thesisPrototype

Conditional

DoD prime award obligations FY2027, no FY27 NDAA

If no Act authorizing appropriations for military activities of the Department of Defense for fiscal year 2027 is enacted into law on or before 2026-12-31, what will USAspending API v2 report for Department of Defense agency 097 prime award obligations for fiscal year 2027, using the obligations field from /api/v2/agency/097/awards/?fiscal_year=2027, transformed to billions USD, in the registered query snapshot first print?

Conditional on: No Act authorizing appropriations for military activities of the Department of Defense for fiscal year 2027 (the FY2027 NDAA; the House-reported bill is H.R. 8800, 119th Congress) is enacted into law on or before 2026-12-31.

Forecast

493.5
436.9–550.1
80% prediction interval
gpt-5.5Run record ↗

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%352493.5635Forecast value

Shaded band: 80% interval (436.9–550.1). Dashed line: point estimate (493.5).

Derived from the point estimate and 80% interval; the agent did not report a full distribution.

Analysis

§Resolution

This cell resolves the registered conditional target for USAspending API v2 agency 097 award obligations, fiscal_year=2027, field obligations, transformed by 1e-9 to billions USD. The target registration commits the expected release window 2027-10-15 through 2027-10-22; I set resolutionDate to the window end, 2027-10-22, and keep the registered query URL as the resolver rather than substituting an agency profile total or search summary.

▸ Reported tool use: local.target_registrationmodel report
Inspect records/targets/2026-08-13-7d60648441d141998baeb2414745ac6c362240c575cb0c305c27ccd3c85a558b.json
↳ Registered target fields inspected this run: catalogSlug us-dod-prime-award-obligations-fy2027-no-fy27-ndaa; period 2027; unit billions USD; dataPointId usaspending.dod.prime_award_obligations.2027.registered_query_snapshot.no_fy27_ndaa; conditionDeadline 2026-12-31; sourceUrl https://api.usaspending.gov/api/v2/agency/097/awards/?fiscal_year={fiscal_year}; field obligations; transform factor 1e-9; expectedReleaseWindow start 2027-10-15 and end 2027-10-22.
▸ Reported tool use: usaspending.awardsmodel report
GET exact registered query for fiscal_year=2017 through 2021
↳ Fetched USAspending agency 097 awards obligations from the registered endpoint: FY2017 obligations 327990949917.96 USD = 327.991B, transaction_count 3689848; FY2018 obligations 365937501255.11 USD = 365.938B, transaction_count 4528752; FY2019 obligations 391333992872.89 USD = 391.334B, transaction_count 4359716; FY2020 obligations 431883672152.30 USD = 431.884B, transaction_count 4059194; FY2021 obligations 395332709399.53 USD = 395.333B, transaction_count 4362064.
▸ Reported tool use: usaspending.awardsmodel report
GET exact registered query for fiscal_year=2022 through 2027
↳ Fetched USAspending agency 097 awards obligations from the registered endpoint: FY2022 obligations 424512414112.80 USD = 424.512B, transaction_count 4377450; FY2023 obligations 467273435599.16 USD = 467.273B, transaction_count 4430019; FY2024 obligations 456945504400.91 USD = 456.946B, transaction_count 4447901; FY2025 obligations 501543513238.03 USD = 501.544B, transaction_count 4516359. FY2026 is incomplete: obligations 298980123138.95 USD = 298.980B through latest_action_date 2026-08-11 with transaction_count 2641896. FY2027 returned detail Field fiscal_year value 2027 is above max 2026, so no current outcome print exists.
▸ Reported tool use: govinfo.billstatusmodel report
Fetch https://www.govinfo.gov/bulkdata/BILLSTATUS/119/hr/BILLSTATUS-119hr8800.xml
↳ GovInfo BILLSTATUS for 119 HR 8800 was updated 2026-08-11T20:28:11Z; number 8800; type HR; congress 119; introducedDate 2026-05-13; title National Defense Authorization Act for Fiscal Year 2027; policy area Armed Forces and National Security. Selected actions fetched include Ordered to be Reported by 44-12 on 2026-06-04, Reported (Amended) by the Committee on Armed Services as H. Rept. 119-698 on 2026-06-15, and Rules Committee resolutions reported on 2026-06-30 and 2026-07-20; no fetched selected action said became Public Law or signed by the President.
▸ Reported tool use: thesis.specs_checkmodel report
curl -sS https://app.thesisinstitute.org/specs.json
↳ Slug check attempted as required: https://app.thesisinstitute.org/specs.json returned an HTTP 404 Next.js page with status text Forecast not found and content including route specs.json. The provided registered slug us-dod-prime-award-obligations-fy2027-no-fy27-ndaa is therefore retained; I did not invent a replacement slug.

Base rate / reference class: comparable complete registered-query prints are FY2017-FY2025 obligations values of 328.0B, 365.9B, 391.3B, 431.9B, 395.3B, 424.5B, 467.3B, 456.9B, and 501.5B. This reference class has mean level 418.1B, median level 424.5B, range 328.0B-501.5B, annual changes +37.9B, +25.4B, +40.5B, -36.6B, +29.2B, +42.8B, -10.3B, and +44.6B, and annual-change sigma 29.5B. FY2026 is excluded from the base rate because it is an incomplete year-to-date value.

Model candidates under thesis_model_candidate_v1, summarized in trace: two-year persistence uses the FY2025 completed registered-query print as point 501.5, p10 463.8, p50 501.5, p90 539.3, 80% interval [463.8, 539.3], 90% interval [453.2, 549.9], interval_method annual-change residual normal, calibration_n 8, train_cutoff FY2025, walk_forward_mae 30.0 over FY2021-FY2025 two-year-origin backtests. A two-year mean-change drift candidate has point 544.9, p10 507.2, p50 544.9, p90 582.6, 80% interval [507.2, 582.6], 90% interval [496.5, 593.3], interval_method annual-change residual normal, calibration_n 8, train_cutoff FY2025, walk_forward_mae 38.5. I select two-year persistence because it has the lower walk-forward MAE for this horizon.

Prior/update/interval: persistence prior = 501.5B from the completed FY2025 registered-query print; historical sample = FY2017-FY2025 complete values with successive annual changes 37.9, 25.4, 40.5, -36.6, 29.2, 42.8, -10.3, and 44.6B; adjustment components = -8.0B judgmental no-FY2027-NDAA adjustment, because authorization delay can slow program direction but USAspending prime award obligations are funded by appropriations and execution, not by the authorization act alone. Interval method = annual-change residual, sigma = 29.5B, nominal 80% half-width = 1.28*sigma = 1.28*29.5 = 37.7B; I widen by 1.5x to 56.6B for the two-fiscal-year horizon and conditional legislative uncertainty. Point = 501.5 - 8.0 = 493.5B; interval = 493.5 +/- 56.6 = [436.9, 550.1]B after rounding to 0.1B.

Sanity and falsification: a +/-56.6B interval around prior-year prints would have covered all 8 fetched annual changes from FY2018-FY2025. Downside risk outside the interval would be a long continuing resolution, delayed appropriations, DoD award timing slippage, or deobligation cleanup that would land below the interval. Upside risk outside the interval would be large late-FY procurement or RDT&E awards, supplemental defense appropriations, or unusually fast USAspending posting that would land above the interval.

Review disposition: accepted the reviewer suggestions by labeling the -8.0B no-NDAA adjustment as judgmental rather than separately empirically calibrated, and retained the explicit FY2027 endpoint failure showing that no resolving print was available at forecast time.

Key drivers

  • two-year persistence benchmark from FY2025
  • no-FY2027-NDAA condition implies authorization delay
  • appropriations, not authorization, drive award budget authority
  • large historical annual award-timing volatility
  • FY2026 current API value is incomplete year-to-date

Tool evidence

This run has no captured tool responses. “Reported tool use” in the analysis is the model’s account, not an archived tool response.

Sources and resolution

Official source
USAspending API v2, agency 097 award summary
Resolution date
October 22, 2027· outcome not recorded
Resolution rule
Resolve from USAspending API v2 agency 097 awards endpoint, fiscal_year=2027, field obligations, for Department of Defense prime award obligations in fiscal year 2027. Use the first registered-query snapshot in the lab-committed expected release window 2027-10-15 through 2027-10-22, multiply obligations by 1e-9, and round to 0.1 billions USD. Later USAspending revisions after that first snapshot are ignored. Resolve only if the condition is satisfied; if any Act authorizing appropriations for military activities of the Department of Defense for fiscal year 2027 is enacted into law on or before 2026-12-31, this conditional cell is unresolved.
Run details

The analysis is the model’s written report. Tool-use descriptions in that report are model claims; the activity artifacts contain the execution record.

thesis.analyst · gpt-5.5 · full · v2.5.9

pre-submit review · completed

Draft is publication-ready on the target contract and forecast-method rubric, with only minor tightening suggested.

  • info optional_suggestion: Consider naming the no-NDAA -8.0B adjustment as judgmental rather than empirically calibrated, since no separate historical reference class for NDAA-delay years is shown.
  • info optional_suggestion: In the final trace, keep the FY2027 endpoint fetch failure explicit so reviewers can see no resolving print was available at forecast time.

disposition not applicable: Review disposition: accepted the reviewer suggestions by labeling the -8.0B no-NDAA adjustment as judgmental rather than separately empirically calibrated, and retained the explicit FY2027 endpoint failure showing that no resolving print was available at forecast time.

disposition not applicable: Review disposition: accepted the reviewer suggestions by labeling the -8.0B no-NDAA adjustment as judgmental rather than separately empirically calibrated, and retained the explicit FY2027 endpoint failure showing that no resolving print was available at forecast time.

Activity artifacts

Complete original trace

§Resolution

This cell resolves the registered conditional target for USAspending API v2 agency 097 award obligations, fiscal_year=2027, field obligations, transformed by 1e-9 to billions USD. The target registration commits the expected release window 2027-10-15 through 2027-10-22; I set resolutionDate to the window end, 2027-10-22, and keep the registered query URL as the resolver rather than substituting an agency profile total or search summary.

▸ Reported tool use: local.target_registrationmodel report
Inspect records/targets/2026-08-13-7d60648441d141998baeb2414745ac6c362240c575cb0c305c27ccd3c85a558b.json
↳ Registered target fields inspected this run: catalogSlug us-dod-prime-award-obligations-fy2027-no-fy27-ndaa; period 2027; unit billions USD; dataPointId usaspending.dod.prime_award_obligations.2027.registered_query_snapshot.no_fy27_ndaa; conditionDeadline 2026-12-31; sourceUrl https://api.usaspending.gov/api/v2/agency/097/awards/?fiscal_year={fiscal_year}; field obligations; transform factor 1e-9; expectedReleaseWindow start 2027-10-15 and end 2027-10-22.
▸ Reported tool use: usaspending.awardsmodel report
GET exact registered query for fiscal_year=2017 through 2021
↳ Fetched USAspending agency 097 awards obligations from the registered endpoint: FY2017 obligations 327990949917.96 USD = 327.991B, transaction_count 3689848; FY2018 obligations 365937501255.11 USD = 365.938B, transaction_count 4528752; FY2019 obligations 391333992872.89 USD = 391.334B, transaction_count 4359716; FY2020 obligations 431883672152.30 USD = 431.884B, transaction_count 4059194; FY2021 obligations 395332709399.53 USD = 395.333B, transaction_count 4362064.
▸ Reported tool use: usaspending.awardsmodel report
GET exact registered query for fiscal_year=2022 through 2027
↳ Fetched USAspending agency 097 awards obligations from the registered endpoint: FY2022 obligations 424512414112.80 USD = 424.512B, transaction_count 4377450; FY2023 obligations 467273435599.16 USD = 467.273B, transaction_count 4430019; FY2024 obligations 456945504400.91 USD = 456.946B, transaction_count 4447901; FY2025 obligations 501543513238.03 USD = 501.544B, transaction_count 4516359. FY2026 is incomplete: obligations 298980123138.95 USD = 298.980B through latest_action_date 2026-08-11 with transaction_count 2641896. FY2027 returned detail Field fiscal_year value 2027 is above max 2026, so no current outcome print exists.
▸ Reported tool use: govinfo.billstatusmodel report
Fetch https://www.govinfo.gov/bulkdata/BILLSTATUS/119/hr/BILLSTATUS-119hr8800.xml
↳ GovInfo BILLSTATUS for 119 HR 8800 was updated 2026-08-11T20:28:11Z; number 8800; type HR; congress 119; introducedDate 2026-05-13; title National Defense Authorization Act for Fiscal Year 2027; policy area Armed Forces and National Security. Selected actions fetched include Ordered to be Reported by 44-12 on 2026-06-04, Reported (Amended) by the Committee on Armed Services as H. Rept. 119-698 on 2026-06-15, and Rules Committee resolutions reported on 2026-06-30 and 2026-07-20; no fetched selected action said became Public Law or signed by the President.
▸ Reported tool use: thesis.specs_checkmodel report
curl -sS https://app.thesisinstitute.org/specs.json
↳ Slug check attempted as required: https://app.thesisinstitute.org/specs.json returned an HTTP 404 Next.js page with status text Forecast not found and content including route specs.json. The provided registered slug us-dod-prime-award-obligations-fy2027-no-fy27-ndaa is therefore retained; I did not invent a replacement slug.

Base rate / reference class: comparable complete registered-query prints are FY2017-FY2025 obligations values of 328.0B, 365.9B, 391.3B, 431.9B, 395.3B, 424.5B, 467.3B, 456.9B, and 501.5B. This reference class has mean level 418.1B, median level 424.5B, range 328.0B-501.5B, annual changes +37.9B, +25.4B, +40.5B, -36.6B, +29.2B, +42.8B, -10.3B, and +44.6B, and annual-change sigma 29.5B. FY2026 is excluded from the base rate because it is an incomplete year-to-date value.

Model candidates under thesis_model_candidate_v1, summarized in trace: two-year persistence uses the FY2025 completed registered-query print as point 501.5, p10 463.8, p50 501.5, p90 539.3, 80% interval [463.8, 539.3], 90% interval [453.2, 549.9], interval_method annual-change residual normal, calibration_n 8, train_cutoff FY2025, walk_forward_mae 30.0 over FY2021-FY2025 two-year-origin backtests. A two-year mean-change drift candidate has point 544.9, p10 507.2, p50 544.9, p90 582.6, 80% interval [507.2, 582.6], 90% interval [496.5, 593.3], interval_method annual-change residual normal, calibration_n 8, train_cutoff FY2025, walk_forward_mae 38.5. I select two-year persistence because it has the lower walk-forward MAE for this horizon.

Prior/update/interval: persistence prior = 501.5B from the completed FY2025 registered-query print; historical sample = FY2017-FY2025 complete values with successive annual changes 37.9, 25.4, 40.5, -36.6, 29.2, 42.8, -10.3, and 44.6B; adjustment components = -8.0B judgmental no-FY2027-NDAA adjustment, because authorization delay can slow program direction but USAspending prime award obligations are funded by appropriations and execution, not by the authorization act alone. Interval method = annual-change residual, sigma = 29.5B, nominal 80% half-width = 1.28*sigma = 1.28*29.5 = 37.7B; I widen by 1.5x to 56.6B for the two-fiscal-year horizon and conditional legislative uncertainty. Point = 501.5 - 8.0 = 493.5B; interval = 493.5 +/- 56.6 = [436.9, 550.1]B after rounding to 0.1B.

Sanity and falsification: a +/-56.6B interval around prior-year prints would have covered all 8 fetched annual changes from FY2018-FY2025. Downside risk outside the interval would be a long continuing resolution, delayed appropriations, DoD award timing slippage, or deobligation cleanup that would land below the interval. Upside risk outside the interval would be large late-FY procurement or RDT&E awards, supplemental defense appropriations, or unusually fast USAspending posting that would land above the interval.

Review disposition: accepted the reviewer suggestions by labeling the -8.0B no-NDAA adjustment as judgmental rather than separately empirically calibrated, and retained the explicit FY2027 endpoint failure showing that no resolving print was available at forecast time.

calibrated forecast · 80% CI
493.5[436.9 · 550.1]
Target metadata

Data point: usaspending.dod.prime_award_obligations.2027.registered_query_snapshot.no_fy27_ndaa

More conditional forecasts