thesisPrototype

Government data

US business and financial operations employment, September 2026

What will the U.S. Bureau of Labor Statistics first report in CPS Employment Situation Table A-19 for employed people in Business and financial operations occupations, total age 16 years and over, not seasonally adjusted, for September 2026, in millions?

Forecast

10.2M
9.8M–10.5M
80% prediction interval
gpt-5.5Run record ↗
9.69.91011January 2026August 2026Oct 202610.2M
historyforecast path80% interval
Probability distribution

Cumulative probability

Chance that the outcome is at or below each value.

0%25%50%75%100%9.4M10.2M11MForecast value

Shaded band: 80% interval (9.8M–10.5M). Dashed line: point estimate (10.2M).

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

Analysis

§BLS CPS A-19 occupation employment forecast

The target is the BLS CPS Employment Situation Table A-19 row 'Business and financial operations occupations,' total age 16 years and over, not seasonally adjusted, for September 2026, first print, transformed from thousands to millions. The resolver is the A-19 NSA detailed occupation table, not the broader A-13 occupation category.

▸ Reported tool use: web.openmodel report
Open BLS Employment Situation schedule at https://www.bls.gov/schedule/news_release/empsit.htm
↳ BLS schedule lists August 2026 release Sep. 04, 2026 at 08:30 AM; September 2026 release Oct. 02, 2026 at 08:30 AM; October 2026 release Nov. 06, 2026 at 08:30 AM. The registered target's 2026-10-15 resolutionDate conflicts with the official Sep. 2026 release date, so I preserve the target contract and state the discrepancy.
▸ Reported tool use: web.openmodel report
Open current BLS CPS Employment Situation Table A-19 at https://www.bls.gov/web/empsit/cpseea19.htm
↳ Current BLS Table A-19, Last Modified Date September 04, 2026: Business and financial operations occupations row has Aug. 2025 total 10,114 thousand and Aug. 2026 total 10,167 thousand, equal to 10.114 and 10.167 million after multiplying by 0.001.
▸ Reported tool use: local.repo.inspectmodel report
Inspect public Thesis generated forecast artifacts that recorded same-row BLS A-19 snapshot values for prior 2026 first prints
↳ Public generated artifact recorded same-row BLS Table A-19 snapshot values: January 2026 10,291 thousand, March 2026 10,195 thousand, May 2026 10,033 thousand, June 2026 9,720 thousand, and July 2026 9,835 thousand, equal to 10.291, 10.195, 10.033, 9.720, and 9.835 million.
▸ Reported tool use: web.searchmodel report
Fetch current Employment Situation release context for August 2026 from BLS archive
↳ BLS August 2026 Employment Situation reports total nonfarm payroll employment +162,000, unemployment rate 4.1 percent, financial activities employment -11,000, and professional and business services employment +10,000 in August 2026.
▸ Reported tool use: web.searchmodel report
Fetch BLS annual CPS occupation context table cpsaat12
↳ BLS annual average table cpsaat12 reports Business and financial operations occupations at 9,945 thousand in 2024 and 10,173 thousand in 2025, equal to 9.945 and 10.173 million.

Base rate / reference class: the recent same-row A-19 reference class available here is 10.291, 10.195, 10.033, 9.720, 9.835, and 10.167 million. Last-print persistence from August 2026 is 10.167 million, the annual 2025 average is 10.173 million, and the recent range is 9.720 to 10.291 million. I treat the August rebound as real but noisy and do not extrapolate it further.

Level, momentum, one-off, and policy mechanisms: level starts at August's 10.167 million. Momentum from June to August is positive after the June trough, but the first half of 2026 had a declining sequence. One-off effects are dominated by CPS occupation-cell sampling noise. Policy mechanism is neutral because no discrete policy change was identified that would shift this occupation count before the September survey.

Prior/update/interval: prior = persistence.last_print from August 2026 = 10.167 million; historical sample = same-row A-19 levels 2026-01, 2026-03, 2026-05, 2026-06, 2026-07, 2026-08; February and April are absent because the draft used the same-row first-print snapshots available from public artifacts and the current official table rather than interpolating unfetched months; successive fetched-print changes = -0.096, -0.162, -0.313, +0.115, +0.332 million, so sigma = 0.252 million using sample standard deviation. Update components = -0.007 million, combining +0.010 million level support from the 2025 annual average and August rebound, -0.015 million from weak financial activities payroll context, -0.002 million residual rounding/noise discipline. Point = 10.167 - 0.007 = 10.160 million. Interval method = realized same-row change dispersion; 80% half-width = 1.28*sigma = 1.28*0.252 = 0.322 million, so bounds are 10.160 - 0.322 = 9.838 and 10.160 + 0.322 = 10.482 million.

Counter-consideration: downside risk outside the interval would come from a renewed occupation-cell drop like May-to-June plus broader household employment weakness, which would land below 9.838 million. Upside risk outside the interval would come from another August-sized rebound together with stronger white-collar classification mix, which would land above 10.482 million.

Review disposition: accepted the reviewer suggestion to explain the missing February and April observations as unfetched rather than intentionally excluded, and retained the explicit note that the registered 2026-10-15 resolutionDate conflicts with the official BLS 2026-10-02 release date and should be separately reviewed; no critique items were rejected.

Key drivers

  • latest same-row A-19 value rose to 10.167 million in August 2026
  • same-row first-print history remains noisy around 10 million
  • August payroll report showed unemployment unchanged at 4.1 percent
  • financial activities payrolls fell while professional and business services rose slightly
  • registered resolutionDate conflicts with official Oct. 02 release date and is preserved

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
U.S. Bureau of Labor Statistics Employment Situation, Table A-19
Resolution date
October 15, 2026· outcome not recorded
Resolution rule
Resolve to the first value published by BLS for September 2026 in CPS Employment Situation Table A-19, row 'Business and financial operations occupations,' Total age 16 years and over, not seasonally adjusted, reported in thousands and multiplied by 0.001 to millions. Use the first official print without later revisions. The official BLS Employment Situation schedule fetched this run lists September 2026 for Oct. 02, 2026 at 08:30 AM, while the registered Thesis target contract sets resolutionDate to 2026-10-15 and expectedReleaseWindow 2026-10-07 to 2026-10-15; this cell preserves the registered target date and records the discrepancy.
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 · fast · v2.5.11

pre-submit review · completed

Draft is publishable as a target-tied forecast, with the main caveat that it records a concrete ledger/calendar discrepancy for the resolution date.

  • info optional_suggestion: Consider adding a short note explaining why February and April 2026 same-row observations are absent from the historical sample if they were unavailable rather than intentionally excluded.
  • info optional_suggestion: Before publication, separately review the registered target window/date because the draft reports an official BLS September 2026 release date of 2026-10-02 while the ledger contract uses 2026-10-15.

disposition not applicable: Review disposition: accepted the reviewer suggestion to explain the missing February and April observations as unfetched rather than intentionally excluded, and retained the explicit note that the registered 2026-10-15 resolutionDate conflicts with the official BLS 2026-10-02 release date and should be separately reviewed; no critique items were rejected.

disposition not applicable: Review disposition: accepted the reviewer suggestion to explain the missing February and April observations as unfetched rather than intentionally excluded, and retained the explicit note that the registered 2026-10-15 resolutionDate conflicts with the official BLS 2026-10-02 release date and should be separately reviewed; no critique items were rejected.

Activity artifacts

Complete original trace

§BLS CPS A-19 occupation employment forecast

The target is the BLS CPS Employment Situation Table A-19 row 'Business and financial operations occupations,' total age 16 years and over, not seasonally adjusted, for September 2026, first print, transformed from thousands to millions. The resolver is the A-19 NSA detailed occupation table, not the broader A-13 occupation category.

▸ Reported tool use: web.openmodel report
Open BLS Employment Situation schedule at https://www.bls.gov/schedule/news_release/empsit.htm
↳ BLS schedule lists August 2026 release Sep. 04, 2026 at 08:30 AM; September 2026 release Oct. 02, 2026 at 08:30 AM; October 2026 release Nov. 06, 2026 at 08:30 AM. The registered target's 2026-10-15 resolutionDate conflicts with the official Sep. 2026 release date, so I preserve the target contract and state the discrepancy.
▸ Reported tool use: web.openmodel report
Open current BLS CPS Employment Situation Table A-19 at https://www.bls.gov/web/empsit/cpseea19.htm
↳ Current BLS Table A-19, Last Modified Date September 04, 2026: Business and financial operations occupations row has Aug. 2025 total 10,114 thousand and Aug. 2026 total 10,167 thousand, equal to 10.114 and 10.167 million after multiplying by 0.001.
▸ Reported tool use: local.repo.inspectmodel report
Inspect public Thesis generated forecast artifacts that recorded same-row BLS A-19 snapshot values for prior 2026 first prints
↳ Public generated artifact recorded same-row BLS Table A-19 snapshot values: January 2026 10,291 thousand, March 2026 10,195 thousand, May 2026 10,033 thousand, June 2026 9,720 thousand, and July 2026 9,835 thousand, equal to 10.291, 10.195, 10.033, 9.720, and 9.835 million.
▸ Reported tool use: web.searchmodel report
Fetch current Employment Situation release context for August 2026 from BLS archive
↳ BLS August 2026 Employment Situation reports total nonfarm payroll employment +162,000, unemployment rate 4.1 percent, financial activities employment -11,000, and professional and business services employment +10,000 in August 2026.
▸ Reported tool use: web.searchmodel report
Fetch BLS annual CPS occupation context table cpsaat12
↳ BLS annual average table cpsaat12 reports Business and financial operations occupations at 9,945 thousand in 2024 and 10,173 thousand in 2025, equal to 9.945 and 10.173 million.

Base rate / reference class: the recent same-row A-19 reference class available here is 10.291, 10.195, 10.033, 9.720, 9.835, and 10.167 million. Last-print persistence from August 2026 is 10.167 million, the annual 2025 average is 10.173 million, and the recent range is 9.720 to 10.291 million. I treat the August rebound as real but noisy and do not extrapolate it further.

Level, momentum, one-off, and policy mechanisms: level starts at August's 10.167 million. Momentum from June to August is positive after the June trough, but the first half of 2026 had a declining sequence. One-off effects are dominated by CPS occupation-cell sampling noise. Policy mechanism is neutral because no discrete policy change was identified that would shift this occupation count before the September survey.

Prior/update/interval: prior = persistence.last_print from August 2026 = 10.167 million; historical sample = same-row A-19 levels 2026-01, 2026-03, 2026-05, 2026-06, 2026-07, 2026-08; February and April are absent because the draft used the same-row first-print snapshots available from public artifacts and the current official table rather than interpolating unfetched months; successive fetched-print changes = -0.096, -0.162, -0.313, +0.115, +0.332 million, so sigma = 0.252 million using sample standard deviation. Update components = -0.007 million, combining +0.010 million level support from the 2025 annual average and August rebound, -0.015 million from weak financial activities payroll context, -0.002 million residual rounding/noise discipline. Point = 10.167 - 0.007 = 10.160 million. Interval method = realized same-row change dispersion; 80% half-width = 1.28*sigma = 1.28*0.252 = 0.322 million, so bounds are 10.160 - 0.322 = 9.838 and 10.160 + 0.322 = 10.482 million.

Counter-consideration: downside risk outside the interval would come from a renewed occupation-cell drop like May-to-June plus broader household employment weakness, which would land below 9.838 million. Upside risk outside the interval would come from another August-sized rebound together with stronger white-collar classification mix, which would land above 10.482 million.

Review disposition: accepted the reviewer suggestion to explain the missing February and April observations as unfetched rather than intentionally excluded, and retained the explicit note that the registered 2026-10-15 resolutionDate conflicts with the official BLS 2026-10-02 release date and should be separately reviewed; no critique items were rejected.

calibrated forecast · 80% CI
10.2M[9.8M · 10.5M]
Target metadata

Data point: bls.cps.employed_people_by_occupation.business_financial_operations.september_2026.first_print

More government data forecasts