The future is now: Corralling diversity and complexity with ‘climate digital twin’ technology

A collaborative paper by Chartis Research and Riskthinking.AI

Chartis’ ClimateRisk50 report looks at the broad vendor community involved in the various quantitative aspects of climate risk events. In this paper, we lay out the requirements and contextual challenges involved in developing a business and operational risk framework in the face of climate-related impact. In Chartis’ view, Riskthinking.AI stands out for its multi-dimensional/multi-layered syntheses, which combine diverse analytical capabilities to address broad business and operational risk challenges. 

.

Executive Summary

This document considers Riskthinking.AI and its tools/technologies and solutions in the context of Chartis’ ClimateRisk50 2024 report.

Our ClimateRisk50 report looks at the broad vendor community involved in the various quantitative aspects of climate risk events, including the full panoply of climate change risk issues. It examines several types of firms: those that develop core climate risk event models for insurance firms, those seeking to model the shift in climate dynamics and those that are translating and/or synthesizing this data and information into financial, business and market impact.

In Chartis’ view, Riskthinking.AI stands out for its multi-dimensional/multi-layered syntheses, which combine diverse analytical capabilities to address broad business and operational risk challenges that can affect multi-asset portfolios with few geographical restrictions.

In this document, we will also lay out the requirements and contextual challenges involved in developing a business and operational risk framework in the face of climate-related impact.

Finally, we briefly explore Riskthinking.AI’s Climate Digital Twin, an approach firms can take to handle complex issues faced by leading providers of climate risk analytics. We also consider why Riskthinking. AI’s model (which has different time paths and incorporates the impact of future climate change) addresses a range of questions that narrower solutions may not. And finally, we will also discuss why we believe that leveraging climate risk event data and climate change models is ‘a business and operational risk activity’ – a matter of firms understanding their locational and spatial risk in whichever direction the regulatory environment evolves.

1. Defining the context and ecosystem

The world of climate risk (including climate physical risk) and analytics is complex and multi-dimensional, with analytical models and techniques that target a variety of requirements and problems with varying degrees of maturity. Natural catastrophe (NatCat) models for such climate-related events as wildfires, hurricanes and storms have been around for many years, used by insurance firms and government agencies. Different stakeholders, including model developers, data providers and insurance carriers

and brokers, have been collecting data and refining these models for some time. In some cases, these datasets (US mortgages, say) are very granular, while the computational environments and data management frameworks are complex.

The range of models available is very diverse in terms of type. At every stage of the process, including calculating core event risk, calculating the evolution of events into the future in the presence of climate change, translating event risk into physical damage and translating loss events into financial impact for different asset classes, firms have choices and options. The macro context, along with demographic details, can also have a substantial impact (see Figure 1). The range of data requirements is equally vast and diverse, covering such areas as real asset data – including its metadata – financial data, hierarchies and climate risk factor historical and forward simulation data.

When the universe of problems is so diverse, varied and complex, it is no surprise that there is a large variety of solutions (and vendors) available to address it. Chartis’ ClimateRisk50 report tabulates and analyzes this diversity. However, several broad structural questions often require a multidimensional approach – in essence, the ‘climate digital twin’ – to marry diverse models and risk factors.

Figure 1: Context is everything - model performance, inputs and options

2. Rethinking the business and operational impact of climate risk

Climate risk and climate change risk affect all firms, regardless of the geography and the nature of the business. Yet the impact is far greater on certain businesses in certain jurisdictions and geographies than on others. All owners of physical assets (e.g., real estate, hospitals, power plants, infrastructure and power networks) are concerned about the physical risk they are exposed to over the life of their infrastructure. Many other businesses (such as logistics, distribution and warehousing) and non- commercial operators (such as government emergency services) also need to plan and structure their thinking about the exposure and challenges their operations face when it comes to climate risk.

Firms with portfolios of financial assets that are exposed to these themes also constitute a critical constituency. Entities with significant data centers, for example, may need to construct appropriate credit or equity shock scenarios. In some situations, risk scenarios may spiral to the point of rendering certain business types or geographies unviable. This is a particular issue when attempting to think decades ahead.

Increasingly, the use of business analytics that leverage locational intelligence and spatial climate risk is becoming a mature discipline. Chartis believes that modeling long-term climate risk and integrating it into an appropriate physical risk framework is a cross-disciplinary activity, in which software, data availability and computation have converged to make it a feasible option for commercial enterprises (see Figure 2). These challenges require both technology and model-oriented solutions.

Figure 2: Modeling the climate - integrating models into risk frameworks.

The organization and structure of diverse datasets is the first step in executing appropriate climate modeling. However, it is vital that firms consider other important activities:

• Layering a broad range of risk factors.

• Ensuring that a diverse range of analytical contexts is considered, including climate event risk, macro factors, demographics and risk factor models from transnational and scientific advisory bodies.

• Taking into account the fact that not all models target similar factors.

• Managing the integration of models and required datasets.

3. The 'Climate Digital Twin' and Riskthinking.AI's answer to complex and climate change risk issues

As we have noted, however, there are many dimensions and approaches to consider when examining the climate risk problem.

Riskthinking.AI combines granular data and stochastic forward-looking models to cover a broad range of climate scenarios. This approach enables the development of the company’s Climate Digital Twin (CDT), which powers all its analytics and products. With this tool, firms can have the physical assets to understand, analyze and forecast the impact of climate change risk and climate uncertainty on their operations and operating businesses. A broad range of businesses, including energy firms, infrastructure providers, logistics providers and real estate firms, can leverage the platform.

The CDT combines broad, global, granular data with a multi-factored modeling approach that explores a broad set of risk factors, including natural and climate-linked factors, macro factors, demographic drivers and financial impact factors, as well as correlating dynamics.

In Chartis’ view, this approach provides a balance between hyper-granular and geographically restricted solutions and those that model and analyze specific perils (such as wildfires in the American Southwest). In taking a wider, more integrated view, Riskthinking.AI can broaden the potential user group to any business, or other organization, looking at the current or future risk exposure of its physical and operational assets.

4. Concluding observations

Climate change and climate risk impact a broad range of institutions across the globe, and climate risk models and data have many dimensions, including being insurance-focused, regulatory-focused or disaster management-focused. Mapping data and model outputs from one dimension to another is a challenge. Some issues, such as regulatory capital risks and the appropriate models that different institutions should adopt, are works in progress.

Unsettled though these issues may be, the business planning and operational optimization of physical asset portfolios, and the measuring of climate risk/change risk impact on physical portfolios, are problems that have available solutions.

Chartis believes that, as the technological and analytical foundations continue to grow rapidly (such
as global satellite networks) and, as a consequence, the rich physical datasets currently available, sophisticated locational and spatial modeling capabilities will offer a strong foundation on which to build.

Riskthinking.AI’s Climate Digital Twin (CDT) approach brings together diverse data, robust computational frameworks and data management tools, combining physical assets, climate hazards, emissions and socioeconomic factors. This approach enables businesses (and other decision-makers) to apply a consistent multi-factor, multi-layered analytical lens to a wide variety of global problems, from infrastructure to real estate portfolios and everything in between. The CDT is also integrated natively with AWS and Google Cloud to provide a simple, secure and scalable way to access data.

Businesses, governmental organizations and NGOs examining the consequences of climate risk and the impact of climate change on their operations or strategies, and considering how to mitigate and control that impact, can benefit from the CDT framework’s careful and deep integration of data and analytics, which enables them to build a data-driven and analytically consistent risk/business scenario management capability.