US continued claims, week ending Aug 1 2026

Framing and exact resolver: this targets the DOL ETA Unemployment Insurance Weekly Claims first print for seasonally adjusted insured unemployment, also called continued claims, for the week ending August 1, 2026. The DOL release table labels this variant as Insured Unemployment (SA), so all anchors below use that same seasonally adjusted variant.

Ledger resolver note: the official statistical source is DOL ETA, but the canonical ledger sourceBinding for resolution is the ALFRED CCSA advance-vintage CSV at alfred.stlouisfed.org with a 1e-06 transform to millions. I keep the forecast tied to that ledger target and use DOL only to verify the first-print release date and source series substance.

Base rate/reference class: the outside-view prior is persistence of the latest official SA insured-unemployment level, with uncertainty calibrated to 2026 weekly changes in the same DOL SA continued-claims series rather than to initial claims or unadjusted state totals.

Prior/update/interval: persistence prior = 1.796 million from the latest DOL first print; historical sample = 27 successive weekly SA insured-unemployment changes in 2026 from January 3 through July 11, giving sigma = 0.023 million. Update components: level is 0.009 million below the latest four-week average of 1.805 million, two-week momentum is -0.025 million from June 27 to July 11, and the July 18 initial-claims fall to 187,000 versus 209,000 in the same July 23 DOL release points to lower near-term inflows; combined adjustment = -0.016 million, so point = 1.796 - 0.016 = 1.780 million. One-week 80% half-width = 1.28*sigma = 1.28*0.023 = 0.029 million; because the target is three weekly steps after the latest continued-claims observation, scale by sqrt(3), giving half-width = 0.051 million and bounds 1.729 to 1.831 million.

Counter-considerations: upside risk is that very low initial claims reverse or benefit durations lengthen, which would land above the interval if continued claims rise by more than about 35,000 from 1.796 million to print above 1.831 million. Downside risk is that July seasonal adjustment and lower layoffs pull insured unemployment below 1.729 million; outside the interval would require a roughly 67,000 or larger decline from the latest level.

Review disposition: accepted the resolver critique by aligning resolutionSourceUrl and rule to the ledger-bound ALFRED CCSA advance-vintage source while explicitly noting DOL ETA as the official first-print source; accepted the release-window critique by tying the August 13 date to the ledger's 2026-08-11 through 2026-08-15 expected window; kept the prior/update/interval arithmetic unchanged.

Prior point = latest observed value = 1.8; 80% interval = [1.8, 1.8].

This run stops before target-specific agent updates; the primary forecast records the adjustment away from this prior.