JPMorgan's July CPI Scenarios: The S&P 500 Flips Sign at a 0.25% Core Print
JPMorgan’s trading desk put out its scenario grid for Wednesday’s July consumer price index, and the number that matters is not the base case. It is 0.25%, because that is where the sign of the expected S&P 500 move changes, and consensus sits five basis points below it.
The grid runs five buckets on core month-over-month. Above 0.30%, assigned a 5% probability, the desk sees the index down 1.5% to 2.5%. Between 0.25% and 0.30%, at 25% probability, down 0.5% to 1.25%. Between 0.20% and 0.25%, the modal outcome at 40%, up 0.25% to 0.75%. Between 0.15% and 0.20%, at 25%, up 0.5% to 1%. Below 0.15%, at 5%, up 1% to 2%. Economists polled by Dow Jones expect headline CPI to rise 0.1% on the month, taking the twelve-month rate to 3.4%, with core up 0.2% and 2.5% year over year.
Core CPI is reported to one decimal place. The scenario boundary that separates a positive expected day from a negative one sits at the second. A seasonally adjusted core reading of 0.249% and one of 0.251% both print as 0.2% on the wire and in every headline that crosses at 8:30, and they land in buckets whose midpoints are 63 basis points apart in the wrong direction. The desk is not being cute here. That is genuinely where the algorithmic reaction function breaks, because the machines read the index level, not the rounded release table, and the second decimal is available immediately.
Run the grid’s own arithmetic and the expected value comes out at roughly plus 14 basis points, which is close enough to zero to be noise. Seventy percent of the probability mass produces a positive close, but the negative thirty percent carries a larger average magnitude. Symmetric misses are not symmetric in payoff: a print five basis points hotter than the modal bucket costs a midpoint of 87 basis points, while a print five basis points cooler pays 75. Only about a tenth of the distribution generates a move larger than 1.25% in either direction, which is a thin set of tails for an event that options desks have historically priced near 1%.
The more useful comparison is against the same exercise two months ago. In June the desk’s most likely scenario was core between 0.25% and 0.30%, with the index somewhere between down 0.5% and up 0.75%, and it flagged elevated tail risk skewed to the hawkish side, with anything above 0.35% worth 2% to 3% of downside. The entire grid has since shifted about ten basis points lower and both tails have compressed. What has not improved is the payout for being right about disinflation. In June a core print at or below 0.2% was worth 1.5% to 2%. Now a print in the 0.20% to 0.25% band is worth 0.25% to 0.75%. The market has already bought the cooling. It is paying a third as much for confirmation.
That repricing happened fast. The S&P 500 closed July at 7,489.81 and traded near 7,769 on Monday against a record high of 7,793.68, which is roughly 4% in seven sessions, most of it delivered by Friday’s payrolls miss. The rally mechanism is worth stating precisely, because it is unusual: stocks are rising on the removal of an expected rate increase, not on the pricing of a cut. The Fed held at 3.50% to 3.75% on July 29 with three officials voting for an immediate quarter point, and fed funds futures now imply roughly a 52% chance of a September hike. That figure had fallen toward the mid-forties on Friday afternoon. It recovered over two sessions while WTI climbed back toward $80 on the Iranian demands over Hormuz and the ten-year yield returned to 4.68% from 4.60%.
That recovery is the part of the setup the scenario grid does not capture. JPMorgan notes that the U.S. has so far avoided an inflation spike from the Middle East conflict, which is true of the data through June and says nothing about the pass-through still in transit. Headline is expected at 0.1% on the month while the annual rate rises to 3.4%, which is a base effect against a soft July 2025 comparison rather than a fresh acceleration, but it does put headline inflation level with the bottom of the funds rate corridor. A committee with three members already voting to hike does not need core to surprise in order to keep the September option live. It needs headline to keep drifting up while energy does the work.
So the asymmetry is not really in the grid. It is between the grid and the meeting. A cool core print buys the equity market 25 to 75 basis points and removes very little from September, because the hawkish case is being made on headline and oil. A hot core print at 0.26% or above costs between half a percent and two and a half, and hardens a hike that is already a coin flip.
The line to watch when the release drops is owners’ equivalent rent. Shelter is roughly a third of the index and it is the single component with enough weight to move core from 0.21% to 0.27% on its own, which is the entire distance between the two outcomes that matter.