The FHA Signal Thesis, Reframed
The borrower split still matters: stress tends to appear first where household cash buffers are thinnest.
This thesis page was revised after the index-family cutover because the earlier version joined changing ADI primitives to a retired engine. The borrower split remains the question: stress tends to show up first where buffers are thinnest. The public ADI framing is narrower.
The family ADI retired the old component weights. It averages five equal-weighted domains after ranking each input inside its own quarterly history. FHA delinquency is supporting mortgage context outside the national composite weight set.
The ADI reference page publishes the maintained national reading with its required band gloss. The score is a mean of input percentiles, and quarter-rank context comes from the separate rank-in-history field.
For source data, use The FHA Signal and the ADI methodology.
Refresh Trace
2026-08-07| Recently changed indicator | Source | Period | Snapshot change |
|---|---|---|---|
| Total Nonfarm Payrolls | U.S. Bureau of Labor Statistics via FRED | 2026-07 | -126K |
| CFPB Consumer Complaint Volume | Consumer Financial Protection Bureau | 2026-07 | +23054 |
| AI Job Posting Share (% of Total) | Indeed Hiring Lab AI Tracker (public CSV) | 2026-07 | +0.4 pp |
| The Tech Drought | Bureau of Labor Statistics, JOLTS | 2026-06 | +14K |
| The Tightening | Chicago Fed via FRED | 2026-07-31 | +0.003 |
Loading comments…