The Fed dot plot — how to read interest-rate projections

The Fed dot plot summarizes individual assessments of appropriate policy rates under participants’ economic assumptions. It helps you read the distribution, but supplies neither a Committee timetable nor a dollar forecast. The key exercise is to calculate the median and dispersion separately, keeping the horizon and source precision.

What the dot plot is and where to find it

The chart belongs to the Summary of Economic Projections, alongside growth, unemployment and inflation assessments. It is collected in connection with March, June, September and December meetings. March/June projections cover the current and next two years; September/December add a third subsequent year. Check the actual release headings, counts and omissions for each horizon rather than assuming an unchanging list.

Each participating Board member and Reserve Bank president submits an assessment of the appropriate policy rate for the years shown and the longer run. The participant count and projection horizon should be checked in the release rather than assumed. For example, the September 2026 SEP covers 2026–2029 and the longer run. The anonymous dots show how many submitted assessments fall at each rate level, without identifying the individual policymakers.

A dot denotes the year-end midpoint of an assessed appropriate target range or the target level, rounded in the chart to the nearest one-eighth percentage point. It is not the yearly average or upper bound. A hypothetical 3.50–3.75% range has midpoint 3.625%. The September 2026 footnote records 18 submitting participants and one omission for 2028 and 2029; the nominal full roster is not the count of every column.

The median, or the middle dot

For each horizon, sort the submitted projections before calculating the median. With nineteen values, it is the tenth ordered value; with an even number, it is the mean of the two middle values. Check the release’s actual participant count instead of assuming every chart has nineteen dots. Also inspect the distribution around the median.

Our seven-value model is fictional, not a Fed meeting. Set A is 3.125, 3.375, 3.625, 3.625, 3.875, 4.125, 4.375%; B replaces one 3.625 with 3.875. Median, the fourth value, rises 3.625% → 3.875%, or 25 basis points. Both ranges remain 3.125–4.375%, width 125 basis points. Only two of seven observations equal each median: the median is not necessarily a majority’s chosen rate.

In a separate even-count example, middle values 3.625% and 3.875% average to 3.750%, which need not be anyone’s dot. Moving an extreme in A from 4.375% to 5.125% leaves its median 3.625% but widens the range to 200 basis points. This example shows one stable median; it does not prove medians are immune to all changes or always lie at a cluster.

The dispersion of the dots shows how far participants’ assessments differ for a given horizon. A narrow range indicates similar submitted rate assessments, while a wider range shows more disagreement. Neither proves that future decisions will follow a stable path: participants can revise their views as the outlook changes, and the chart is not a policy commitment.

The longer-run dot, or the neutral rate

The “longer run” column is not a forecast for a particular year. It records each participant’s assessment of where the policy rate would converge under appropriate policy and without further shocks. Its median is a reference point for those individual assessments, not a rate the Committee has jointly promised to reach.

The 2016 Fed staff note interprets longer-run nominal assessments as estimates of longer-run neutral rates under its conditions, while distinguishing shorter-horizon neutrality. They are not observed constants, current real neutral rates or a promised destination. Subtracting today’s nominal rate from this column alone does not measure current restraint.

Changes in the longer-run median can reveal a shift in participants’ assessments of the eventual policy-rate level. Compare successive SEP releases and their ranges before drawing that conclusion. A higher US assessment alone does not guarantee lasting dollar appreciation; foreign rates, risk conditions and what markets already expected also matter. See the fundamental analysis material for the wider context.

Why the dots are projections, not promises

Projections reflect individual economic assumptions and policy assessments, not votes or a binding timetable. The 12 June 2024 conference explicitly separated them from a Committee plan or decision. New information can change assessments; the document does not guarantee a particular speed or direction of revision.

Retain historical definitions as well as dates. The December 2023 summary table showed a rounded 2024 federal-funds median of 4.6%; the June 2024 decision still held the target at 5.25–5.50%. A year-end projection never fixed particular earlier meeting dates. For a separate fictional exercise, current midpoint 4.375% minus set A median 3.625% gives 75 basis points, equivalent to three 25-basis-point steps under that assumed step size, without determining dates or a sequence.

Projection participants and policy voters are different groups. The FOMC normally has twelve voting members: the seven Board governors, the president of the Federal Reserve Bank of New York and four other Reserve Bank presidents serving rotating terms. All Reserve Bank presidents participate in discussions, but not all vote at every meeting. Vacancies or absences can reduce the number of projections or voters.

Match horizons and keep individual paths unknown

Compare the same year across releases, with its participant count, median and dispersion; the first column’s year can change. FedWatch uses futures contracts for monthly-average EFFR and assumptions to derive probabilities. A dot is an individual year-end policy assessment. Matching a date alone does not make these interchangeable; retain market information recorded before publication.

Anonymous columns cannot recover individual paths. In our fictional three-person model, first-year projections are 3%, 4%, 5%, next-year projections 2%, 3%, 4%; medians fall 4% → 3%. Assignment A pairs 3→2, 4→3, 5→4, all down one percentage point. Assignment B pairs 3→4, 4→3, 5→2, one up and two down. Both give identical columns. Joining sorted ranks cannot identify the same person, so a declining median does not show everyone expects cuts. Neither set predicts USD, bonds or equities.

Compare the projections with the Chair's press conference, which is scheduled separately from the statement and projection release. Kevin Warsh became Board Chair on 22 May 2026; Powell's earlier remarks remain historical material. Verify the current speaker and event time on the Federal Reserve website, then separate policy explanations from your pre-defined trading rules.

Your next step

1. Copy definitions and counts. Record release date, exact year, middle pair where needed, minimum and maximum. Dispersion is not a confidence interval or vote.
2. Keep source precision. Do not invent missing digits from rounded summary numbers. The diagram’s shorthand error is a calculation effect, not a policy change.
3. Check the calendar. Identify the four SEP meetings and separate communication times; publication does not guarantee elevated volatility. Relate them to other releases such as CPI inflation.
4. Separate interpretation from action. Read statement and later answers with their conditions, match market measures and keep contradictory cases. A fixed wait is not a guarantee of low risk.