Insight: Physical Climate Risk in Asset Management

Winners Consulting Services Co. Ltd. (積穗科研股份有限公司), Taiwan's expert in Enterprise Risk Management (ERM), draws urgent attention to a 2025 quantitative study published on arXiv that demonstrates, for the first time using a mathematically rigorous jump-diffusion model, that physical climate risks — including floods, heat waves, and water scarcity — cause portfolio expected losses to increase continuously over time, with significantly higher exposure concentrated in asset-intensive industries. For Taiwanese manufacturers, financial institutions, and listed companies, this research signals that climate physical risk must be elevated to a priority tier within their ERM frameworks, risk matrices, and KRI systems without delay.

About the Authors and This Research

This paper is co-authored by Michele Azzone, Matteo Ghesini, and Davide Stocco — three researchers with backgrounds in quantitative finance and mathematical modeling. Michele Azzone is affiliated with the Department of Mathematics at Politecnico di Milano (Milan Polytechnic), one of Italy's most prestigious technical universities, and specializes in stochastic processes, credit risk quantification, and financial mathematics. While his current h-index stands at 1 with 6 cumulative citations — indicative of an early-career researcher — the methodological contribution of this paper is notable: it bridges the gap between climate science narratives and rigorous financial risk modeling in a way that very few academic papers have achieved to date.

The paper's core innovation is the integration of climate-driven downward jumps into the classical Vasicek credit risk model, creating a multivariate firm value model with jumps. Unlike qualitative climate risk assessments, this framework is calibrated using observable company-level data — specifically, asset intensity and geographical exposure — making it directly applicable to portfolio risk management and enterprise-level KRI design. For Taiwanese corporate risk officers and board-level risk committees, the significance of this work lies not in its citation count, but in the practical quantitative logic it introduces for translating physical climate events into measurable financial losses.

Climate Jump-Shock Model: Quantifying Expected and Unexpected Portfolio Losses from Extreme Weather

The central research question this paper addresses is: how can asset managers and enterprise risk managers scientifically quantify the financial impact of extreme climate events — floods, heat waves, and water scarcity — on their portfolios and corporate assets? Traditional credit risk models, including the widely-used Vasicek framework, do not account for the discontinuous, shock-like nature of climate-driven asset impairment. This paper's methodological contribution is the introduction of downward jumps into the stochastic dynamics of firm asset values, specifically designed to mirror the negative shocks that extreme climate events impose on corporate assets and productivity.

The model calibration methodology relies on two key inputs: asset intensity (the ratio of a company's physical assets to its total assets or revenues, indicating how exposed a firm is to physical climate damage) and geographical exposure (where the company's assets are physically located and what climate hazards those locations face). By combining these two dimensions, the model generates sector-specific and geography-specific risk profiles that can be directly incorporated into portfolio loss calculations.

Core Finding 1: Expected Losses Rise Continuously Over Time, with Pronounced Sectoral Divergence

When the multivariate jump model is applied across multiple sectoral indices, the results show a consistent upward trajectory in expected losses (EL) over time. This is not a one-time shock effect — the cumulative nature of climate physical risk means that even without a single catastrophic event, the persistent probability of climate-driven asset impairment compounds into ever-increasing expected portfolio losses. More importantly, the research reveals pronounced differences in exposure across sectoral indices. High asset-intensity sectors — including manufacturing, energy, utilities, and infrastructure — face materially higher physical climate risk exposure compared to asset-light service industries. For Taiwan's export-oriented manufacturing sector, which includes semiconductor fabrication, precision machinery, and petrochemicals — industries with high fixed-asset intensity concentrated in geographically specific industrial zones — this finding carries direct and immediate relevance.

Core Finding 2: Additional Capital Buffers Are Required, Especially for Asset-Intensive Industries

The paper's second major finding, with direct implications for both corporate ERM and regulatory policy, is that existing capital structures are insufficient to absorb the unexpected losses (UL) generated by extreme climate events. The authors explicitly recommend, from an environmental policy perspective, the need for additional capital buffers in sectors with high asset intensity. This conclusion aligns with the direction of global financial regulation — the Basel Committee on Banking Supervision has been progressively incorporating climate risk into capital adequacy frameworks since 2021, and Taiwan's Financial Supervisory Commission (FSC) has been advancing TCFD-aligned disclosure requirements since 2022. For Taiwanese listed companies, this finding signals that climate physical risk is not merely a disclosure compliance issue, but a capital adequacy and financial resilience challenge that must be integrated into the core of their COSO ERM 2017 risk response strategies.

Implications for Taiwan's Enterprise Risk Management (ERM) Practice

For Taiwanese enterprises, the implications of this research are both strategic and operational. At the strategic level, this study provides quantitative validation for a shift that many Taiwanese corporate risk managers have been slow to make: elevating physical climate risk from a peripheral ESG narrative to a core element of enterprise risk management. ISO 31000:2018, in Section 6.4.2 on risk identification, explicitly requires organizations to identify all potential risk sources from external context — extreme climate events are a textbook example of such external context risks. Yet in practice, many Taiwanese companies' risk matrices still treat climate events as residual or low-probability risks rather than as a distinct, quantifiable risk category with its own KRI indicators.

At the operational level, the COSO ERM 2017 framework's "Strategy and Objective-Setting" component requires that organizations consider how risk appetite intersects with strategic choices. For asset-intensive Taiwanese manufacturers with significant fixed investments in geographically concentrated industrial zones (such as the Taichung Industrial Park, Kaohsiung's Linhai Industrial Zone, or the Central Taiwan Science Park), the climate jump-shock model's findings suggest that current risk appetite statements may be underestimating physical climate exposure by failing to account for the compounding nature of expected losses over multi-year horizons.

Three specific ERM practice improvements are immediately actionable for Taiwanese enterprises: First, risk matrices must be updated to include physical climate risk as an independent risk category, with sub-categories for flood risk, heat stress, and water scarcity, each with defined probability and impact ratings calibrated to the company's asset intensity and geographic footprint. Second, KRI systems must incorporate climate geographic exposure indicators — such as the percentage of fixed assets located in flood-prone zones, or the proportion of production processes dependent on water-intensive operations. Third, board-level risk governance must establish climate scenario stress-testing capabilities, producing quantified estimates of expected losses and unexpected losses under different climate pathways, as required by the FSC's TCFD disclosure roadmap for companies with paid-in capital exceeding NTD 10 billion.

How Winners Consulting Services Helps Taiwanese Enterprises Integrate Physical Climate Risk into ERM

積穗科研股份有限公司（Winners Consulting Services Co. Ltd.） assists Taiwanese enterprises in implementing ISO 31000 and COSO ERM frameworks, building risk matrices and KRI systems, and strengthening board-level risk governance capabilities. In response to the physical climate risk challenge highlighted by this 2025 research, Winners Consulting offers the following targeted services:

1. Physical Climate Risk Inventory and Risk Matrix Integration: Following ISO 31000:2018 risk identification principles, Winners Consulting systematically maps the climate geographic exposure of each facility, supply chain node, and asset cluster. Flood risk, heat wave exposure, and water scarcity risk are integrated into the existing enterprise risk matrix as independent risk categories, with risk ratings calibrated to asset intensity levels consistent with this paper's sectoral findings. The result is a risk matrix that genuinely reflects the quantitative logic of climate jump-shock modeling rather than relying on qualitative guesswork.
2. Climate KRI Design and Scenario Stress Testing: Drawing on the jump-shock model's calibration logic, Winners Consulting designs measurable climate KRIs — including flood-exposed asset ratios, heat-stress production loss estimates, and water dependency risk scores — and builds scenario stress-testing models under the COSO ERM 2017 "Risk Response" component. These models quantify expected losses and unexpected losses under different extreme climate pathways, providing the capital buffer analysis that this paper's findings recommend for asset-intensive industries.
3. Board-Level Climate Risk Governance and TCFD Disclosure Preparation: Winners Consulting establishes board-level climate risk oversight structures aligned with the FSC's TCFD disclosure roadmap, ensuring that physical climate risk identification, assessment, response, and reporting flows meet the ISO 31000 risk governance cycle. This directly prepares listed companies for the TCFD disclosure requirements that apply to companies with paid-in capital exceeding NTD 10 billion from 2025 onward.

Frequently Asked Questions

How should Taiwanese enterprises integrate physical climate risks such as floods and heat waves into their ERM frameworks?

The most effective approach is to add a dedicated "Physical Climate Risk" category to the existing risk matrix, with sub-categories calibrated to the company's asset intensity and geographic exposure. ISO 31000:2018 Section 6.4.2 requires identification of all external context risks — extreme climate events are a direct application of this requirement. Enterprises should design climate-specific KRIs (such as the percentage of fixed assets in flood-prone zones), conduct annual climate scenario stress tests, and report results to the board risk committee. Companies that have already implemented COSO ERM 2017 can directly leverage the "Strategy and Objective-Setting" and "Risk Response" components to structure this integration.

What are the regulatory compliance requirements for Taiwanese listed companies regarding climate risk disclosure?

Under the FSC's Corporate Governance Sustainability Development Roadmap, listed companies with paid-in capital exceeding NTD 10 billion are required to complete TCFD-aligned climate risk disclosure by 2025, with all other listed companies required to follow by 2027. TCFD requires disclosure of both physical risks (such as extreme weather impacts on assets) and transition risks (such as carbon pricing and regulatory changes). Enterprises that have implemented a COSO ERM 2017-aligned risk management framework can directly map to TCFD's four disclosure pillars — Governance, Strategy, Risk Management, and Metrics & Targets — significantly reducing compliance preparation costs.

What is the difference between ISO 31000 and COSO ERM for climate risk management? Which should a Taiwanese enterprise adopt?

ISO 31000:2018 is a principles-based guidance framework applicable to all organizations and all risk types, emphasizing risk management culture and a continuous improvement cycle. COSO ERM 2017 is more focused on integrating risk appetite with enterprise strategy and provides more operationally specific architecture. For climate physical risk management, the two are complementary rather than competing: ISO 31000 provides the foundational principles and process structure for risk identification, assessment, and monitoring, while COSO ERM's "Strategy and Objective-Setting" and "Risk Response" components provide the strategic decision-making structure for climate scenario analysis and capital allocation. Winners Consulting typically recommends implementing both frameworks in parallel to maximize the completeness and international comparability of the enterprise ERM system.

How long does it take to implement a climate physical risk ERM system, and what are the specific steps?

A complete implementation typically requires 90 to 180 days, structured in four phases: Phase 1 (Days 1–30): Current state diagnostic — mapping existing ERM mechanism coverage gaps against ISO 31000 and TC