How CME FedWatch Probabilities Work: From Fed Funds Futures to FOMC Outcomes
Learn how CME FedWatch uses 30-Day Fed Funds futures, monthly average EFFR, non-meeting-month anchors, and stated assumptions to estimate FOMC rate outcomes.
Direct answer
CME FedWatch starts with 30-Day Federal Funds futures, whose named monthly contracts reflect market pricing for average daily EFFR over their contract months. CME then applies a published set of calendar and rate-change assumptions to estimate probabilities for scheduled FOMC meeting outcomes. The output is an estimate under those assumptions, not a directly observed probability or a guaranteed policy result.
Start with a named monthly Federal Funds futures contract
The input is not a generic “Fed rate” quote. A 30-Day Federal Funds futures contract has a named month and contract-defined calculation. CME's methodology describes the price as incorporating market expectations of the average daily EFFR during that futures contract month.
That monthly window is essential. An FOMC meeting may take place after some days of the month have already occurred and before other days remain. The monthly futures price therefore does not simply identify one meeting outcome without a calendar-aware interpretation.
What the CME FedWatch Tool means explains what a displayed meeting probability is before this guide explains how CME derives it.
Read 100 minus price as a monthly implied-average rate
CME's methodology uses the convention that a named 30-Day Federal Funds futures price can be read as 100 minus the implied average EFFR for that contract month. This is a product-defined calculation convention. It is not a statement that the price equals today's EFFR, a target-rate midpoint, or the outcome of the next meeting.
The New York Fed defines EFFR as a volume-weighted median of overnight federal funds transactions and publishes it for the prior business day. That published rate belongs to a different time record from a future month's implied average. Do not substitute a spot EFFR value for a monthly futures input or use a rounded screen quote without retaining its field and timestamp.
SOFR futures price and implied rate shows the same general 100-minus-rate convention in a different contract family, where the reference period and calculation rule are not the same.
Use the non-meeting-month anchor in CME's methodology
CME's published methodology begins from a full month without an FOMC meeting. For such an anchor month, it uses the implied monthly average EFFR to connect the end-of-month rate before it and the start-of-month rate after it. From there, the method works through surrounding meeting months.
This step is why a FedWatch probability cannot be reproduced by pointing to one price and saying “100 minus that number.” The method uses the named contract months, their calendar relationship to scheduled meetings, and an order of operations for propagating implied rates across the timeline.
The contract month is a product-defined label, not a universal policy calendar. Futures contract month codes is a useful companion when a short symbol has obscured the underlying month-year record.
The probability tree depends on explicit rate-change assumptions
To translate expected EFFR changes into meeting-outcome estimates, CME's methodology makes specific assumptions. Among them are rate moves in uniform 25-basis-point increments, a proportional EFFR response, a published schedule of FOMC meetings, and an EFFR lower bound of zero. It also describes an unconditional probability tree built from those inputs.
These assumptions make the output readable, but they are not facts about the future. CME states that calculated probabilities are estimates and can vary if an assumption is not met. A table of possible outcomes should therefore be read as a model result with conditions, not as a measurement of an unknown decision that has already happened.
Preserve the input time, the meeting, and the methodology version
A careful record includes the selected FOMC meeting, the futures contract month, the exact price field and observation time, the displayed outcome probabilities, and the methodology source. Use the same observation time for all contracts in a multi-meeting comparison when possible; otherwise label the time mismatch rather than treating it as an insignificant detail.
The output also needs a separate comparison from the FOMC's own published projections. FedWatch versus the FOMC dot plot distinguishes the market-price model from the Summary of Economic Projections. For the underlying distinction between SOFR and EFFR futures, see SOFR futures versus Fed Funds futures.
This guide describes a public methodology for interpreting a market-data tool. It does not estimate a current FOMC result, endorse the assumptions, or recommend trading, hedging, or changing a portfolio. Official policy material, current exchange documentation, and current market data should be consulted for their separate purposes.
Common questions
Does FedWatch use one futures contract for each FOMC meeting?
It uses 30-Day Federal Funds futures tied to named monthly periods and applies calendar relationships in its methodology. A meeting outcome is not simply the same object as one monthly contract price.
Why does the methodology use a month without an FOMC meeting?
CME uses a non-meeting month as an anchor for connecting implied monthly average EFFR values with adjacent start- and end-of-month rate estimates. That is part of the stated calculation process.
Is 100 minus the futures price the next target rate?
No. In this context, the convention gives an implied average EFFR for a named contract month. It is not automatically today's EFFR, the target rate, or a single meeting result.
Why are 25-basis-point assumptions important?
The assumptions determine how CME turns an expected EFFR change into discrete possible meeting outcomes. If the assumptions or underlying pricing differ, the calculated probability output can differ too.
Can a FedWatch probability remain unchanged until a meeting?
There is no such guarantee. The tool is based on market pricing and a stated methodology, so its output can change as the relevant inputs or observation time change.