Abstract: This episode of the Sustainable Finance Podcast features Dr. Ron Dembo, founder and CEO of RiskThinking.AI, discussing innovative approaches to modeling climate risk in finance. Dr. Dembo explains the importance of probabilistic models, or “future data,” over deterministic forecasts to address the uncertainties of climate change. He highlights key challenges, such as the lack of detailed data on physical assets and the limitations of traditional methods, while showcasing RiskThinking.AI’s Climate Earth Digital Twin Platform. The discussion emphasizes the evolving role of regulators, insurers, and banks in adopting stochastic methods and provides practical guidance for leveraging these tools to integrate climate risk into financial systems effectively.
To view the podcast in its entirety, please click here.
RiskThinking.AI and Future Data
Paul: Let’s start with the core idea behind your work at RiskThinking.AI. It seems important to you that people understand this isn’t just forecasting or prediction. Can you elaborate?
Ron: Absolutely. RiskThinking.AI focuses on the concept of “future data.” Imagine consulting a thousand scientists, each with unique models and assumptions. While no single model is entirely accurate, together, they form a valuable ensemble of possibilities. When pilots chart flight plans, they anticipate turbulence and adjust mid-flight. Similarly, we must treat climate projections as dynamic, adapting as new information emerges. Our approach emphasizes preparing for uncertainty rather than relying on rigid scenarios.
Challenges in Climate Data and Physical Risk Assessment
Paul: What are the pain points organizations face when using climate data effectively?
Ron: The first major challenge is understanding the physical nature of the assets being assessed. Banks often lack detailed data about the physical infrastructure of their counterparties, which impedes accurate risk pricing. Secondly, representing climate futures requires viewing data as probability distributions rather than fixed numbers. This involves modeling location-specific hazards—such as sea-level rise or heat waves—over time. Without this detailed, stochastic perspective, many financial decisions are based on incomplete information.
Stochastic Modeling vs. Deterministic Approaches
Paul: Do you see regulators and industries improving in their understanding and integration of stochastic models?
Ron: Regulators and insurers are recognizing the limitations of traditional deterministic models. However, many lack the mathematical expertise to handle radical uncertainty. While some, like regulators in Canada, are taking steps to understand and implement these methods, the transition is slow. The insurance industry, deeply rooted in backward-looking data, is beginning to explore stochastic approaches but struggles with access to advanced technology.
Mark-to-Market Pricing and Climate Risk Integration
Paul: Will we ever achieve a mark-to-market pricing system that inherently includes climate risk?
Ron: Yes, we can and will. For example, pricing loans currently ignores the potential impact of climate change on collateral values. By integrating stochastic data into these calculations—like stress-testing physical assets—we can refine loan pricing to reflect true climate risks. This is already underway in collaboration with regulators and some financial institutions.
Tools and Resources for Understanding Climate Risks
Paul: Where can our listeners learn more about your work and explore these concepts?
Ron: Our website, RiskThinking.AI, offers resources and a free tool where users can explore climate impacts under various scenarios. For instance, you can compare companies’ climate risks or analyze stochastic flooding data. We also share insights on LinkedIn and encourage direct outreach via Climate@RiskThinking.ai. We’re happy to engage with anyone interested in these discussions.
Closing Thoughts
Paul: Thank you, Dr. Dembo, for sharing your insights. It’s been a pleasure having you on the Sustainable Finance Podcast. To our listeners, stay tuned for more conversations on this evolving field.
Ron: Thank you, Paul. I look forward to continuing the dialogue.