Last week, the open question was whether a hot July CPI print would back the Fed into a corner and force a hike. This week, the opposite arrived — and it reopened the same coin flip from the other side.
Here's the thing most readers get wrong: they treat "soft CPI" as a directional resolution. Cooler inflation, dovish Fed, done. But the print that dropped Tuesday didn't resolve anything. It just moved the disagreement to a different market.
Start with what actually happened. Headline CPI rose 3.5% year-over-year in June, down sharply from 4.2% in May, and fell 0.4% on the month. Two-year Treasury yields dropped up to 14 basis points to 4.14%, the largest single-day move since February.
The tape did exactly what you'd expect a relief rally to do. Gold reversed off its lowest level since July 1 to close up 1.6%. The Nasdaq rose 0.90%, with Nvidia up 4.06% and Micron up 4.92%. The dollar index fell 0.6%.
So far, so clean. Now look under the hood, because the real story isn't headline — it's the number the Fed actually watches.
Core CPI was unchanged on the month, after rising 0.2% in May. As metals trader Tai Wong put it, "headline dive lower but more importantly, core unchanged." That flat print is what pulled July and September hike bets off the table.
But headline inflation is the number energy and food push around. Brent crude climbed above $87 on fresh attacks near the Strait of Hormuz — the exact channel that made June's headline look tame and threatens to make the next one look hot again.
Which brings us to the contradiction at the heart of this week. The CPI report was genuinely soft. The Fed's preferred gauge is still climbing.
Per the Cleveland Fed Nowcast, core PCE was tracking 3.43% year-over-year in June and 3.47% in July — and notably, those forecasts are still rising. Core CPI annualizing to roughly 2.80% is one thing. Core PCE running toward 3.4% is a different animal, and it's the animal Chair Warsh's Fed says it targets.
Core CPI annualizing to roughly 2.80% is one thing. Core PCE running toward 3.4% is a different animal, and it's the animal the Fed says it targets.
That gap is why the coin flip didn't close. It moved.
Watch how the prediction markets split the difference. On Polymarket, the near-term contracts screamed relief: "Fed Decision in July?" priced No change at 90.5%, and one wrap put the July-hold probability as high as 93.5%. Nobody thinks the Fed hikes into a soft CPI three weeks before the meeting.
But push the horizon out and consensus dissolves. The "Fed rate hike in 2026?" contract sits at 51.5% Yes on roughly $4.1 million matched — a literal coin flip, and one that just fell 15 points from 66.5%.
Read those two numbers together. The market is near-certain about July and split down the middle about the year. That's not a dovish pivot. That's traders pricing a soft print as a delay, not a reversal.
The September ladder tells the same story with more texture. "No change" leads at 56.5%, but a 25bp increase still holds 38.5% — a hike path that refuses to die. Cuts are priced as tail outcomes: 25bp decrease at 4.0%, anything bigger below that.
So the distribution of outcomes for September is: probably nothing, meaningfully maybe a hike, almost certainly not a cut. One soft CPI report did not put easing on the table. It just softened the odds of the next tightening.
This is the mechanism worth internalizing. A downside surprise in headline CPI removes the immediate forcing event — it takes July off the board — without touching the structural argument for hikes, which lives in core PCE and the capex-driven price pressures that have kept inflation above target for years.
The Fed knows this, which is why the official posture didn't soften. In his first semi-annual hearing, Warsh reiterated the commitment to the 2% target while emphasizing the employment mandate — the classic dual-mandate hedge that commits to nothing.
And the timing of the soft print was, for Warsh, a gift. As Natixis economist Christopher Hodge noted, "After coming out swinging as a hawk, a hotter-than-expected CPI print could have backed him into a corner and forced a hike." The cool number bought him breathing room — the freedom to stay hawkish in tone without having to prove it with action in July.
That's the trap in reading this week as dovish. The soft CPI doesn't constrain the Fed toward cuts. It relaxes the constraint that would have forced a hike, leaving Warsh free to keep the tightening bias alive on his own timeline.
Fort Washington's Dan Carter framed the base case cleanly: "Near-term hikes are off the table... Our base case is the Fed will remain on hold." Hold — not cut. The relief rally in Treasuries was the market re-steepening the curve around a pause, not pricing an easing cycle.
There are two other reasons the coin flip stays live, and both cut against the relief narrative.
First, oil. The same Hormuz disruption lifting Brent is precisely the kind of energy-driven shock that hits headline first and bleeds into core later. If crude holds elevated, the next CPI could reverse Tuesday's move — which is exactly why swap markets and Polymarket are keeping the hike rungs populated rather than collapsing them.
Second, the labor market gives the Fed no clean read. Chicago Fed's Goolsbee has described a labor market frozen in place — hiring and layoffs both stalled — which he calls puzzling, not reassuring. A frozen labor market removes the growth-scare that would normally push the Fed toward cuts, leaving inflation as the only mandate sending a clear signal. And that signal, via core PCE, still points up.
So the honest read on this week is narrow. The soft June CPI killed the July hike and eased September odds. It did nothing to the 2026 path, which is why the year-ahead contract snapped back to 50/50.
The tell to watch is whether "No change" for September can hold above its five-reading average of 53.9% — or whether probability leaks back into the 38.5% hike rung as oil and the July 30 core PCE print land. Note that PCE report drops after the July 28–29 meeting, so the Fed decides July on incomplete information and defends the decision with data it hasn't seen yet.
The burden of proof did shift, as the gold desk argued — the hawks now have to justify further tightening against a cooling headline. But "burden shifted" is not "case closed." Core PCE at 3.4% and rising is a heavy counterweight, and one flat core CPI month doesn't lift it.
One cooler report does not make a trend. The market knows this — it's why the near-dated contracts scream relief and the year-out contract sits on a knife's edge. The soft print didn't answer the question. It just handed the question forward, to oil, to PCE, and to a Fed chair who now has every reason to stay quiet and keep his options open.