The Federal Open Market Committee (FOMC) meets September 15-16, 2026, with the policy decision and a fresh Summary of Economic Projections (SEP) due Wednesday, September 16, per the committee’s published calendar. The federal funds target range stands at 3.50% to 3.75%. What separates this meeting from the two before it is the dot plot: the June SEP lifted the median year-end 2026 projection to 3.8%, above the current 3.625% midpoint, with nearly half the committee marking at least one additional hike. September 16 is the first test of whether that median survives.
For organizations that discount long-tail claim liabilities, the SEP is a reserve input, not a markets story. Year-end reserve studies are being scoped right now, and the discount rates they carry will be benchmarked against whatever rate path the September projections confirm or unwind. The April 29 meeting produced four dissents, the widest committee split since October 1992, so the path is genuinely contested inside the room. A single-point discount rate carried forward from last year’s study is the exposure, and the cut-driven rate decline many 2026 models penciled in never arrived.
Who it affects
Captives reporting discounted reserves under statutory accounting, public entities following Governmental Accounting Standards Board (GASB) standards, and large self-insured employers reporting under generally accepted accounting principles (GAAP) whose actuaries present both nominal and discounted estimates. The effect concentrates in the longest-duration books: workers’ compensation lifetime medical and pension cases, excess and umbrella casualty, and hospital professional liability held through a captive. Programs that fund at nominal, undiscounted levels see no change in cash needs. Every program that reports, collateralizes, or prices off a present value does. The mechanics of how captives discount reserves determine how much of the move reaches the balance sheet.
The mechanism: the dot plot sets the discount rate
Discounting divides each future claim payment by the discount rate compounded over the years until it is paid. Raise the rate and the present value falls; lower it and the present value rises. The June median of 3.8% sits above the current 3.625% midpoint and is consistent with one further 25 basis point hike before December, so the committee’s own central projection was for tighter policy, not looser. If the September SEP confirms that median, yields at the maturities that match long-tail payout patterns reprice upward, and duration-matched discount rates rise with them. If the median falls back, the repricing runs the other way. Either outcome moves booked reserves without a single change in claims experience.
The arithmetic, before the statement drops
The swing is computable now. Two illustrative payout patterns, discounted at an illustrative 4.0% base rate; substitute your stated rate and the method holds.
Pattern one is a front-loaded, 10-year workers’ compensation tail: 20% of the unpaid balance pays in year one, then 16%, 13%, 11%, 9%, 8%, 7%, 6%, 5%, and 5% through year 10. Discounted at 4.0%, $100 of nominal unpaid claims carries a present value of $85.30. Re-run the identical payments at 4.5% and the present value falls to $83.72. A 50 basis point move in the discount rate changes the booked reserve by 1.8% on this pattern.
Pattern two is a level payout over 15 years, the shape closer to an excess casualty or long-tail medical professional liability book: one-fifteenth of the balance each year. At 4.0%, the present value of $100 is $74.12; at 4.5% it is $71.44. The same 50 basis points moves the reserve 3.6%, roughly double the first pattern’s sensitivity.
The driver is duration, the payout-weighted average time to payment, not the calendar length of the tail. The front-loaded pattern carries an effective duration near 3.8 years; the level pattern near 7.0. The working rule: the percentage change in present value roughly equals duration times the change in rate. That is why two books with similar settlement horizons can swing by materially different amounts, and why the sensitivity run belongs on your payout pattern, not a generic one.
Direction matters when you brief it. A confirmed hike trims the discounted reserve, which flatters the balance sheet today and sets up a reversal if the committee later cuts. A dovish SEP that unwinds the June median raises present values and can force a discount-driven strengthening with zero deterioration in claims. Compare the nominal and discounted columns in your actuarial report to isolate how much of any reserve movement is assumption-driven versus experience-driven; a discount move lands entirely in the gap between those columns.
Where this shows up in your reserves
Three places. First, the assumptions memo of your actuarial report: the stated discount rate, and whether it references a spot yield or a full curve. Second, the discounted-versus-undiscounted reconciliation, where a rate move lands in full. Third, for captives, the funding discount inside the confidence-level study and the tail rows of Schedule P, the statutory exhibit for loss development, where a stale rate compounds longest. Public entities should check the discount rate disclosure in the GASB footnote. If your program discounts off a spot rate pegged earlier this year, that peg is what the September SEP puts at risk.
What this means for your next review
Put the September 16 SEP on the agenda for the meeting where you scope the year-end study, and fold it into your interim monitoring calendar rather than waiting for the annual report. Ask the actuary to run the discount rate two ways: at the path the September projections imply, and flat at the current selection. The gap between those runs, applied to your actual payout pattern, is the number to carry into the audit committee conversation. If the June median survives into September, a discount rate selected under easing assumptions overstates present values; the reserve looks conservative until the rate is corrected, and the correction then reads as a release.
Decision-maker checklist
- Pull the stated discount rate from your most recent actuarial report and confirm whether it assumed rate cuts that have not arrived.
- After September 16, re-benchmark the rate to yields at the duration that matches your actual payout pattern, not a generic maturity.
- Request sensitivity runs at 25 and 50 basis points against your own triangle-derived payout pattern, and record the duration your book actually carries.
- Keep the statutory, GAAP, and internal funding discount bases separate; one SEP move does not apply uniformly across the three.
- For captives, ask whether the funding discount in the confidence-level study uses a spot rate or a full yield curve, and how each behaves under a hike.
An independent reserve review brings a second pair of eyes that is free of the third-party administrator’s or fronting carrier’s incentive structure. We’re working on a directory of independent reviewing actuaries. If you’d like to be considered, get in touch.
Sources
- Federal Reserve: FOMC meeting calendars and information (September 15-16, 2026 meeting, with Summary of Economic Projections)
- FOMC statement, July 29, 2026 (federal funds target range of 3.50% to 3.75%)
- FOMC minutes, June 16-17, 2026 meeting (released July 8, 2026; median year-end 2026 projection of 3.8%)
- Summary of Economic Projections, June 17, 2026
- FOMC statement, April 29, 2026 (four dissents; inflation characterized as elevated)