Fed Rate Cut Odds: How Many Cuts In 2026, According To The Market
Five FOMC meetings into 2026, the Fed has not cut once, and Kalshi's $6.7 million "how many cuts this year" market has all but called the year over: exactly zero cuts trades near 87 cents. We put the same question to eight AI models, price-blind, with a fetched dossier of every 2026 Fed decision, the July statement's dissent pattern, and the latest inflation and jobs prints. After a revision round in which each model read the others' anonymized reasoning, the panel came back far less certain than the crowd: 41% on zero cuts, and one cut as its median path. That 46-point gap on a $6.7 million macro board is the whole article.
The Quick Answer
Kalshi prices exactly zero 2026 rate cuts at roughly 87 cents, with exactly one cut at 10 cents and everything else in the rounding. Our eight-model, price-blind panel blends to 41.1% on zero, 28.9% on one, and 17.1% on two, which makes at least one cut a 59% proposition against a market that pays it like a longshot. The market's median outcome is no cuts at all; the panel's median outcome is one. The full count board, the seat-by-seat numbers, and the argument that moved two models in opposite directions during the revision round are all below.
The Market At A Glance
| Venue | Kalshi, a CFTC-regulated event-contract exchange (18+; availability varies by state, as of August 2026) |
| The Event | "How many times will the Fed cut rates in 2026?" (Kalshi event KXRATECUTCOUNT-26DEC31), 21 exact-count contracts from 0 through 20, about $6.7 million traded |
| The Contract | Per the RATECUTS contract terms, the count is the number of downward changes in the upper limit of the fed funds target range in 2026, as documented by the FOMC itself: up to 25bps is one cut, a move larger than 25bps counts as two, larger than 50bps as three. Revisions made after expiration do not count. |
| Settles | December 31, 2026 |
| Prices Below As Of | August 1, 2026 (live quote midpoints; sub-cent longshots from last trades) |
These are model estimates, not predictions of fact and not financial advice. Kalshi event contracts trade on a CFTC-regulated exchange; you must be 18 or older and in an eligible state to participate.
Want the re-scored board after the September Fed meeting? The FOMC meets September 15-16, and the panel re-runs this exact distribution when the decision lands. The email signup at the end of this page gets the updated board first.
The Meeting-By-Meeting Path
The structure of this market is a calendar problem before it is an economics problem. The Federal Reserve's 2026 meeting calendar shows five meetings already held (January 27-28, March 17-18, April 28-29, June 16-17, July 28-29) and three remaining: September 15-16, October 27-28, and December 8-9. Every 2026 meeting so far has ended in a hold, with the target range parked at 3-1/2 to 3-3/4 percent since the December 2025 cut.
That calendar does the heavy structural lifting:
- Zero, one, two, or three cuts are reachable with ordinary 25bp moves at the remaining scheduled meetings.
- Four or more cuts requires at least one 50bp move (which the contract counts as two) or an unscheduled intermeeting action. The contract counts changes to the target range, not meetings, so a crisis response can add multiple counts at once.
- A September hike, which the Fed's own July dissent pattern put in play, would not subtract from the count. The contract counts cuts only.
The July statement is the tell the panel leaned on hardest. It kept the "inflation remains elevated" language, attributed part of the problem to supply shocks "in certain sectors, including energy," and drew three dissents in favor of a quarter-point hike from Beth Hammack, Neel Kashkari, and Lorie Logan. Contrast January, when the two dissents (Stephen Miran, Christopher Waller) pushed the other way, for a cut. Over six months the committee's argumentative wing flipped from dovish to hawkish.
Against that sits the data the Fed is holding through: June CPI at 3.5% year over year, down from 4.2% in May, with core easing to 2.6%; and a June jobs report of just +57,000 payrolls, with April and May revised down a combined 74,000 and the trailing-year average near +36,000 a month. Falling inflation, stalling payrolls, hawkish dissents. Both sides of this market have real material.
The Full Count Board
Here is the whole distribution: the market's price for each count, the panel's blended probability, and the gap. The blend is the equal-weight mean of the eight models' round-two numbers, generated after each model read the others' anonymized round-one reasoning and could revise.
| 2026 Cuts | Market Price | AI Blend | Gap (blend% − mkt¢) |
|---|---|---|---|
| Exactly 0 | 86.6¢ | 41.1% | −46 |
| Exactly 1 | 10.3¢ | 28.9% | +19 |
| Exactly 2 | 2.7¢ | 17.1% | +14 |
| Exactly 3 | 0.9¢ | 7.2% | +6 |
| Exactly 4 | 0.7¢ | 3.1% | +2 |
| Exactly 5 | 0.2¢ | 1.5% | +1 |
| 6 or more (15 contracts) | ~1.5¢ combined | 1.1% | 0 |
These are model estimates, not predictions of fact and not financial advice. Market figures are Kalshi quote midpoints as of August 1, 2026; the 6-and-up bucket sums fifteen deep-longshot contracts that only trade on thin last prices. All 21 contract prices sum to about 102.7 cents, so the raw prices run slightly rich of true probabilities.
Every Seat's Number
| 2026 Cuts | Fable | Opus | Sonnet | GPT | Gemini | GLM | Kimi | DeepSeek | Blend |
|---|---|---|---|---|---|---|---|---|---|
| Exactly 0 | 43.0 | 38.2 | 35.0 | 40.0 | 45.0 | 38.0 | 44.0 | 46.0 | 41.1 |
| Exactly 1 | 28.0 | 28.1 | 32.0 | 32.0 | 30.0 | 30.0 | 27.0 | 24.0 | 28.9 |
| Exactly 2 | 17.0 | 18.6 | 17.0 | 17.0 | 15.0 | 18.0 | 16.0 | 18.0 | 17.1 |
| Exactly 3 | 7.0 | 9.6 | 8.0 | 7.0 | 6.0 | 7.0 | 7.0 | 6.0 | 7.2 |
| Exactly 4 | 3.0 | 3.5 | 4.0 | 2.5 | 2.5 | 3.0 | 3.5 | 3.0 | 3.1 |
| Exactly 5 | 1.2 | 1.2 | 2.5 | 1.0 | 1.0 | 2.0 | 1.5 | 1.5 | 1.5 |
| 6 or more | 0.8 | 0.8 | 1.5 | 0.5 | 0.5 | 2.0 | 1.0 | 1.5 | 1.1 |
These are model estimates, not predictions of fact and not financial advice. The panel: Claude Fable, Claude Opus, Claude Sonnet, ChatGPT (GPT-5.5), Gemini 3.1 Pro, GLM 5.2, Kimi K3, and DeepSeek V4, each pricing the full distribution in a single pass on August 1, 2026, without seeing any market prices. All eight seats returned coherent distributions in both rounds, and every number lands in our back-tested Model Verdict ledger, where past price-blind calls are graded against settled results.
Zero Cuts At 87 Cents: The Bucket The Market May Overprice
The market's favorite is not a mystery. The Fed has held five straight times, and it held in July while three of its own members argued for a hike. Every model treated that dissent trio as the single loudest fact on the card. As Kimi K3 put it, "the July debate was hold-vs-hike, not hold-vs-cut." Claude Fable's version: "committees rarely swing from debating a hike to delivering a cut inside two meetings."
So why does the panel still put zero cuts at only 41%? Because three meetings is a long runway for data this soft. The case, in the models' own words:
"The Fed's own reaction function keys off the unemployment rate, not the payroll pace, and that number is not screaming." — Claude Opus
"By December, two more payroll prints in the +40-60K range plus continued core CPI cooling toward 2.5% give the doves enough cover for a single insurance cut alongside fresh projections." — Claude Sonnet
Most of the panel's one-cut mass is concentrated in December, the meeting with fresh projections and four more payroll prints of visibility. The seats split on whether October is live at all: several treated a no-projections meeting as one the committee skips, while Claude Opus pushed back that the Fed's October 29, 2025 cut came at exactly such a meeting, so the calendar excuse only goes so far. The market says the whole cut window basically does not exist. The panel says December alone is nearly a 3-in-10 outcome.
One And Two Cuts: Where The Panel Sees Daylight
The two biggest positive gaps on the board are exactly one cut (+19 points against a 10-cent price) and exactly two (+14 against about 3 cents). The two-cut number is not a "cuts in September and October" story; almost none of the models believe that. It runs through a mechanism the revision round surfaced and spread across the panel: a Fed that waits too long tends to move bigger when it finally moves, and this contract counts a single 50bp move as two cuts. Kimi K3 made the case in round one: "a hawkish committee that waits too long with unemployment drifting toward 4.5-4.6% tends to move bigger when it finally moves, and one meeting can deliver two contract-units." DeepSeek, revising upward in round two, credited exactly that argument, "a path I had underweighted by spreading mass across two-meeting 25bp sequences."
The revision round is worth pausing on, because it moved the two most extreme seats toward each other from opposite directions, each citing a specific argument rather than the group average:
"I significantly walked back my extreme 75% probability on zero cuts... I had underpriced the possibility that the center of the committee could pivot by December if the ~36k/month payroll trend persists." — Gemini 3.1 Pro, whose zero-cut number went from 75 to 45
"I raised 0 from 14% to 38% and cut 2 from 31% to 18%... the Fed saw the weak June jobs report and favorable CPI BEFORE its July meeting and responded with three hike dissents — a revealed reaction function I had underweighted." — GLM 5.2, moving the other way
Two models entering from opposite ends and converging on roughly 40/30/17 is the panel's real headline: the distribution, not any single count, is the answer.
Dark Horses The Panel Won't Dismiss
Per house rule, the board below the favorites, and what would have to happen:
- Three cuts (7.2% blend vs a 0.9¢ price). A real autumn labor break, with the Fed starting late and going meeting-by-meeting. Claude Opus is the panel's high number at 9.6: the already-turning core "can produce three consecutive 25s once they start, and starting late means going fast."
- Four cuts (3.1% vs 0.7¢). Requires a 50bp move plus follow-through. GLM sketched the path: "a 50bp move at September (counting as 2) plus two 25bp cuts is the realistic path if payrolls go negative."
- Five-plus (2.6% combined vs under 2¢). Pure crisis mechanics: multiple 50s or intermeeting action. As DeepSeek put it, if the labor data cracks hard, "the Fed would move fast and could deliver a 50bp cut that counts as two — history says that scenario fat-tails."
None of these are calls. They are the panel refusing to price the tail at zero while the market rounds it away.
The Trap: What The Market Might Know
Honesty requires steelmanning an 87-cent favorite. Kalshi's per-meeting September market, which we cover on our standing Fed September 2026 board, was pricing a September hike as more likely than a hold when we pulled prices on August 1 (about 57 cents on a 25bp raise). The count market and the meeting market agree with each other: traders think this Fed's next move is up, not down, presumably on the same energy-supply-shock inflation the July statement named. Our panel worked from a card whose data window closed with the July 29 statement. If August Fed-speak or an energy re-spike has hardened the hike case since, the market is pricing information the models never saw, and the honest read of the 46-point gap shrinks accordingly. What the gap cannot be, on the panel's numbers, is fully justified: even the most hawkish seat (DeepSeek at 46% zero) prices "no cuts ever in 2026" at barely half the market's confidence.
A hike, if it comes, does not settle this market by itself. The count runs through December 31, and cuts after a hike still count.
What Moves This Market Between Meetings
This board reprices on data, not just decisions. Each monthly CPI release and each jobs report between now and December is a live event for every bucket, which is why the panel's own reasoning kept returning to the same two series we already score on standing boards: the CPI verdict board and the jobs-report verdict board. The June prints that fed this panel (CPI released July 14, jobs released July 2) frame the tension: headline inflation fell for the first time in five months on an energy drop, while payrolls came in at +57,000 against roughly +115,000 expected. Whether the Fed reads that as disinflation progress or as a supply-shock head-fake is the entire zero-versus-one argument.
Count Markets Vs Meeting Markets: Reading Both
The annual count market and the per-meeting decision markets answer different questions, and their prices discipline each other. Zero cuts at 87 cents implies three consecutive cut-free meetings; if you multiply through what the per-meeting markets imply about September, October, and December, the products should roughly agree, and when they do not, one of the boards is stale. That is the same cross-checking habit our permanent Fed decision board applies meeting by meeting. The count market's advantage is that it forces a full-year view: it cannot dodge the December question the way a September-only trader can.
How To Read These Prices
Worked Example: What An 87-Cent "Zero Cuts" Contract Pays
Buy YES on "exactly 0 cuts" at 86.6 cents and you collect one dollar if the Fed finishes 2026 without a cut: a 13.4-cent gross profit, minus Kalshi's trading fees, which scale with how close a price sits to 50 cents. Implied break-even: the contract must win about 87% of the time. The panel says 41%. If the panel is right, the same dollar is far better spent on the other side of the distribution, where one cut at 10.3 cents pays roughly 9-to-1 against a blended 28.9% probability. If the market is right, the panel is underweighting a hawkish regime shift. That is the disagreement, stated as prices.
Three honest cautions before anyone acts on a number above:
- The prices sum rich. All 21 contracts add to about 102.7 cents, so each raw price mildly overstates its true implied probability; re-scale before comparing to the blend.
- The longshot rungs are thin. Everything from three cuts up trades on sparse books at sub-cent last prices; entry and exit at the screen price is not promised.
- Capital is parked until December 31. This contract settles at year-end no matter how obvious the answer becomes in September; early certainty comes only from selling at the then-current price.
Kalshi is a CFTC-regulated event-contract exchange, not a sportsbook; you must be 18 or older and located in an eligible state, and availability varies by state. If the event-contract format is new to you, the mechanics live in our guide to how to bet sports on Kalshi. For everyday sports prices, the same line-reading discipline runs through our live odds screen and free expert picks today; the Fed shopping itself happens on Kalshi. These are model estimates, not predictions of fact and not financial advice.
Re-Score Schedule
This page is the permanent home for the 2026 cut-count question, and it re-scores on a fixed calendar: after the September 15-16 decision, after October 27-28, and after December 8-9, with each prior cycle folded into a track-record section. When the market settles on December 31, the settlement retrospective lands here too, graded against these exact numbers. The signup box below gets each re-score first.
FAQ
How many FOMC meetings are left in 2026? Three. The Federal Reserve's calendar lists September 15-16, October 27-28, and December 8-9. The first five meetings of 2026 (January, March, April, June, July) all ended in holds, with the target range at 3.50-3.75% since December 2025. September and December include updated economic projections.
How does the Kalshi rate-cut count market settle? Each contract pays out on the exact number of cuts from January 1, 2026 through year-end, counted by the FOMC's own changes to the target range. Per the contract terms, a cut of up to 25 basis points counts as one cut, and a 50-basis-point move counts as two, so a single meeting can add two to the count. Revisions made after expiration are not counted.
The Bottom Line
The market looked at five straight holds and three hike dissents and called the year: 87 cents on zero cuts, with its sister market leaning toward a September hike. The panel looked at the same statement plus 3.5%-and-falling inflation, +57,000 payrolls, and a December meeting with fresh projections, and refused to call it: 41% zero, 29% one, 17% two, with the December insurance cut as the median path. One of these two readings is meaningfully wrong, and unlike most macro arguments, this one grades itself by New Year's Eve. To be explicit one final time: these are model estimates, not predictions of fact and not financial advice.
Hottest Prediction Markets Right Now
- 2028 Democratic presidential nominee · $166M traded
- Rocket Classic Winner · $85M traded
- 2027 Pro Football Champion · $52M traded
- Pro Baseball Champion · $41M traded
- Bitcoin price at the end of 2026 · $30M traded
- 2027 Pro Basketball Champion · $19M traded
Every market above links to our full AI model verdict; browse them all on the OddsShopper prediction markets hub.
Verdicts generated August 1, 2026, price-blind, revised once after an anonymized cross-read. Prices as of August 1, 2026 (Kalshi quote midpoints, fetched 12:14 UTC). Panel: Claude Fable, Claude Opus, Claude Sonnet, ChatGPT (GPT-5.5), Gemini 3.1 Pro, GLM 5.2, Kimi K3, DeepSeek V4.
