Methodology v1.3: the restatement note
The Deposit Mortality Index publishes under a standing rule: the index never moves quietly. Every methodology change is classified, disclosed, and quantified before restated figures appear anywhere. This page is that disclosure for version 1.3.
The formal restatement
With the October 2026 scoring, the Deposit Mortality Index adopts corrected Step 3 balance weights drawn from published 2022 Survey of Consumer Finances figures, replacing an approximate vector used since launch. Under the Methodology Change Policy this is a Class 2 input-fidelity change: component definitions, weights, and tier thresholds are unchanged. Every scored institution was computed under both specifications. Effect: composite scores decline by a median of 2.6 points (5th to 95th percentile: −3.0 to −2.1; largest single change 3.6), with no institution scoring higher; 920 of 8,429 institutions move down one tier (664 Elevated to Moderate, 237 High to Elevated, 19 Severe to High); the national ten-year expected transfer exposure declines from $1.52 trillion to approximately $1.46 trillion. The correction lowers the index's published estimates. Institutions that purchased reports on the outgoing specification within the preceding two quarters receive restated figures at no charge. Report vintages are labeled; figures published before October 2026 reflect the superseded specification and are not retroactively edited.
What changed, in plain language
Step 3 of the methodology converts a footprint's population age mix into a deposit-dollar age mix, because older households hold larger balances. The weights for that step were launched as a rounded approximation of the Survey of Consumer Finances shape, carrying an internal instruction to verify against the published tables before publication. The verification, performed in September 2026, found the approximation overweighted the 55-to-64 band: actual SCF balances are nearly flat from 45 through 64 and rise sharply at 65. Version 1.3 replaces the approximation with the verified published values.
The correction moves every estimate in one direction: down. No institution scores higher under v1.3 than it did under v1.2, and the national exposure figure falls about four percent. An index whose first restatement lowers its own headline is the index working as designed.
What it means for published figures
- The October 2026 edition of The State of Deposit Mortality is computed entirely under v1.3 and explains the restatement in its opening section, including the decomposition of every headline change.
- The July 2026 edition remains archived at its permanent URL exactly as published. It reflects the superseded specification and is vintage-labeled, not retroactively edited.
- Institution reports purchased on the outgoing specification within the preceding two quarters receive restated figures at no charge: [email protected].
- Quarterly refreshes are not methodology changes. Data inputs refresh on their natural cycles against the frozen scale; this note covers the one specification change in the v1.3 release.
Read the documents:
The State of Deposit Mortality 2026: October Edition
DMI Methodology v1.3, complete specification · Methodology Overview, 3 pages
The State of Deposit Mortality 2026: July Edition (archived)
The Deposit Mortality Index is a research estimate derived entirely from public data using the disclosed DMI methodology. DMI scores are statements of opinion based on that methodology and those data sources; they are not credit ratings, examination ratings, or assessments of safety and soundness. Deposit Mortality Index, DMI, and DMI Verified are trademarks of Wunderwood Corporation d/b/a Prismm. © 2026 Wunderwood Corporation d/b/a Prismm. All rights reserved. The Deposit Mortality Index was created by Martha Sylla Underwood. Questions: [email protected].