New frontiers in climate-risk analytics

In recent years, record-breaking temperatures and extreme weather events have highlighted the overwhelming impact of greenhouse-gas (GHG) emissions on the global climate.

Moreover, the costs of such events are mounting. For example, five of the worst natural disasters in US history have occurred since 2005, causing economic damage totaling $523 billion in inflation-adjusted terms. And America has suffered 22 major natural disasters in the last year alone.

But translating the outputs of climate-change models into specific potential impacts, and gauging the financial materiality of climate hazards, presents challenges for both businesses and investors.

The rapid uptake of model-driven climate data has fueled concerns about unintended misuse in the context of financial decision-making and disclosures, as well as about material misstatements in financial reports and greenwashing. These risks are particularly problematic in the case of long-term capital investments in public infrastructure, which often have a multi-decade operational lifespan.

Financial market participants’ need for climate information varies, in terms of both granularity of assessment (regarding specific assets or asset classes, regions, and sectors) and time horizons.

But it is difficult to assess measures to mitigate climate exposures without specific data on entities’ past performance. This may include how businesses have been affected by historic events such as flooding, the timing and geographic scale of hazards and their impact, and the effectiveness of adaptation.

While there is no one-size-fits-all solution when it comes to pricing in climate-related risks and opportunities, some processes have high priority. For example, standardization can help to avert maladaptation to climate change by ensuring the consistent application of data sets and taxonomies, as well as reduce reliance on climate-model outputs and proxies.

Standardized, geographically specific disclosures relevant to credit risks would also allow for comparable assessments of climate-related risks and opportunities – and their potential impact.

Another approach – enhanced climate-risk analytics – involves supplementing climate-model outputs with entityspecific data, including asset-level data and financial information. A clear view of an entity’s assets makes it much easier to understand the possible financial impact of the physical effects of climate change.

This analysis can also facilitate dialogue with decisionmakers to understand their perspective about the acute and chronic climate risks they face, and how they manage, monitor, and mitigate them.

Finally, using multiple climate scenarios enables decisionmakers to consider a broader range of possible outcomes. This helps them to build organizational resilience and identify risks and opportunities before they emerge, in turn enabling more productive deliberation about the interventions that may be required.

Although climate-risk analytics, dialogue with entities, and expert judgment can all improve analysis, the next generation of climate models will need to be more sophisticated to account better for global warming’s complexities. Climate hazards do not occur in isolation or respect sectoral and geographical boundaries. And the further progression of climate change may give rise to new, complex interdependencies and interactions that data providers are unable to resolve due to the siloed nature of existing models.

Non-equilibrium models, which assume more complex relationships between climate variables, could be one viable alternative. Similarly, integrated assessment models offer the potential to group multiple models together in order to understand the impact chains that join environmental, socioeconomic, and climatic systems. IAMs can also assess the effects of GHG-mitigation efforts and adaptation actions on the climate system and, in turn, gauge the efficacy of associated strategies.

But non-equilibrium models and IAMs are not a panacea. For example, IAMs cannot measure the economic damage caused by certain events, such as severe storms, or calculate the costs associated with adaptation.

Moreover, such models are typically calibrated to the change in global mean temperature. This limits their insights regarding changes in extreme events such as storms and flash floods, which are a major concern for many financialmarket participants, including insurers. Furthermore, models such as IAMs are inherently complex, produce large outputs, and are expensive to run, meaning that many of the challenges facing the current generation of climate models are likely to apply to the next generation as well.

There currently is no perfect solution for assessing the financial effects of physical climate change, but this should not be an excuse for inaction. Enhanced climate-risk analytics can provide a clearer picture of how bad – or expensive – global warming could become for businesses. While technology will develop apace to help companies’ climate-risk assessments, analytical judgment is needed more than ever to interpret model outputs and inform better decision-making. After all, in a fast-changing field like climate-risk analytics, the past provides only a narrow, shortterm view of the future.

Such an approach will also help to prevent the unintended consequences and misuse of climate-model outputs by financial-market participants who increasingly need to disclose publicly their exposures to climate risks. Firms and investors can then prepare better for a range of possible future outcomes.