§Forecast DoD FY2026 Prime Award Transactions
Framing and exact resolver: this is the USAspending API v2 agency 097 awards endpoint, fiscal_year=2026, field transaction_count, transformed to millions. The registered target uses a resolve-by-bound window ending 2026-10-22; the official submission-period endpoint visible this run lists recent reveal and certification dates but not yet the future FY2026 fiscal-month-12 row, so I keep the ledger resolutionDate and note that the exact future reveal row was not observable yet.
Fetched USAspending agency 097 awards endpoint for completed fiscal years 2021 and 2022, reading transaction_count and converting by 1e-6.↳ Fetched FY2021 transaction_count 4566000 = 4.566 million; FY2022 transaction_count 4318000 = 4.318 million.Fetched USAspending agency 097 awards endpoint for completed fiscal years 2023 and 2024, same agency 097 and same transaction_count field.↳ Fetched FY2023 transaction_count 4084000 = 4.084 million; FY2024 transaction_count 3965000 = 3.965 million.Fetched USAspending agency 097 awards endpoint for FY2025 and the public submission-period schedule context.↳ Fetched FY2025 transaction_count 3786000 = 3.786 million; submission_periods showed FY2026 fiscal_month 9 and FY2025 fiscal_month 12 entries, with FY2025 fiscal_month 12 submission_reveal_date 2025-12-06T02:57:27Z.Base rate/reference class: the same-source annual flow reference class is FY2021-FY2025 DoD prime award transaction counts, current official API readings rather than known first registered snapshots, ranging from 3.786 million to 4.566 million, with a mean of 4.144 million and a downward drift of about 0.195 million per year over the last four year-to-year steps.
Level and momentum: a linear trend/extrapolation prior was considered but only partially used because the decline from FY2021 through FY2025 appears to be flattening; I therefore anchor on FY2025 persistence rather than extending the full trend mechanically.
Policy and mechanism adjustment: FY2026 defense spending remains large, so the main policy effect is on composition and timing rather than a collapse in transaction activity. I add a small +0.07 million adjustment versus pure recent-trend persistence for late-year modifications and awards tied to ongoing defense procurement activity.
Prior/update/interval: persistence prior is FY2025 at 3.786 million; historical sample is FY2021-FY2025 same USAspending API annual values of 4.566, 4.318, 4.084, 3.965, and 3.786 million; adjustment components are +0.04 million for stabilization after the FY2021-FY2025 decline and +0.03 million for FY2026 late-year defense award/modification flow, giving point 3.786 + 0.074 = 3.860 million. For this annual flow-series interval I use dispersion of the annual values themselves: sigma = 0.305 million, so 80% half-width is about 1.28*sigma = 1.28*0.305 = 0.390 million, implying 3.860 +/- 0.390 = 3.470 to 4.250 million. The level-based sigma is intentionally conservative relative to recent year-to-year change volatility because the first registered snapshot can be affected by reporting lag and late procurement transaction clearing.
Counter-considerations: upside risk is a heavier-than-usual late FY2026 modification surge or delayed FY2025-FY2026 procurement data clearing into the first snapshot, which would land above the interval if the count exceeds 4.25 million. Downside risk is a reporting lag, shutdown-related processing disruption, or sustained simplification of award actions, which would land below the interval if the count is under 3.47 million.
Review disposition: accepted the interval critique by explicitly justifying the level-dispersion interval as a conservative annual-flow and first-snapshot uncertainty proxy, accepted the prior critique by stating that linear trend extrapolation was considered and discounted, and clarified that historical values are current API readings rather than known first registered snapshots.