💰 quantitative finance

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

This paper presents a geospatial framework for conducting climate risk stress tests in California, integrating physical hazard mapping and sector-specific exposure analysis to assess the financial impacts of climate risks on banking and key industries like agriculture, real estate, and tourism.

Original authors: Satya Narayana Panda, Aishworzo Saha

Published 2026-04-21

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Satya Narayana Panda, Aishworzo Saha

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are the captain of a massive ship (the banking system) sailing through the Pacific Ocean. For a long time, you've been worried about storms that might happen 50 years from now. But this paper argues that the storm is already here, and it's hitting specific parts of your ship right now.

The authors, Satya Panda and Aishworzo Saha, are saying: "Stop looking at the whole ocean as one big blur. We need to look at the map."

Here is the paper explained in simple terms, using some everyday analogies.

1. The Core Problem: The "Zip Code" Blind Spot

Traditionally, banks look at risk like a grocery store manager. They ask, "Do we have too many apples? Too many oranges?" (Sector risk). If they have a mix of different industries, they think they are safe.

The Paper's Insight:
This is like a grocery store thinking it's safe because it sells apples, oranges, and bananas. But what if a giant hailstorm hits only the orchard where they get their apples? Suddenly, they have no apples, and the oranges don't matter.

In California, the "hailstorm" is climate change (wildfires, droughts, floods). The paper argues that a bank's portfolio might look diverse on paper, but if all their "apples" (loans) are in the same geographic zip code that is about to burn down, they are in deep trouble. You can't manage climate risk without a geospatial map.

2. The Solution: A "Weather Forecast" for Money

The authors propose a new framework—a Geospatial Stress Test. Think of this as a super-advanced weather forecast, but instead of predicting rain, it predicts financial ruin.

They break it down into three layers:

Layer 1: The Hazard Map. Where are the wildfires? Where is the drought? (The "Storm Clouds").
Layer 2: The Exposure Map. Which houses, farms, and businesses are standing right under those clouds? (The "Vulnerable Houses").
Layer 3: The Wallet Impact. If those houses burn or the farms dry up, how much money does the bank actually lose? (The "Broken Wallet").

3. The Four "Transmission Channels" (How the Storm Hits)

The paper explains how a physical disaster turns into a financial disaster. Imagine a domino effect:

Cash Flow Impairment: A farmer in the Central Valley loses their water. They can't grow crops. They can't pay their loan. Result: The bank loses money.
Collateral Repricing: A homeowner in a wildfire zone wants to sell their house. No one wants to buy it because it might burn down next year. The house value drops. Result: The bank's "safety net" (the house) is now worth less than the loan.
Operating Disruption: A hotel in a coastal city gets flooded. They can't open for the summer. They can't pay their staff or their mortgage.
Market Spillovers: If a whole town is hit by a disaster, the local economy crashes. Even businesses that weren't directly hit (like a local bakery) might fail because no one has money to spend.

4. The Scenarios: "What If" Stories

The paper suggests banks shouldn't just guess; they should run specific "What If" stories (Scenarios):

The "Orderly Transition": The government slowly raises taxes on carbon. Businesses adjust slowly. It's a gentle slope.
The "Disorderly Transition": The government suddenly bans something. Markets panic. Prices crash overnight.
The "Physical Shock": A massive wildfire season hits. Homes burn, crops die.
The "Compound Scenario": The worst of all worlds. A wildfire hits while interest rates are high and the economy is weak.

5. Why California?

California is the perfect "Laboratory" for this. It has everything:

Wildfires (burning the West).
Droughts (drying the Central Valley farms).
Floods (threatening the coast).
Heat (straining cities).

Because California has so many different types of climate risks in one place, it's the best place to test if this new "Map-Based" banking system works.

6. The Takeaway for Everyone

For Banks: Stop just saying "We care about the environment." Start mapping your loans. If you have 50% of your loans in a high-fire zone, you need to set aside more money to cover potential losses, just like you do for bad credit.

For Investors: Don't just look at a company's stock price. Look at where their factories and offices are. If they are in a flood zone, their stock might be a "time bomb."

For Regulators: It's not enough to make companies tell you about their climate risks (Disclosure). You need to force them to calculate how those risks will break their balance sheet (Stress Testing).

The Bottom Line

This paper is a call to action. It says: Climate change isn't just an environmental problem; it's a geography problem.

If you don't look at the map, you won't see the storm coming until it's too late. By using this new framework, banks and investors can see exactly where the water is rising and move their assets to higher ground before the flood hits.

Based on the paper "Climate Risk Stress Testing in California: A Geospatial Framework for Banking and Climate-Exposed Sectors" by Satya Narayana Panda and Aishworzo Saha, here is a detailed technical summary.

1. Problem Statement

The paper addresses the critical gap in current financial risk management systems: the failure to adequately integrate granular physical climate hazards with financial portfolio performance.

The Core Issue: Traditional risk models often capture borrower, sector, and macroeconomic factors but fail to map these to specific geographic locations or the interaction between physical risks (e.g., wildfires, drought) and transition risks (e.g., regulatory changes).
The Context: California serves as a high-stakes laboratory due to its diversified capital markets and concentrated exposure to specific hazards (wildfires, drought, coastal flooding, extreme heat).
The Consequence: A balance sheet may appear diversified at the sector level but remain highly concentrated geographically, leading to unquantified vulnerabilities in mortgage collateral, agricultural production, tourism revenue, and municipal tax bases.

2. Methodology: The Geospatial Framework

The authors propose a three-layered, finance-centered framework designed to link local hazard intensity to portfolio-level financial outcomes.

A. Three-Layer Architecture

Hazard Measurement: Quantifies physical hazards ( $H_{g,s}$ ) for a geographic unit $g$ (e.g., ZIP code, county) under scenario $s$ . Metrics include wildfire intensity, drought severity, flood probability, and heat stress.
Financial Exposure Mapping: Links these hazards to the specific location and sectoral composition of loans, properties, businesses, and local fiscal bases.
Balance-Sheet Transmission: Translates exposures into expected loss, asset repricing, and capital strain.

B. Mathematical Formulation

The framework utilizes scenario-contingent expected loss models:

Expected Loss (EL): Defined as $EL_{i,s} = PD_{i,s} \times LGD_{i,s} \times EAD_{i}$ $E L_{i, s} = P D_{i, s} \times L G D_{i, s} \times E A D_{i}$ .
- The Probability of Default ($PD$) is modeled as a function of physical hazards ( $H$ ), transition risks ( $T$ ), local economic fragility ( $U$ ), and adaptation capacity ( $A$ ):
  $PD_{i,s} = PD^0_i \exp(\beta_1 H_{g,s} + \beta_2 T_{j,s} + \beta_3 U_g - \beta_4 A_i)$
Climate Value-at-Risk (Climate-VaR): An illustrative metric for valuation-sensitive portfolios that aggregates mark-to-market repricing ( $\Delta V$ ) and credit impairment ($EL$):
$Climate\text{-}VaR_s = \sum_{i=1}^{N} w_i \Delta V_{i,s} + \lambda \sum_{i=1}^{N} EL_{i,s}$

C. Scenario Design

The framework moves beyond point forecasts to "disciplined stress narratives" combining acute shocks and structural change:

Orderly Transition: Gradual policy tightening affecting carbon-sensitive cash flows.
Disorderly Transition: Abrupt regulation causing spread widening and valuation adjustments.
Physical Risk Shock: Clusters of wildfires, floods, or droughts causing defaults and collateral impairment.
Compound Scenario: Physical shocks combined with tighter financing conditions, creating joint pressure on capital and liquidity.

D. Data Strategy

Implementation requires merging granular datasets:

Climate/Physical: NOAA, state climate records, CAL FIRE, FEMA hazard layers.
Financial/Real Estate: County assessor files, property transactions, bank call reports, HMDA data.
Economic: Agricultural production statistics and municipal finance disclosures.

3. Key Contributions

Channel Organization: The paper categorizes climate risk into four specific financial transmission channels:
- Cash-flow impairment.
- Collateral repricing.
- Operating disruption.
- Broader market spillovers.
Geospatial Integration: It proposes a novel approach to link county- and asset-level hazards directly to portfolio-level expected loss, moving away from aggregate sector analysis to location-specific risk assessment.
Empirical Agenda & Survey: It outlines a practical research design and survey instrument to benchmark current institutional preparedness across banking, agriculture, real estate, and tourism sectors.

4. Results and Testable Predictions

While the paper primarily establishes a framework rather than presenting a completed empirical estimation, it derives three specific, testable predictions for future research:

Credit Spreads: Borrowers in high-intensity hazard zones (wildfire, drought, flood) should exhibit higher credit spreads and weaker performance, even after controlling for borrower quality and local economic conditions.
Collateral Discounts: Areas with climate stress that weakens property-market liquidity should show stronger collateral discounts, even if the underlying hazard changes slowly.
Concentration Risk: Institutions with geographically concentrated exposures will face disproportionately large stress losses compared to institutions with similar aggregate sector compositions but better geographic diversification.

5. Significance and Implications

For Financial Institutions: The paper argues that climate risk programs must move beyond broad ESG language to concrete exposure mapping, underwriting sensitivity, and collateral resilience. Institutions must track geographic concentrations and re-estimate sensitivity to local hazards.
For Regulators: While disclosure regimes (e.g., California SB 253/261) improve comparability, the paper emphasizes that disclosure alone is insufficient. Financial stability depends on converting climate data into capital planning, pricing, and governance decisions.
For Academic Research: It bridges the gap between academic finance, prudential supervision, and applied risk management by treating climate risk as a finance problem with geographic structure.
Sectoral Impact: The framework highlights that sectors like agriculture (water stress), real estate (collateral value), and tourism (cash flow) are not peripheral but central transmission mechanisms for climate shocks to the broader financial system.

In conclusion, the paper asserts that credible climate stress testing requires a framework that is explicitly geospatial, sector-aware, and tied to core balance-sheet metrics to treat climate risk with the same rigor as credit, liquidity, and market risk.