Expectations for the next five meetings
| Meeting | 2.25%current rate | 2.50%+25 bp | 2.75%+50 bp | 3.00%+75 bp | 3.25%+100 bp |
|---|---|---|---|---|---|
| 10 September 2026in 26 days | 6% | 94% | — | — | — |
| 29 October 2026in 75 days | 5% | 78% | 17% | — | — |
| 17 December 2026in 124 days | 3% | 43% | 46% | 8% | — |
| 4 February 2027in 173 days | 2% | 37% | 46% | 14% | 1% |
| 18 March 2027in 215 days | 2% | 28% | 44% | 22% | 4% |
Calculated from futures prices, as of 14 August 2026
How often were markets wrong?
Just before a meeting, markets are almost always right; months out, far less often. The comparison with the actual rate path shows how big that gap really is.
To the review →The 10 September 2026 meeting
How the expectation for this date has shifted over recent weeks, and which rate level could still be likely.
To the detail page →The expected rate path
The rate path markets expect over the coming months.
To the rate path →Methodology
Eight times a year, the Governing Council of the European Central Bank decides on policy rates. What this site shows is the deposit facility rate, the most important of the three, because overnight and money-market rates follow it. And it does not show what the ECB will do, but what markets are betting on.
That expectation sits in the prices of futures on €STR, the euro overnight rate. Anyone buying or selling one commits to a rate and puts their own money behind it.
Each of those contracts covers the period between two Governing Council meetings. The price difference between the contract before and the one after a meeting therefore shows how much of a rate change the market has priced in for that specific meeting; divided by the usual step size of 25 basis points, that yields the probability of a move. For later meetings, the distributions of all meetings in between are chained together. Updated twice per trading day. Full methodology →
It is still not reliable. Markets are wrong regularly, and the further out the meeting, the more often. Exactly how often is in forecast vs. reality, where every past expectation sits next to the actual rate path.