99.80, EUR/USD breaks below 1.1600; December hike odds push past 70% | Pressured: breaks the 200-day (4,433), toward 4,220, with the safe-haven bid already fading | 2Y and 10Y yields up; equities and bitcoin sell off, growth and tech hit hardest |
| In line | Core +0.3% MoM, headline ~4.0% YoY led by May's energy | Dollar holds near 99.80, off its highs; EUR/USD heavy but holds 1.1600 | Drifts near 4,300 on a thinner floor now the geopolitical bid is fading; the trimmed lens caps the hawkish follow-through | Front end little changed, the >70% December hike intact; risk drifts |
| Cool | Core +0.2% MoM or lower, soft shelter/supercore | Dollar dips, EUR/USD pops; with oil already falling the dovish read has more traction than before | Relief bounce toward 4,400, but the fading safe-haven bid caps it | Yields ease, equities and bitcoin bid; December hike odds slip from above 70% |
Now overlay the trimmed lens on each row. A hot but narrow beat, an energy-driven 4% headline with a benign trimmed core, mutes the hawkish reaction and lets Warsh hold: read it as closer to the "in line" column for the dollar and rates even if the headline screams "hot." A hot and broad beat, where the trimmed-mean and median rise too, is the genuine hawkish scenario and reinforces the move on every gauge. The breadth is the tie-breaker between the two.
The greenback rose through May and June on safe-haven demand and the inflation-and-energy story, but it has come off the boil with the de-escalation, easing to around 99.80 on the index and down modestly on the day even as it holds most of the month's gains. EUR/USD has so far defended the 1.1600 level. In an inflation-led regime the dollar's reaction function is skewed: a hot core extends the trend, while a soft core produces a dip that has repeatedly faded. Only a genuine break of the oil story turns dollar weakness structural.
Gold is the cleaner two-way trade and it goes into the print on the back foot, having slipped to about $4,297, below $4,300 and its lowest since late March, surrendering its year-to-date gains as rate-hike fears built (it traded near $4,440 only a week earlier, with the 200-day moving average around $4,433 now acting as resistance rather than support). Two forces meet here. The headwind is a hawkish Fed, higher real yields and a firm dollar, all of which weigh on a non-yielding asset. The tailwind is the geopolitical bid and gold's role as an inflation hedge. A hot core presses gold toward $4,220; a soft core lets it reclaim the 200-day and squeeze back toward $4,400. The Middle East risk is the floor under the trade; the CPI core decides which side of the 200-day gold closes the week.
The front end of the curve is the purest expression of the print: a hot core lifts the 2-year yield as hike probability is repriced, and a flatter or inverted curve follows. Equities trade the textbook inflation reflex, with long-duration growth and technology the most rate-sensitive, so a hot core is a headwind for the Nasdaq in particular. Crypto trades as a high-beta liquidity proxy: it leans risk-off into a hawkish surprise and rallies on a dovish relief, while keeping a partial "debasement hedge" bid that can blunt the downside. Right now it is on the hawkish side of that: bitcoin has slipped to around $61,600 alongside the broad de-risking and the jump in December hike odds, its debasement bid not enough to offset higher-for-longer rates.
Strip everything back and the dollar's run, the gold sell-off and the hike debate are all the same trade: an oil-driven inflation scare. The one development that flips the regime is not a single CPI beat or miss. It is a de-escalation in the Gulf that pulls crude lower, and that is now under way: Israel and Iran have agreed to halt attacks, crude has dropped about 4% on the day and roughly 20% from its 2026 peak, and the energy premium is visibly draining out of the complex. That is the regime-flip in motion, and it is why the inflation risk is migrating from energy to the domestic core.
The catch is that it is fragile rather than finished. The Strait of Hormuz remains partly disrupted under a US to Iran blockade, so the de-escalation is real but incomplete, and a single headline can reverse the move. The cleaner read: the May CPI out Wednesday is backward-looking and will still carry May's elevated energy, but with crude now falling the market will increasingly look through the headline to a forward path where the energy contribution rolls off and Warsh's trimmed gauge reads cooler. The hawkish risk that remains is breadth, the core, shelter and services, not the barrel.
CPI is a top-tier volatility event, and we treat it as a set of structured inputs rather than a single number: headline against core, core against shelter and supercore, the print against the energy path, all weighed through the Fed's current reaction function rather than the old cut-timing reflex. Expect a whipsaw. The first move trades the headline and frequently reverses once the components are digested, so the edge is usually in the second move, after the market has worked out whether the core confirms the hawkish path or quietly undercuts it.
A near-4% headline, a core that the Fed needs to see fall and will not tolerate rising, an oil shock still in the system and a central bank that has swapped the cut debate for a hike debate: that is the setup. The dollar holds the cards into the FOMC, gold leans on its geopolitical floor while it fights the 200-day, the front end is the scoreboard, and the genuine game-changer sits in the oil market, not in any single line of the CPI table. Trade the core, respect the energy story, and wait for the second move.
Figures are consensus and analyst estimates, prior official releases and market levels (sources include the BLS, Trading Economics, the Cleveland Fed inflation nowcast and median CPI, the Dallas Fed trimmed-mean PCE, Kiplinger, Morningstar, Goldman Sachs, JPMorgan, CME FedWatch, and Chair Warsh's public remarks on trimmed-mean inflation reported by CNBC, Yahoo Finance and Charles Schwab) as of publication, and are subject to revision. This article is TTerminal's own market analysis and is not investment advice.
The macro calendar, the CPI cross-asset reaction, the dollar, gold, yields and the risk map, all live across forex, indices, commodities and crypto.
Start 7-day free trial ↗