01
CMS Provider Data Catalog
The authoritative federal database of Medicare/Medicaid-certified providers. Updated monthly. Paginated API with 1,500-row caps — full ingestion requires sequential pagination.
Overall star ratings (1–5 scale)
Health inspection ratings and deficiency history
Staffing ratings and hours per resident day
Quality measure ratings (long-stay and short-stay)
Special Focus Facility designation status
Ownership type and chain affiliation
Abuse icon and penalty history (count and dollar)
CMS.gov · Federal
02
Payroll-Based Journal (PBJ)
CMS-mandated daily staffing submission from every certified SNF. The most granular federal staffing dataset available — captures actual worked hours by staff type, not self-reported estimates.
RN hours per resident day (actual, not reported)
LPN and CNA hours per resident day
Weekend staffing coverage
Zero-RN-day frequency per quarter
Contract vs. employee staff composition
11-quarter history (Q1 2023 – Q1 2026)
CMS PBJ · Federal
03
HCRIS Cost Reports (CMS-1728-20)
Annual financial cost reports submitted by every Medicare-participating SNF, home health agency, and hospice provider. The only federal source for facility-level financial data including revenue, costs, and payer mix.
Total revenue and operating costs
Operating margin (derived)
Medicare, Medicaid, and private pay day mix
Cost per visit (home health)
Case mix index
HCRIS · Federal
04
HHVBP Performance Data
CMS Home Health Value-Based Purchasing program scores for all participating agencies. The federal accountability system for home health quality — 41% of agencies fall outside its measurement threshold.
Total performance score (0–100)
Payment adjustment percentage
VBP participation status
Benchmark comparisons (national and state)
CMS HHVBP · Federal
05
FMCSA Safety Database
Federal Motor Carrier Safety Administration records for all registered commercial carriers. Refreshed daily. The authoritative source for carrier operating authority, safety ratings, and crash history.
Operating authority status (active / inactive)
Safety rating (satisfactory / conditional / unsatisfactory)
Crash records (24-month and all-time)
Vehicle and driver out-of-service rates
Insurance filing history and cancellations
SMS BASICs scores (7 categories)
FMCSA · Federal
06
Census ACS County Data
American Community Survey 5-year estimates at the county level. Used to contextualize provider markets — population age distribution, density, and demographic factors that affect demand and access.
Population 65+ by county
Population density classification
Rural / urban classification
Care desert scoring inputs
U.S. Census Bureau · Federal
07
FDIC Call Reports
Quarterly financial reports filed by all FDIC-insured institutions. The primary spine of the Bank Health Score. 1,131,904 quarterly records ingested from 1992 to present.
Capital adequacy — Tier 1 ratio, risk-based capital ratio
Asset quality — noncurrent loan ratio, net charge-off rate
Earnings — ROA, net interest margin, efficiency ratio
Liquidity — loan-to-deposit ratio, uninsured deposit %
Real estate concentration — total domestic real-estate loans vs. equity
Federal · Quarterly
08
FDIC Enforcement Actions
All formal enforcement actions issued by the FDIC against supervised institutions — consent orders, cease and desist orders, civil money penalties, and removal/prohibition orders.
Action type — C&D, consent order, civil money penalty
Effective date and termination date
Active vs. terminated status
10,841 total actions · 8,113 active
Federal · Monthly update
09
FFIEC CRA Ratings
Community Reinvestment Act examination ratings published by the FFIEC. CRA ratings below Satisfactory are a regulatory constraint on M&A activity and branch expansion.
Overall rating — Outstanding, Satisfactory, Needs to Improve, Substantial Noncompliance
Examination date
83,093 ratings ingested · historical record from 1990
Federal · Annual
10
FDIC Summary of Deposits
Annual branch-level deposit data for every FDIC-insured institution. Used to assess franchise footprint, deposit market share, and geographic concentration.
Branch-level deposit balances
Geographic footprint — states and MSAs
76,120 branches · $18T total deposits · 2025 vintage
Federal · Annual
11
OSHA Enforcement Data
Federal OSHA inspection and violation records dating to 1970. Includes citation type (willful, repeat, serious), penalty amounts, abatement dates, fatality flags, and severe injury reports (amputations, hospitalizations, eye loss) since January 2015.
5.16M inspections · 1970–2026 · all federal jurisdictions
12.77M violation records · $6.47B in current penalties
103,750 severe injury reports · 27,358 amputations
Federal · Quarterly
12
DOL Wage & Hour Division
Department of Labor WHD investigation records covering wage theft, FLSA violations, child labor violations, and civil money penalties. Every investigation with findings is documented with back wages recovered and violation counts.
365,393 investigations with findings
$4.73B in documented back wages recovered
FLSA violation counts · repeat violator flags · CMP amounts
Federal · Quarterly
13
FFIEC Call Report — Schedule RC-B
FFIEC bulk Call Report data for securities holdings. Contains held-to-maturity (HTM) and available-for-sale (AFS) unrealized gains and losses — the primary input for SVB-pattern failure mode detection in Bank Health Score v2.0. Q1 2026 vintage. 4,262 records ingested.
HTM and AFS amortized cost and fair value
HTM and AFS unrealized gains / losses by security category
Total unrealized loss as % of Tier 1 equity capital (derived)
Bank of America 33.8%, Wells Fargo 20.8% — validates against public disclosures
FFIEC · Quarterly
14
FFIEC Call Report — Schedule RC-O
FFIEC bulk Call Report schedule for deposit insurance assessment data. Contains RCON5597 — the bank's self-reported estimate of uninsured deposits, the authoritative figure used by the FDIC for insurance assessment. Preferred over FDIC API estimate for large institutions. Mandatory reporters: banks with assets above $1B. 1,006 institutions with non-null values ingested.
RCON5597 — estimated uninsured deposits (self-reported by institution)
Coverage: institutions with assets > $1B (mandatory reporters)
Preferred input for B-03 cliff penalty calculation when available
FFIEC · Quarterly
15
FFIEC Call Report — Schedule RC-N
FFIEC bulk Call Report data for past due and nonaccrual loans. The primary forward-looking credit quality signal in Bank Health Score v2.1. Institution-level delinquency data by category — not available from the FDIC BankFind API. Q1 2026 vintage. 4,262 records ingested with 99.8% nonaccrual coverage.
RCFD/RCON1403 — Total loans past due 30–89 days, accruing
RCFD/RCON1406 — Total nonaccrual loans
RCFD/RCON1407 — Total loans past due 90 or more days, accruing
Noncurrent total — nonaccrual + past due 90+ (primary Asset Quality input)
RCFD prefix takes precedence; RCON used as fallback for smaller institutions
FFIEC CDR · Federal · Quarterly
16
CFPB Home Mortgage Disclosure Act (HMDA)
Annual mortgage application and origination data filed by all HMDA-reporting lenders with the Consumer Financial Protection Bureau. 4,351 institutions ingested covering 5,760,991 originations and 627,874 denials in 2024. Institution-level denial rates, origination volumes, and average loan amounts joined to bank scores via RSSD identifier. Reveals mortgage underwriting posture and market participation not captured by FDIC Call Report financials.
Total applications, originations, denials, withdrawals per institution
Denial rate = denied / (originated + denied) — standard fair lending denominator
Average loan amount per institution — proxy for market segment positioning
2,246 FDIC banks matched via RSSD — 100% coverage of active institutions with HMDA filings
2,105 non-bank lenders (Rocket, UWM, CrossCountry, Guild, et al.) included for market context
hmda_denial_flag: TRUE for institutions above P75 denial rate (16.0%)
CFPB · Federal · Annual
22
USASpending.gov Award Data Archive
Annual contract award archives published by the U.S. Treasury. Full transaction-level obligation data for every federal contract, FY2021–FY2025. Five annual archive files ingested — approximately 1.8GB compressed each, 33M rows total. The authoritative federal source for vendor obligation history, agency exposure, and contract type composition.
Recipient UEI and legal name — primary entity identifier
Awarding agency (top-level and sub-agency)
Obligation amount per transaction and cumulative per fiscal year
Contract type — fixed-price, cost-reimbursement, IDV, grant, other
NAICS code and place of performance
33,090,050 award records · FY2021–FY2025 · 159,091 unique vendors
USASpending.gov · Federal · Annual
23
SAM.gov Active Exclusions
System for Award Management public extract of all active debarments, suspensions, proposed debarments, and voluntary exclusions. Published by the General Services Administration. Every entity prohibited from receiving federal awards is listed with the excluding agency, exclusion type, and effective date. 167,617 records ingested from the V2 public extract.
UEI — matched to USASpending vendor universe on exact key
Exclusion type — debarment, suspension, proposed debarment, voluntary exclusion
Excluding agency and effective date
163 exclusion-matched vendors with documented FY2025 obligation history
Combined FY2025 obligations among matched vendors: $202.7M
SAM.gov · GSA · Federal
24
SAM.gov Responsibility/Qualification Records (FAPIIS)
Federal Awardee Performance and Integrity Information System records, now published as Responsibility/Qualification data via the SAM.gov Data Bank. Contains all government-entered integrity records required by FAR Subparts 9.105-2, 9.406-3, and DFARS 209.105-270. Includes terminations for default and cause, administrative agreements, nonresponsibility determinations, defective pricing adjudications, indictments, convictions, and trafficking findings. 14-calendar-day delay in public posting per SAM.gov policy. 3,009 records ingested; 2,217 matched to scored vendor universe.
UEI — matched to contractor_scores on exact key
Record type — termination for cause/default, admin agreement, nonresponsibility, defective pricing, indictment, conviction
Agency name, record date, contract/grant number
995 vendors with FAPIIS records in scored corpus · 232 tier changes on F-01.2 update
Vintage: 2025-09-04 static extract · Manual refresh from SAM.gov Data Bank
SAM.gov · GSA · Federal
32
DOL/EBSA Form 5500 — Annual Return/Report
The base annual filing every covered employee benefit plan submits to the Department of Labor, IRS, and PBGC under ERISA. The spine of PensionGuard — plan type, participant counts, funding and benefit arrangements, and the schedules attached. Keyed by sponsor EIN and three-digit plan number. Sourced from the EBSA public Form 5500 datasets, latest available filing year.
Plan identity — sponsor EIN + plan number (PN), plan name, plan type
Participant counts — active, retired, separated-vested, total
Base 5500 and 5500-SF treated as mutually exclusive per plan
989,418 plans scored across four models
DOL/EBSA · Federal · Annual
33
Schedule MB — Multiemployer Actuarial
The enrolled actuary's certification for multiemployer defined-benefit plans, filed as an attachment to the Form 5500. Carries the certified funded percentage and the PPA zone status that drive the multiemployer funding score. No assumed field mappings — funded percentage and zone code read from their distinct form lines.
Certified funded percentage (line 4a — actuarial assets ÷ accrued liability)
PPA zone status (line 4b — critical, endangered, and so on)
1,200 multiemployer DB plans
DOL/EBSA · Federal · Annual
34
Schedule SB — Single-Employer Actuarial
The enrolled actuary's certification for single-employer defined-benefit plans. Carries the Adjusted Funding Target Attainment Percentage (AFTAP) and funding-target detail behind ERISA §436 — the statutory bright lines that restrict and then freeze benefit accruals as funding falls.
AFTAP and funding-target attainment detail
§436 restriction bands: ≥80%, 60–<80%, <60% (accruals frozen)
42,142 single-employer DB plans; 306 below the 60% line
DOL/EBSA · Federal · Annual
35
Schedule H — Large-Plan Financial
The audited financial schedule filed by plans above the audit threshold. Carries the independent auditor's opinion and the schedule of delinquent participant contributions — the basis for the large-plan fiduciary score. Limited-scope audit elections are separated from genuine adverse and disclaimer opinions so a routine election is never scored as distress.
Independent auditor opinion (unqualified / qualified / disclaimer / adverse)
Delinquent participant contributions (line 4a) — gross, as filed
Limited-scope (103-8/103-12) elections separated from genuine disclaimers
82,352 large plans
DOL/EBSA · Federal · Annual
36
Form 5500-SF — Small-Plan Short Form
The condensed annual filing for smaller, audit-exempt plans — the long tail of the private retirement system. Scored on a three-tier basis (no auditor-opinion axis exists below the audit threshold) from the compliance signals present: delinquent contributions and reported losses.
Delinquent participant contributions and reported losses
Three-tier scoring — no auditor opinion below the audit threshold
863,724 small plans
DOL/EBSA · Federal · Annual
37
PBGC Special Financial Assistance Roster
The Pension Benefit Guaranty Corporation's public record of multiemployer Special Financial Assistance applications and approvals, updated weekly. Matched to each plan by EIN and plan number to separate SFA-rescued plans — stabilized through 2051 — from those still declining, a distinction the zone code alone does not carry.
Application status — approved, pending, denied, withdrawn
Matched to Form 5500 plans by EIN + plan number
Among critical-and-declining plans, 94% of participants federally rescued
PBGC · Federal · Weekly
40
IPEDS Finance (F) — Audited Financial Statements
The institutional financial statements every Title IV institution reports to the National Center for Education Statistics. Filed on three distinct forms by accounting regime — GASB (public), FASB (private-nonprofit), and proprietary — each with its own field dictionary. The source of the derived Composite Financial Index for the 2,138 institutions the Department of Education does not score. Field-level measure definitions read from the IPEDS dictionaries; no assumed column mappings.
Total revenues, total expenses, and operating results
Net assets — unrestricted, restricted expendable, permanently restricted
Endowment assets and long-term debt
Four CFI strength factors: primary reserve, viability, return on net assets, net operating revenue
Three accounting regimes scored on their own dictionaries — GASB and FASB are not interchangeable
Keyed on IPEDS UnitID · FY2022–23 vintage
NCES IPEDS · Federal · Annual
41
ED Financial Responsibility Composite (FSA)
The Department of Education's own solvency score for Title IV institutions, published on a −1.0 to +3.0 scale. Below 1.0 is a federal determination that an institution is not financially responsible; 1.0–1.49 places it in the federal oversight zone. Published only for private-nonprofit and for-profit institutions — public colleges and universities carry no federal composite at all. Where ED publishes a composite, CampusIndex carries that federal number through unchanged as the solvency anchor rather than re-deriving it.
Composite score (−1.0 to +3.0) · 2,852 institutions anchored on the federal number
Federal thresholds: <1.0 not financially responsible; 1.0–1.49 oversight zone
Coverage limit: publics are not scored by ED — the gap CampusIndex closes
Institution name, type, and OPEID · AY2022–23 vintage
ED Federal Student Aid · Federal · Annual
42
IPEDS Enrollment (EF) — Multi-Year Trajectory
Fall enrollment headcounts across four collection years. The trajectory axis that turns a point-in-time solvency reading into a direction of travel — a funded balance sheet with collapsing enrollment carries different risk than a stable one. Also the source of the signal-loss flag: an institution that drops out of the enrollment file entirely has, in practice, stopped reporting.
Fall total enrollment, four collection years (2021–24)
Compound annual growth rate and trajectory band (derived)
Signal-loss flag — institution absent from the current enrollment file
49 institutions currently flagged SIGNAL_LOSS; 24 subsequently confirmed closed in ED PEPS
NCES IPEDS · Federal · Annual
43
ED PEPS — Closed-School Record
The Department of Education's Postsecondary Education Participation System record of closed institutions, matched to the scored population by primary OPEID. This is the closure ground truth CampusIndex calibrates against — the difference between a scoring model and a validated one. ED discloses a compile lag in recent months, so the most recent periods undercount closures; the taper is a reporting artifact, not evidence that closures are slowing.
Closed-school roster with closure date, matched via primary OPEID
Clean holdout: closures on/after 2024-07-01, scored on 2021–24 features, non-terminal tier at last scoring (N=90)
78.9% of scored institutions that later closed were flagged high-risk first (71 of 90)
Disclosed lag: recent-month closures undercount per ED's own compile notice
ED PEPS · Federal · Periodic
44
IPEDS HD — Institutional Directory & Crosswalk
The institutional directory and identity file. Supplies the UnitID↔OPEID crosswalk that ties the finance, enrollment, FSA composite, and closure records to a single institution without a manual match — IPEDS is keyed on UnitID, while FSA and PEPS are keyed on OPEID.
UnitID ↔ OPEID crosswalk — the federal-to-federal join key
Institution name, sector, control, and accounting regime
Institutional identity spine for the 5,772 scored institutions
NCES IPEDS · Federal · Annual
SNF Compliance Tier
Three-tier classification derived from CMS star rating and SFF designation. Non-compliant: SFF/SFF Candidate or rating ≤ 2. At-risk: rating 3. Compliant: rating ≥ 4.
Inputs: CMS overall_rating · special_focus_facility
Staffing Opportunity Score
Composite score for staffing agency prospecting. Combines RN HPRD gap vs. CMS benchmark, contract staff dependency, zero-RN-day frequency, and stability trend. Outputs LOW / MOD / HIGH / URGENT.
Inputs: PBJ rn_hprd · contract_pct · zero_rn_days · stability_score
HHA Overall Grade
Letter grade (A through F) derived from quality outcome score, patient experience score, VBP performance, and growth trajectory. Grades are AuditPoint analysis — not a CMS designation.
Inputs: CMS quality score · HHVBP score · HHCAHPS · trajectory
Acquisition Attractiveness
Composite scoring for acquisition screening. Weights compliance history, financial stability, payer mix, staffing quality, and market position. Labeled as AuditPoint analysis in all reports.
Inputs: HCRIS financials · CMS ratings · PBJ staffing · payer mix
WeighStation Score
100-point composite carrier risk score. Components: safety (FMCSA records), crash risk (24-month window), insurance stability (cancellation history), authority integrity (revocation history).
Inputs: FMCSA SMS · crash records · insurance filings · authority history
Bank Health Score
100-point composite bank health score. Five components: Capital Adequacy (25%), Asset Quality (25%), Earnings (20%), Liquidity (15%), Regulatory Standing (15%). Risk tiers: STRONG (80–100), ADEQUATE (70–79), WATCH (40–69), STRESSED (0–39). Version 2.1 (June 2026) added FFIEC RC-N loan delinquency data (Q1 2026) as the primary Asset Quality input, replacing FDIC API noncurrent ratios with fresher institution-level delinquency data. Version 2.1 also introduced a ROA floor override: institutions with ROA below −3.0% are capped at a composite score of 39 (STRESSED ceiling) regardless of other component scores. This prevents a single-quarter improvement in delinquency data from obscuring catastrophic earnings failure — the defining characteristic of terminal bank distress. Version 2.0 (May 2026) introduced an updated uninsured deposit penalty schedule, a tier cap preventing STRONG classification for institutions with >75% uninsured deposits, and a Run Architecture Flag identifying institutions matching the SVB/First Republic deposit concentration pattern. Methodology informed by FDIC Staff Study “Dissecting Depositor Flight” (Chu et al., 2026).
Inputs: FDIC Call Reports · FDIC Enforcement Actions · FFIEC CRA Ratings · FFIEC RC-B · FFIEC RC-O · FFIEC RC-N
Credit Union Health Score
100-point composite credit union health score. Five components: Capital Adequacy (25%) — net worth ratio scored against NCUA Prompt Corrective Action thresholds (well-capitalized ≥7%, adequately capitalized ≥6%); Asset Quality (25%) — noncurrent loan ratio (2+ months delinquent) and net charge-off rate; Earnings (20%) — return on assets against the system median of 0.94%; Liquidity (15%) — loan-to-share ratio against the NCUA system median of 70%; Regulatory Standing (15%) — NCUA PCA capital classification. Risk tiers: STRONG (80–100), ADEQUATE (65–79), WATCH (40–64), STRESSED (0–39). Version 1.0 (June 2026). Net income is derived from the income statement identity (interest income less interest expense plus non-interest income less operating expenses less provision) for large credit unions where NCUA Form 5300 reports net income as zero due to mid-year transfer to undivided earnings. Percentile ranks computed both system-wide and within asset-tier peer group. Institution-level enforcement orders are not incorporated: the NCUA does not publish institution-level cease and desist orders in bulk machine-readable format keyed to charter number — this is a documented data gap, not a scored zero. The regulatory standing component reflects PCA capital classification only.
Inputs: NCUA Call Reports (Form 5300) · NCUA Credit Union Registry (FOICU)
WorkRecord Score
100-point composite employer compliance score. Four components: OSHA Citation Severity (35%), WHD Wage Violations (25%), OSHA Penalty Scale (20%), Recency Weighting (20%). Risk tiers: CLEAN (80–100), MONITORED (60–79), CITED (40–59), CRITICAL (0–39).
Inputs: OSHA inspections · OSHA violations · OSHA severe injuries · DOL WHD investigations
Market Density Score
Agency or facility density relative to population 65+ by county. High density = competitive market. Very low density = potential care desert. Used in HH acquisition and state market reports.
Inputs: Census ACS 65+ population · provider count per county
NonProfits Score
100-point composite score for tax-exempt organizations. Four components: Financial Sustainability (30%), Program Efficiency (25%), Executive Accountability (25%), Transparency (20%). Peer-group benchmarking via NTEE classification (25 groups). Risk tiers: STRONG (75–100), STABLE (50–74), WATCH (25–49), CRITICAL (0–24).
Inputs: IRS Form 990 financials · Part VII exec comp · Schedule J · IRS BMF NTEE code
Hospice Compliance Index
100-point composite score for Medicare-certified hospice providers. Five components: Utilization Integrity (30%), Care Quality (25%), Patient Experience (20%), Financial Integrity (15%), Enrollment & Ownership Risk (10%). National norms bootstrapped from live CMS claims data (mean 37.31, SD 15.84). Risk tiers: COMPLIANT (75–100), MONITOR (50–74), ELEVATED (25–49), CRITICAL (0–24). Model scored all 6,943 providers on 2026-05-14 before CMS suspension list release — computed_at timestamp preserved as permanent audit trail.
Inputs: CMS Hospice Care Index (claims) · CAHPS Hospice Survey · HOPE/HIS quality measures · HCRIS cost reports · CMS provider enrollment data
Contractor Risk Score
100-point composite vendor risk score for federal contractors (F-01.2). Four components: Federal Dependency (35%) — obligation volume and YoY trend, adjusted for large commercial entities where federal revenue is incidental; Agency Concentration (30%) — HHI across awarding agencies, with elevated penalty for DOGE-risk civilian agency exposure; Contract Type Mix (20%) — fixed-price vs. cost-reimbursement vs. IDV composition; Exclusion & Integrity (15%) — graduated penalty model incorporating SAM.gov active exclusion status (hard cap) plus FAPIIS integrity records: terminations for default/cause (−20 to −25 pts), administrative agreements (−20 pts), nonresponsibility determinations (−15 pts), defective pricing (−10 pts), indictments/convictions (−30 to −35 pts). Lookback windows applied per record type (3–5 years; no limit for criminal findings). Risk tiers: STABLE (65–100), WATCH (45–64), EXPOSED (25–44), CRITICAL (0–24). DOGE-risk flag applied independently to 11,202 vendors with >50% obligation concentration in identified high-exposure civilian agencies. Tier cutoffs differ by model version by design: the population-wide F-01.1 score uses STABLE ≥70 / WATCH ≥50 / EXPOSED ≥30, while the 995 FAPIIS-enriched F-01.2 vendors use the more lenient breaks above (65/45/25) so the FAPIIS integrity penalty is not double-counted against the tier threshold.
Inputs: USASpending.gov award archives · SAM.gov active exclusions · SAM.gov Responsibility/Qualification records (FAPIIS) · HHI concentration calculation · YoY obligation trend (FY2023–FY2025)
FoodGuard Score
100-point composite food safety score for USDA FSIS-inspected establishments. Four components: Recall Severity (40%) — Class I/II/PHA history with recency weighting and a permanent Class I floor; Recall Frequency & Pattern (25%) — total recall count, repeat Class I flag, active recall flag, and pounds recovered; Recall Cause (20%) — operator culpability weighting by reason (contamination > labeling > import violation); Labor Compliance Signal (15%) — WorkRecord employer score for EIN-matched establishments, neutral for unmatched. Hard overrides: active recall caps at WATCH; Produced Without Inspection or repeat Class I caps at FLAGGED. Risk tiers: SAFE (75–100), WATCH (50–74), FLAGGED (25–49), SEVERE (0–24). Labor signal coverage disclosed per establishment — 24% of establishments have a verified EIN match.
Inputs: USDA FSIS MPI Directory · FSIS Recall Database · WorkRecord employer_compliance_scores (EIN cross-reference)
SeaSignal Vessel & Operator Score
100-point composite vessel and operator safety scores for U.S.-operating commercial vessels. Derived from USCG PSIX inspection and deficiency records — deficiency frequency, severity, detention history, and resolution status. Operator scores aggregate vessel-level performance across each operator's fleet. Risk tiers: COMPLIANT / MONITOR / ELEVATED / CRITICAL. 44,241 vessels and 2,192 operators scored.
Inputs: USCG PSIX inspections · deficiency records · detention history · NOAA AIS activity summaries
RailSignal Score
100-point composite rail carrier safety score for Class I railroads. Three components: Safety (35%) — FRA accidents per million tons hauled, anchored to the industry benchmark of 0.1786 accidents/Mt/yr (recalibrated 2026-07-08 on the corrected 3-year basis); Hazmat Risk (40%) — combines FRA hazmat-involved accident share (industry 21.5%) with PHMSA release severity (industry 1,851 units/incident); Operational (25%) — casualties per accident (industry 0.114). All components scored against fixed absolute industry benchmarks, not relative rank. Risk tiers: STRONG (80–100), ADEQUATE (65–79), WATCH (50–64), ELEVATED (0–49). Six Class I carriers scored. Scores as of 2026-07-08; benchmarks recalibrated same day on the corrected 3-year basis; PHMSA release-severity extract through 2025-07-10.
Inputs: STB Freight Commodity Statistics · FRA Form 6180.54 accidents · PHMSA Form 5800.1 hazmat incidents
PensionGuard Tiers
Four-tier funding and fiduciary scores for U.S. retirement plans, on the shared STRONG / ADEQUATE / WATCH / STRESSED scale (small-plan fiduciary uses three tiers — no auditor opinion below the audit threshold). Four models, scored on their own statutory or fiduciary basis rather than one blended composite: single-employer DB on ERISA §436 AFTAP bands; multiemployer DB on PPA zone status cross-referenced against the PBGC Special Financial Assistance roster; large-plan fiduciary on auditor opinion and delinquent-contribution disclosures (limited-scope elections separated from genuine adverse and disclaimer opinions); small-plan fiduciary on the audit-exempt 5500-SF signals. Disclosed limits: withdrawal-liability exposure is not reported at plan level on Form 5500 and is not estimated; delinquent-contribution figures are gross as filed, because correction status lives in attachment schedules, not the structured data. 989,418 plans scored.
Inputs: DOL/EBSA Form 5500 · Schedule MB (funded % + zone) · Schedule SB (AFTAP) · Schedule H (auditor opinion + delinquent contributions) · Form 5500-SF · PBGC SFA roster
CampusIndex Composite
Closure-risk tier for Title IV colleges and universities: solvency read against enrollment trajectory. Solvency is FSA-primary — where the Department of Education publishes a Financial Responsibility composite, that federal number is the anchor, carried through unchanged (2,852 institutions); where ED publishes none, chiefly the public sector, a Composite Financial Index is derived from IPEDS audited financials on the four standard strength factors (2,138 institutions). A further 782 institutions carry no solvency filing and are scored on enrollment trajectory and closure signals alone. Tier boundaries: STRONG ≥2.0, ADEQUATE 1.5–1.99, WATCH 1.0–1.49, STRESSED <1.0, with an acute CRITICAL band where distressed solvency meets enrollment collapse. The 1.5 and 1.0 breaks are ED regulatory thresholds; the 2.0 STRONG cut is an AuditPoint analytical split and is disclosed as such. Calibration: of institutions CampusIndex scored that subsequently closed, 78.9% (71 of 90 in a clean holdout) were flagged high-risk — WATCH, STRESSED, or CRITICAL — before closing, approaching the ~83% closure-prediction benchmark published by the Federal Reserve Bank of Philadelphia. Disclosed limits: (1) survivor spine — institutions frequently drop from the IPEDS spine the year they close, so many closures are never scored at all; the correct claim is "of institutions we scored that later closed," never "of all closures"; (2) composite lag — solvency derives from audited financials describing a fiscal year that closed one to two years earlier, making this a structural annual signal, not a real-time one; (3) forensic suppression — 96 institutions whose strength factors ran on a near-zero denominator (a component-reporting artifact, not distress) are suppressed from named-distress findings, leaving a published distress universe of 521 institutions. 5,772 institutions scored.
Inputs: IPEDS Finance (F) · ED FSA Financial Responsibility composite · IPEDS Enrollment (EF) · ED PEPS closed-school record · IPEDS HD crosswalk