CLIMA

Score Calculation Methodology

Climate-Linked Infrastructure & Mortgage Analytics

Technical Documentation for Investors & Stakeholders


Overall CLIMA Score Formula

CLIMA Score = (A × 30%) + (B × 20%) + (C × 20%) + (D × 15%) + (E × 15%)

Key Principle: Higher score = Safer (lower climate risk). All factors normalized to 0-10 scale.

Five Scoring Factors

A. Climate Risk Progression 30%
Historical disasters (15%) + Climate projections (85%). Uses FEMA data + NOAA/NASA SSP scenarios.
B. Relief Allocations per Capita 20%
FEMA Public Assistance funding per capita. Higher relief = more disaster history = lower score.
C. Municipal Fiscal Stress 20%
BLS unemployment (50%) + FRED bonds (25%) + Census finances (25%). Lower stress = higher score.
D. Personal Finance Resilience 15%
Census ACS: Income (40%), Value-to-Income (30%), Housing burden (30%). Better finances = higher score.
E. Insurance Availability 15%
Census ACS income/home value ratio. More affordable = higher score.

Score Interpretation

Score Range Grade Risk Level Description
9.0-10.0 A+ Minimal Risk Excellent climate resilience
7.0-8.9 A Low Risk Low climate-linked mortgage risk
5.0-6.9 B Moderate Risk Standard underwriting appropriate
3.0-4.9 C Elevated Risk Consider pricing adjustments
0.0-2.9 D/F High Risk Very high risk - consider declining

Data Sources (All Real, Public Data)

FEMA OpenFEMA API — Disaster declarations (1953-2024) and Public Assistance funding
U.S. Census Bureau ACS — Income, home values, housing costs, demographics (5-year estimates)
Bureau of Labor Statistics — Unemployment rates via LAUS (Local Area Unemployment Statistics)
FRED (Federal Reserve) — Municipal bond spreads, Treasury yields, fiscal indicators
NOAA/NASA Climate Projections — SSP2-4.5 and SSP5-8.5 scenarios for flood, wildfire, heat, wind
NFIP Claims Data — Flood loss severity validation (30.3% average on 717 claims)

Model Validation Performance

0.842 Default Prediction AUC — Exceeds 0.80 industry standard
0.946 Disaster Prediction AUC — Exceeds 0.80 by 18%
Both models validated — 145,452 observations and 68,485 FEMA disasters, 68,485 disasters, forward-looking validation
68,485 disasters validated — 71-year backtest (1953-2024), 13x larger than typical studies
70% loss reduction — Approved segment has 70% fewer major disasters than rejected segment
No look-ahead bias — All predictions use only past data, proper time-series validation

Key Design Principles

Future-focused — 85% weight on climate projections for 30-year mortgage commitments
Transparent — All data sources are public and reproducible, no black-box models
Fair lending aligned — Geographic factors only, no demographic discrimination (ECOA)
Production ready — Real-time API integration with authoritative sources

Claims & Validation

Key claims are traceable to: DATA_SOURCES.md, REPRODUCIBILITY.md, EXTERNAL_TECH_AUDIT_FORMULAS_AND_MODELS.md, and backtest/elite_constants.py. Regulatory: inputs for CECL/Basel III; aligned with OCC/FDIC/Fed climate risk guidance (not certified).

CLIMA — Climate-Linked Infrastructure & Mortgage Analytics
For more information, visit our technical documentation or contact us for API access.