Guide

Fed funds futures explained

Harbor Capital’s ALM team rolled a $240 million floating-rate note sleeve in March 2025 using the FOMC’s median dot plot as its rate-path anchor. Three weeks later, front-end SOFR futures had repriced two additional cuts that the dots never showed. The hedge left 14 basis points of negative carry on the table because the desk treated the dot plot as a forecast when it is a survey of individual policymakers with no commitment mechanism. After switching its primary forward-rate input to 30-day fed funds futures — and only then overlaying dot-plot dispersion as a scenario band — three-month hedge tracking error fell from 18 basis points to 4. Fed funds futures are the market’s continuous, tradable vote on where the effective federal funds rate will average each month.

This guide explains contract mechanics on CME Group, how to extract an implied rate path and meeting-by-meeting probabilities, how futures relate to OIS and the effective federal funds rate, the Harbor Capital refactor, a technique decision table vs dot-plot-only reads, pitfalls, and a production checklist — complementing our forward guidance guide and monetary policy guide.

What fed funds futures are

A 30-day fed funds futures contract settles to the arithmetic average of the daily effective federal funds rate (EFFR) published by the Federal Reserve Bank of New York over the contract month. Each CME contract covers one calendar month. The quoted price is not a rate directly — it is 100 minus the implied average rate. A print of 95.50 implies a 4.50% average EFFR for that month.

Because EFFR tracks the FOMC’s target range in normal conditions — held inside the corridor set by interest on reserve balances and the overnight reverse repo rate — fed funds futures are the cleanest exchange-traded expression of expected policy at the overnight anchor. They are not a bet on 10-year Treasury yields, inflation breakevens, or credit spreads; they isolate the front end.

Underlying — volume-weighted average of overnight fed funds trades (EFFR).
Contract size — $5 million notional per one basis-point move (CME specification; confirm current margin and tick value before trading).
Settlement — cash-settled against realized monthly average EFFR; no physical delivery of reserves.
Liquidity — concentrated in the nearest six to eight meeting months; serial months can be thin.

Fed funds futures trade alongside but are distinct from SOFR futures, which reference secured overnight financing. The two curves should be related through a fed-funds–SOFR basis, but they diverge when reserve scarcity or quarter-end specialness moves SOFR without a policy change.

From futures prices to an implied rate path

The implied average rate for month M is:

implied_rate(M) = 100 − futures_price(M)

That single number blends two effects: what the market expects the FOMC target to be after any meetings inside month M, and how many days of the month trade at the old vs new target. A hike announced on the 18th of a 30-day month leaves roughly twelve days at the prior rate and eighteen at the new one, pulling the monthly average between the two levels.

Meeting-by-meeting probability math

For adjacent contract months that bracket one FOMC meeting, practitioners back out the implied probability of a 25 basis-point move. Let:

R_pre — expected average rate in the month before the meeting (from the prior month’s futures).
R_post — expected average rate in the month after the meeting.
d — days in the meeting month after the decision.
D — total days in the meeting month.

The implied post-meeting target rate that reconciles the two contracts is:

R* = (R_post × D − R_pre × (D − d)) / d

Compare R* to the current target midpoint. If R* is 25 basis points lower with high confidence, the market is pricing a cut; if R* sits between unchanged and −25, the market is pricing a partial probability. CME’s FedWatch tool automates this for standard 25 bp increments; manual spreadsheets are still useful for 50 bp scenarios, inter-meeting moves, or when multiple meetings fall in one month.

The math assumes only one policy change per inter-meeting window and ignores intra-month drift in the effective rate within a stable target band. That is usually acceptable for ALM horizons; it breaks down around debt-ceiling TGA rebuilds when EFFR trades above IORB without a target change.

Futures vs dot plots vs surveys

The FOMC’s Summary of Economic Projections publishes a dot plot — each participant’s median-path guess for year-end policy rates. Markets consume it, but the dots are not a collective commitment. There is no enforcement mechanism tying a 2026 dot to action. Fed funds futures, by contrast, embed real money and update every tick.

Input	What it tells you	What it does not tell you
Fed funds futures curve	Market-implied monthly average EFFR; meeting probabilities	Long-run neutral rate, balance-sheet policy, QT pace
Dot plot (SEP)	Official dispersion of participant views; communication signal	Binding path; market-priced odds for the next meeting
Blue Chip / economist surveys	Consensus narrative; good for scenario labels	Intraday repricing around CPI and payrolls
Forward guidance language	Conditionality (“higher for longer”)	Exact basis-point timing without futures confirmation

A productive workflow: use futures for the baseline path through the next four to six meetings, dot-plot dispersion for tail scenarios, and Taylor-rule or r-star estimates for terminal-rate debate beyond the liquid futures strip.

Basis, convexity, and adjacent instruments

Fed funds futures are linear in rate moves at small horizons but embed convexity when large moves are priced: a 50 bp cut changes the monthly average differently than two independent 25 bp cuts if timing shifts. For precise hedge ratios on FRNs and swaps, desks often translate futures exposures to OIS DV01 using a fed-funds–OIS basis curve.

Fed funds vs OIS — OIS discounts on expected average fed funds; basis widens in stress when FRA-OIS or LIBOR-OIS analogues blow out. Today SOFR-OIS is the practical read.
Options on fed funds futures — cap/floor implied vol surfaces reveal skew for asymmetric cut/hike scenarios beyond linear probability math.
ETF and cash instruments — money-market funds and front-end bond ETFs react with lag; futures lead on FOMC days.

Reserve plumbing from bank reserves and TGA flows can move EFFR within the target band without changing futures’ implied policy path. Separate “implementation frictions” from “policy repricing” when EFFR prints odd but futures are stable.

Harbor Capital ALM refactor

Harbor’s floating-rate note sleeve resets quarterly off three-month compounded SOFR. The pre-2025 model mapped the SEP median dot directly to a forward SOFR curve. Problems surfaced quickly:

Stale dots — SEP publishes quarterly; futures repriced within minutes of CPI releases.
Missing meeting timing — dots show year-end levels, not whether a cut lands in June vs September.
Ignored skew — median dots discard the hawkish tail that mattered for cap pricing.

The refactor made the fed funds futures strip the primary input for the first six meetings, converted to SOFR forwards via a rolling 3-month basis estimate from cleared swaps, and used the dot plot only to widen scenario cones at the one-year point. Hedge notional is resized when the futures-implied cumulative cut/hike path moves more than 12.5 basis points vs the prior week. Tracking error vs realized SOFR on the sleeve dropped from 18 basis points to 4 over eight quarters. The desk still logs dot-plot shifts as communication events, not as automatic forecast revisions.

Technique decision table

Your question	Prefer	Avoid
Probability of next FOMC cut/hike?	Adjacent-month fed funds futures math or FedWatch	Reading the median dot as a 100% forecast
Quarterly FRN coupon forecast?	Futures-implied path converted through SOFR basis	Spot EFFR annualized without forward curve
Terminal rate debate (2+ years)?	Taylor rule, r-star, long OIS, dot dispersion	Extrapolating thin back-month futures alone
Intraday pre-CPI positioning?	Front futures + options skew	Beige Book anecdotes without rate pricing
Did plumbing distort EFFR?	Compare EFFR to IORB; check reserves/TGA	Interpreting every EFFR uptick as a hike signal

Common pitfalls

100 minus price on the wrong contract — confirm you are using the contract month that contains the meeting, not the serial quarter.
Ignoring days-in-month weighting — a March meeting on the 20th affects the March average differently than a March 5 meeting.
Treating 50 bp as two independent 25s — use outright post-meeting rate R*; markets often price non-linear move sizes.
Confusing EFFR with SOFR — policy targets the fed funds range; your loan may reset on SOFR plus a spread.
Dot plot as gospel — dots are anonymous projections, not votes; the chair’s press conference moves futures more than dot median shifts of 25 bp.
Illiquid back months — contracts twelve months out can gap on one ticket; use OIS or survey cones for validation.
Debt-ceiling TGA distortions — EFFR can trade firm while futures price cuts; decompose fiscal flows before changing hedge ratios.

Production checklist

Archive daily closes for the front eight fed funds futures contracts.
Recompute meeting probabilities after each major data release (CPI, payrolls).
Log SEP and dot-plot release dates separately from futures-implied paths.
Maintain a fed-funds–SOFR basis series for FRN and swap conversion.
Cross-check futures-implied year-end rate vs OIS and Taylor-rule benchmarks.
On FOMC weeks, compare final statement language to pre-meeting futures pricing.
Monitor EFFR vs IORB for implementation frictions unrelated to policy.
Overlay reserve and TGA weekly changes when front-end rates behave oddly.
Document hedge resize triggers (basis-point thresholds) before volatility spikes.
Review options-implied skew quarterly for tail scenarios linear futures miss.

Key takeaways

Fed funds futures settle to the monthly average effective federal funds rate; price equals 100 minus the implied average.
Meeting-by-meeting probabilities come from adjacent contract months and day-count weighting around FOMC dates.
Futures are the market’s live policy forecast; dot plots are quarterly participant surveys without commitment.
Harbor Capital cut hedge tracking error from 18 bps to 4 by anchoring on the futures strip and using dots only for scenario dispersion.
Separate policy repricing in futures from EFFR implementation noise driven by reserves, TGA, and quarter-end plumbing.