T
Tellsign
Distress Index™ · Methodology
Phase 0 · Public Sources Only · Version 1.0 · May 2026

Distress Index™
Methodology

The Distress Index aggregates public regulatory, financial, and SEC filing data into a single 0–100 score per institution. This page documents exactly how that score is calculated — every weight, every threshold, every exclusion. Nothing is proprietary. If you can read a public filing, you can reproduce any score.

The Six Signal Lenses

Signals are grouped into six lenses. Each lens has a fixed weight — how many index points a single qualifying signal contributes at full age weight — and a cap on how many signals from that lens can score simultaneously. The cap prevents a flood of routine disclosures (30 board-change 8-Ks, for example) from inflating the score beyond what the underlying distress picture warrants.

LensNameWeightCapMax pts*Phase 0 data source
C
Confirmed Action
A federal regulator publicly took action — consent orders, Matters Requiring Attention (MRAs), formal agreements, civil money penalties, prohibition orders.
25 pts4100 pts
OCC enforcement database
FDIC enforcement actions
Federal Reserve enforcement actions
CFPB enforcement database
F
Financial Deterioration
Capital ratio below regulatory thresholds, net interest margin (NIM) decline, deposit outflow, efficiency ratio deterioration, or credit-quality deterioration — each crossing a material threshold from FDIC Call Report data.
20 pts360 pts
FDIC BankFind Suite — quarterly Call Report data
D
Disclosed Risk
Risk-factor language in SEC 10-K (annual) or 10-Q (quarterly) filings: material weakness in internal controls, BSA/AML compliance deficiency, cybersecurity incident disclosure, or going-concern qualification ('substantial doubt').
15 pts115 pts
SEC EDGAR full-text search (EFTS) — 10-K and 10-Q filings
L
Leadership Transition
Executive departure or new appointment for CRO (Chief Risk Officer), CFO (Chief Financial Officer), CIO (Chief Information Officer), or Chief Compliance Officer. Presence is the signal — a bank with 30 board changes doesn't score 30× a bank with one new CRO.
Partial — public BHCs via SEC EDGAR only
10 pts110 pts
SEC 8-K Item 5.02 filings
Public press releases
O
Operational Signal
Core banking system evaluation or replacement announced in an SEC 8-K filing, material cybersecurity incident disclosure (mandatory under SEC Rule 10-5(b)), or digital banking platform initiative. Each indicates a live technology procurement cycle.
Partial — SEC EDGAR filings only
8 pts18 pts
SEC EDGAR 8-K current-event filings (EFTS full-text search)
U
Unverified
Industry press, trade publication clustering, news feed output. Not citable in the same way as regulatory filings. Excluded from the scored index unless the user explicitly opts in. Never the sole basis for outreach.
Off by default
5 pts15 pts
Free-tier news RSS aggregation
* Maximum contribution per lens at full age weight (signal filed within 89 days). All six lenses active simultaneously at full weight = 193 raw points → capped at 100. With U-lens enabled: 198 raw pts.

One within-lens exception. Within the Financial (F) lens, peer-comparison signals — an institution ranking below its FDIC peer-group median on a Call Report ratio — carry a reduced 8 points rather than the full 20. Being below a peer median is true for roughly half of all banks on any given metric by mathematical design, so it is weaker evidence than an absolute deterioration signal (for example, a capital ratio crossing a regulatory threshold), which keeps the full 20-point F weight. The peer signal still counts toward the F-lens cap; only its point value is reduced.

Scoring Formula

The algorithm has three steps. All three are deterministic — run it twice on the same data and you get the same answer.

  1. Gather signals. All signals for the institution from the ingestion pipeline. Signals below 70% confidence are excluded (see Confidence Threshold below). Unverified (U-lens) signals are excluded unless the user has opted in.
  2. Apply cap and decay per lens. Signals are sorted most-recent first. The cap (C=4, F=3, D=1, L=1, O=1) limits how many signals from each lens contribute to the score. Each contributing signal is multiplied by its age-decay multiplier (see Age Decay table).
  3. Sum and cap. The weighted, decayed contributions are summed across all lenses and capped at 100. The lens_breakdown field always shows the full signal count — the cap only affects scoring, not display.
Formula
Raw Score = Σ ( Lens Weight × Age Decay )summed over each lens, up to that lens's signal cap
Final Score = min( 100, Raw Score )

Age Decay

A consent order from two years ago matters less than one from last month. Age decay discounts older signals without removing them. A signal is never worthless — it just contributes less as time passes.

Signal ageDecay multiplierEffect
0 – 89 days1.00×Full weight
90 – 364 days0.75×
365 – 729 days0.50×
730+ days0.25×Signal still counts; heavily discounted
Date unknown0.50×Conservative fallback

Score Bands

The 0–100 score is bucketed into five bands. The bands are thresholds, not percentile rankings — an institution at 81 is Critical because of what its signals say, not because it ranks in the top 5% of the universe.

80 – 100
Critical
Multiple material signals across high-weight lenses. Immediate outreach warranted — the institution is likely in active procurement mode.
60 – 79
High
Significant signal activity across several lenses. Strong buying-trigger candidate; pipeline this account this week.
40 – 59
Elevated
Active signal presence. Worth adding to pipeline and scheduling a first touch within 30 days.
25 – 39
Watch
Early or aging signals. Monitor for escalation — next regulatory filing or executive change will clarify the picture.
0 – 24
Quiet
No active signals, or all signals aged below scoring threshold. Not a priority this quarter.

Worked Example

First National Bank of Example — fictional institution. Signal set as of May 2025. All eight signals have been collected; the algorithm processes them as follows.

Note: F-lens cap is 3, so all three F-lens signals score. C-lens cap is 4, so both C-lens signals score. All signals filed within 90–364 days receive 0.75 decay; signals under 90 days receive 1.00.

SignalLensFiledAgeDecayWeightContribution
Formal agreement — OCCCOct 2024210d0.75×2518.75
Consent order — CFPBCDec 2024150d0.75×2518.75
Net interest margin below 2.5%FNov 2024182d0.75×2015.00
Tier 1 capital ratio at 8.1%FJan 202590d1.00×2020.00
Domestic deposit decline >10%FJan 202590d1.00×2020.00
Material weakness — BSA/AML controlsDDec 2024150d0.75×1511.25
New CRO appointedLFeb 202560d1.00×1010.00
Core banking initiative — 8-KOJan 202590d1.00×88.00
Raw score (sum of contributions)121.75
Final Score = min(100, 121.75)100Critical

Example only — fictional institution and signal dates. Not derived from any real institution.

Confidence Threshold

Every signal carries a confidence score (0–100%) representing the estimated probability that the signal corresponds to a genuine buying event within 180 days. Signals below 70% confidence are excluded from scoring.

Phase 0 confidence defaults by lens: C=95%, F=90%, D=75%, L=80%, O=80%, U=40%. U-lens signals fail the 70% threshold by default, reinforcing the opt-in requirement. Signals with no confidence score (NULL) are included — not yet scored ≠ invalid.

Unverified Lens (U) — Opt-In Rules

  • U-lens signals are excluded from the Distress Index by default. They do not affect any institution's score unless the user explicitly enables them.
  • U-lens signals can appear in the signal timeline (labelled “Unverified”) as color commentary, but are never the sole basis for a briefing claim.
  • When opted in, U-lens carries 5 pts, capped at 1 signal. Even full opt-in cannot raise an institution's Quiet score above Watch on U-lens data alone.
  • Every U-lens signal carries the same citation requirements (source_url, source_excerpt, as_of_timestamp) as verified lenses. The “Unverified” label refers to the data class, not to a missing citation.

Data Sources (Phase 0)

Every Phase 0 source is publicly available and free to ingest. No paid feeds, no licensed workforce data, no premium APIs. Every signal links directly to the originating document.

CConfirmed — Regulatory Enforcement
  • OCC Enforcement Actions Search (apps.occ.gov/EASearch)
  • Federal Reserve enforcement actions (federalreserve.gov)
  • CFPB enforcement actions database (consumerfinance.gov)
  • FDIC enforcement actions (orders.fdic.gov) — backup: Federal Register
FFinancial — Call Report Data
  • FDIC BankFind Suite — quarterly Call Report data (api.fdic.gov/banks)
  • Fields: Tier 1 capital, total capital, net interest margin, efficiency ratio, domestic deposits, nonperforming loans
DDisclosed — SEC EDGAR Filings
  • SEC EDGAR Full-Text Search (EFTS) — efts.sec.gov
  • Forms: 10-K (annual), 10-Q (quarterly)
  • Terms: substantial doubt, material weakness, BSA/AML, cybersecurity incident, significant deficiency
LLeadership — Executive Transitions
  • SEC EDGAR 8-K filings — Item 5.02 (departure/appointment of directors/officers)
  • Public press releases (cross-referenced via EDGAR)
OOperational — Current-Event Filings
  • SEC EDGAR 8-K full-text search (EFTS)
  • Terms: 'core banking', 'digital banking', 'Item 1.05' (cybersecurity incident)
  • 8-Ks are mandatory current-event filings — higher signal quality than job postings
UUnverified — News & Industry Press
  • Free-tier news API or RSS aggregation
  • Off by default — excluded from scored index unless opted in
  • Never the sole basis for outreach

Phase 0 Limitations

Phase 0 is built entirely on free public data sources. This is a deliberate architectural choice — it establishes the citation standard and the product experience before any paid feeds are introduced. It also means there are real coverage gaps, documented here.

  • L-lens is partial. Leadership transitions for publicly traded bank holding companies are captured via SEC 8-K filings. Private community banks (~90% of the universe) do not file 8-Ks for executive changes. Full leadership coverage requires licensed workforce data (Revelio Labs, Live Data Technologies), available at the Wire tier.
  • O-lens is partial. Operational signals are sourced from SEC EDGAR 8-K filings — so only public BHCs that file 8-Ks. Private-bank RFPs and job-posting data require licensed feeds, available at Wire tier.
  • FDIC enforcement actions are partially unavailable. The FDIC migrated enforcement data to a new system (orders.fdic.gov) in early 2026 with no public API. OCC, Federal Reserve, and CFPB enforcement sources are unaffected.
  • Refresh rate is weekly (Brief tier). Wire tier will receive daily and real-time alerts. Phase 0 scores are recomputed every Monday 06:00 ET.
  • No licensed financial data. The F-lens uses publicly available FDIC Call Report data, refreshed quarterly with a ~45-day lag after quarter-end. Licensed intra-quarter data (SNL, S&P Market Intelligence) is out of Phase 0 scope.
  • Credit union universe is not included. Phase 0 covers FDIC-insured banks only. NCUA-regulated credit unions (~4,287 federally insured) are a separate universe toggle available as a future add-on.

The Tellsign Distress Index™ is a signal-aggregation indicator derived from publicly available regulatory filings, public financial data, and public SEC filings. It is not a regulatory rating, a determination of financial condition, an investment recommendation, or a legal opinion regarding any financial institution.

All source excerpts are reproduced verbatim from the cited public documents. Users are responsible for independent verification before taking any action based on Tellsign output. Tellsign does not guarantee the accuracy, completeness, or timeliness of the underlying source data.

Tellsign · Distress Index™ Methodology v1.0 · May 2026Sign in to Tellsign →