Climate Risk Stress Testing in California: A Geospatial Framework for Banking and Climate-Exposed Sectors

Battle Count: 70

The paper is primarily focused on banking risk management, climate stress testing, and financial stability rather than quantitative trading strategies. While the Climate-VaR metric and expected loss formulations could inform risk-adjusted position sizing or sector allocation decisions, the paper does not address trading signals, execution, market microstructure, or alpha generation. Its relevance to quantitative trading is indirect, through portfolio risk management and sector exposure monitoring for climate-sensitive assets.

Implementing the full framework requires: (1) assembling and harmonizing multi-source geospatial data (NOAA, CAL FIRE, FEMA, county assessor records) at ZIP/county level; (2) linking hazard layers to loan-level and portfolio-level financial data (HMDA, call reports); (3) estimating scenario-contingent PD and LGD models with physical and transition risk covariates; (4) constructing the Climate-VaR aggregation across heterogeneous instruments; (5) designing and running multiple stress scenarios including compound events; (6) maintaining data pipelines for ongoing portfolio monitoring. The interdisciplinary nature (climate science + geospatial analysis + financial modeling) adds significant complexity.

The paper is primarily a conceptual framework and research design proposal rather than an empirical implementation. It outlines data sources (NOAA, CAL FIRE, FEMA, HMDA, bank call reports) and mathematical formulations but does not present actual estimation results, code, or a completed empirical exercise. Reproducibility would require assembling the multi-source geospatial and financial datasets described and implementing the three-layer architecture independently.