CPI Locks the Door

Prediction markets now price elections, policy, sports, and the economy in real time—often faster and sharper than traditional forecasting. Our job is to help you read those signals and act on them.

Glad to have you here. Let's find the edge.

A week ago the September Fed decision was still a coin flip. This morning’s inflation sheet closed ranks.

August consumer prices rose 0.4% on the month and 3.4% on the year — exactly as expected — while core prices, stripping out food and energy, climbed 0.3%, a tenth hotter than forecast. Gasoline jumped 3.9% and did most of the headline work; shelter ticked up 0.3% again. By early afternoon, Polymarket priced a 25-basis-point September hike at 81% (Kalshi 80%). CME’s futures-based FedWatch sat near 90%. Five days from the vote, the hold trade that was still alive after Waller is a minority price. — The Editor

What the markets think this week — after the last major inflation print before the meeting. Friday’s CPI turned a near-coin-flip hike into an 80-cent contract. Weekly arrows are versus the September 8 issue (Friday Sep 4 snapshot). Bars are colored by category; verified live Friday afternoon before this draft.

The week’s biggest live-market move was the September hike itself.

Last Friday, after the 162,000 payrolls beat, Polymarket priced a 25-basis-point September increase at 52% and a hold at 49%. Thursday’s PPI — wholesale prices up 0.4%, energy up 4.2% — nudged futures-implied odds into the high 60s. Then August CPI arrived: headline in line, core a shade hot, gasoline doing the heavy lifting on the top line.

By this afternoon the September 25bp hike sat at 81% on Polymarket and 80% on Kalshi. A hold was 17% / 20%. A separate Polymarket contract on any hike in 2026 climbed to 89%. The policy-sensitive two-year yield rose about five basis points to 4.59%. That is a 29-point swing in a week on the same contract — from “either way” to “mostly priced.”

Waller’s published if-then had named inflation as the swing factor. The sheet did not cool enough to keep a hold as the base case. What remains is the meeting itself, not another major print.

Core inflation is the consumer price index with food and energy removed.

Those two categories swing hard on weather, harvests, and oil shocks, so economists often treat the leftover “core” as a cleaner read on the underlying trend the Fed can actually lean on. Headline CPI is what households feel at the pump and the grocery aisle. Core is what policymakers argue over when they ask whether inflation is sticky or fading.

Friday’s split made the distinction concrete. Headline matched the forecast at +0.4%, mostly because gasoline rose 3.9%. Core came in at +0.3% on the month — a tenth hotter than expected — with shelter still climbing. A cooler core would have been the reading that could have put a hold back in front after 162,000 jobs. A hotter core is why the hike contract jumped into the 80s even though the annual core rate, at 2.4%, still looks milder than the 3.4% headline. Same release, two numbers, different jobs.

Eighty-one percent is not one hundred. The market that captures the next move still settles on September 16, and a minority price on a hold means something would still have to happen for that minority to win.

The books already include Friday’s CPI and last week’s payrolls. They do not include a surprise from the Fed’s own statement, a dissent that markets read as a warning shot, or an eleventh-hour data revision nobody is watching. Chair Warsh has already said that if the numbers do not improve, “we have work to do.” A hike that merely confirms that bias would leave the 81% roughly where it is. A hold after an in-line-to-hot core print would be the shock — and would force the hold contract out of the teens in a hurry.

Watch the statement language and the new dot plot more than another macro sheet. The last major print before the vote has printed. It resolves September 16.

Last week we handed September to the CPI sheet. The sheet arrived. Here is the grade.

CPI as the swing factor — green. After Waller, we wrote that a cooler print could put a hold back in front even after 162,000 jobs, and a hotter one would stack with payrolls. Core printed a tenth hot. The September hike ran from 52% to 81% on Polymarket in a week. An 80% price that pays later is still only a price — but the direction of the move matched the map we published.

The midweek hold — red as a lasting base case. Hold was briefly the favorite at 57% after Waller. Payrolls knocked it off; CPI buried it at 17%. A base case that lasts 24 hours and then fails twice is a lesson in how fast “favorite” decays.

The fading 2026 cut — green. We said the cut trade was dying near 8–9%. Kalshi and Polymarket now price zero cuts this year at about 93% (any cut roughly 7%). That call held.

Mortgage quotes, auto loans, and credit-card APRs do not wait for the Fed’s press conference. They move when markets decide a hike is the likely path — and this week they decided.

An 81% September hike price is the market’s way of saying borrowing costs are more likely to tick up than to ease next week. The funds rate has sat at 3.50–3.75% all year; a quarter-point increase would push the top of that range to 4.00%. Gasoline is already higher — AAA had regular near $4.29 on Friday — so the same week that locked in hike odds also made the commute more expensive. Savers still collect the high yields that come with a restrictive stance; anyone waiting on cheaper financing is waiting on a path the books have mostly closed for September.

None of this is advice. It is Friday’s inflation signal, translated into the costs people carry into the week of the vote.

The legal map did not redraw this week. It waited.

The Ninth Circuit’s August 28 ruling against Kalshi — sports event contracts are not CEA swaps, so federal law does not preempt state gambling rules — still stands opposite the Third Circuit’s earlier Kalshi-friendly read. New Jersey’s September 2 petition asking the Supreme Court to take that split is still just a petition: the justices have not granted certiorari. Minnesota’s preliminary injunction against the first state ban remains in force. The CFTC’s Rule 40.11 “gaming” proposal is still pending after its July comment close.

Quiet weeks matter too. A circuit split with a live cert petition is a fall story, not a resolved one. The docket holds; the Court has not spoken.

After the same CPI print, CME FedWatch sat near 90% for a September hike while Polymarket sat at 81% and Kalshi at 80%. Same news. Different prices. That is not a bug to “fix.” It is a feature to understand.

FedWatch is built from fed-funds futures — a deep institutional market with its own margins, participants, and contract design. Polymarket and Kalshi are prediction markets: yes/no contracts that pay a dollar if a named outcome happens, traded by a different crowd with different size and different hours. When two books disagree, they are not grading the same homework. They are averaging different sets of bets under different rules.

For a retail reader, the useful habit is not to pick a winner. It is to notice the spread. When the books cluster, confidence in the price is higher even if the event is still uncertain. When they gap — 90% here, 81% there — you are looking at methodology and clientele, not at a single “true” probability. Markets can be wrong. Two markets can also be differently wrong.

Headline inflation matched the forecast because energy jumped. Core was only a tenth hot. Should a gas-driven print lock in a rate increase?

Yes. Households do not live in core CPI. Gasoline up 3.9% in a month is inflation they feel, and Chair Warsh has already warned that underlying trends have not improved. Energy that stays high can pass through into other prices and into expectations. Waiting for a cooler core while the pump climbs, in this view, ignores the mandate people actually experience.

No. The Fed’s own framework leans on core and on the PCE gauge precisely because food and energy are noisy. A one-month gasoline spike tied to Middle East risk is a supply shock, not proof that demand needs another quarter-point of restraint — especially with annual core at 2.4%. Hiking on the top line, in this view, treats a volatile category as if it were sticky services inflation.

We take no side. The books are 80–81% hike. They can still move before Wednesday.

Each week, Prediction Markets Alert reaches a curated and growing list of professionals across prediction markets, event derivatives, trading firms, exchanges, and institutional participants. Our audience includes decision-makers from trading desks, market operators, research providers, and fintech platforms. If your firm would like to engage with a sophisticated readership at the intersection of prediction markets and financial innovation, we offer high-visibility sponsorship opportunities in both the newsletter and future reports.

📩 Inquire at [email protected] for details, audience metrics, and pricing.

DISCLAIMER

Prediction Markets Alert is for educational and informational purposes only. Nothing in this newsletter is investment, trading, or betting advice, or a recommendation to buy, sell, or participate in any market or contract. Probabilities shown are market prices, not predictions or guarantees. Markets can be — and often are — wrong. Prediction market platforms are not available or legal in all jurisdictions. It is your responsibility to verify eligibility and legality where you live before using any platform. PMA receives no compensation from any platform, market, or company mentioned in this newsletter. All figures are verified against live markets as of send time and will change after publication. If you or someone you know has a gambling problem, help is available: call or text 1-800-GAMBLER.