The ability to assess and manage climate-related financial risk has emerged as a critical priority for financial institutions. Climate stress tests must be precisely implemented to align with emerging global standards and local regulations, while also enhancing strategic risk management practices, as summarized in our “Overcoming the Challenges of Climate Risk Regulations - A Guide to Climate Stress Test Implementation” guide. Here, we describe the business benefits that derive from an integrated approach, mainly focusing on transition risk.
A robust solution for climate risk management can do more than fulfill compliance mandates: it can unlock a competitive advantage by enabling users to assess the impact of climate scenarios on financial instruments with exceptional detail. The ability to differentiate investment assets by their specific sensitivities to climate change can be leveraged to improve asset allocation and achieve sustainability objectives, such as Net-Zero 2050 commitments.
It is essential to deploy risk management software that allows banks and insurance companies to prepare for present and future requirements. Mapping climate scenarios to traditional risk factor stress testing is a useful first step, but it bears limitations in terms of accuracy, granularity and flexibility. Let us describe how these can be overcome by incorporating complex climate scenario-based projections into the core valuation and simulation engine.
Mapping Climate Scenarios to Risk Factor Shocks
Financial firms must run climate stress tests to fulfill quantitative regulations (e.g., Own Risk and Solvency Assessment reports). A simple methodology, where climate scenarios are mapped to existing market risk factors, has been widely adopted by early regulatory exercises such as the European Insurance and Occupational Pensions Authority (EIOPA) 2022 Institutions for Occupational Retirement Provision(IORP) climate stress tests. This simplified approach is likely to be insufficient for future requirements, including those based on the Network for Greening the Financial System (NGFS) framework[1], and is at present inadequate to drive business decisions because of the coarseness of the results obtained.
Typically, the shocks are differentiated, at most, by country and sector categories such as NACE (Nomenclature statistique des Activites economiques dans la Communaute Europeenne) Section or equivalent. For example, industries falling under "D - Electricity, Gas, Steam and Air Conditioning Supply" would all receive the same adjustment regardless of significant differences in their activities.
Figure 1: Comparing impact of granular calculation of NGFS scenarios (NDC, NZ50, Low Demand, Fragmented World, Delayed Transition, <2C) vs a "traditional" stress test (EIOPA 2022) in the electricity sector. While the traditional approach results in similar adjustments for all 4 subsectors, the NGFS scenarios, calculated here using “Algo Climate Scenarios for Transition Risk, powered by CLIMAFIN”, shows instead high variations. Under most policy trajectories, different subsectors take a mixture of positive and negative adjustments, as the impact depends on the specific activity and technology. The graphic has been generated with the Algo Workspace Analyzer
To increase the granularity, the sector classification can be refined down to 4-digit NACE codes. Section D of our example would allow us to distinguish between activities like the production and distribution of electricity—an illustrative example is depicted in Figure 1, where a traditional stress test is shown alongside the more refined classification. If one were to follow the risk factor mapping approach, however, this transformation would already imply an explosion of the number of risk factors to map equity shocks, creditworthiness and interest rate spreads to all the subsections.
The next step is to modify the classification and group firms into sets that share similar sensitivities to physical impacts and climate policy changes. NACE codes can, to that end, be mapped into granular Climate Policy Relevant Sectors (CPRS granular)[2] which distinguish across carbon emission profiles. NACE code D35.11 “Production of Electricity,” for instance, can be linked to "CPRS Fossil-Fuel Energy" and "CPRS Renewable Energy" in different proportions, depending on the production technologies. The resulting taxonomy is an excellent basis for propagating climate scenarios, as it can be conveniently connected with the Integrated Assessment Models (IAM) leveraged by NGFS scenarios.
Increasing Granularity from Sectors to Legal Entities
When it comes to improving investment portfolios, it is important to go even beyond detailed subsector and country classification to generate different climate adjustments for each (listed and unlisted) security. Climate models should, therefore, account for the specific characteristics of each Legal Entity (LE) and the instruments issued by it. Elaborating on the previous example of the Electricity sector, risk and investment managers should be able to differentiate the impact of mitigation policies on companies with different technologies and varying ability to convert them to a lower greenhouse gas emission profile.
For physical risk, the geolocation of the physical assets of each firm must be assembled and overlaid to local damage functions, quantifying the impact of different hazards on each region. A similar level of granularity can be obtained for transition risk assessment.
Many firms operate in more than one sector, so a sophisticated model must consider a weighted vector of NACE codes per LE, and map it to one or more CPRS. Additional LE-level characteristics should also be leveraged to refine climate impacts, including credit risk indicators (PD and LGD), dividend rate, return on capital, debt-to-capital ratio, stranding coefficients and technology mix of the firm.
Through the NGFS framework, sector-specific and LE-specific trajectories can be generated for variables such as energy demand and carbon prices, which in turn can be propagated to shock creditworthiness, spread and equity values. The specific model applied to produce these figures is very important, as described in the "Overcoming the Challenges of Climate Risk Regulations: A Guide to Climate Stress Test Implementation" whitepaper, and discussed in a future publication.
Benefits of an Integrated Approach
Such a granular level of calculation can hardly be sustained within a traditional market risk factors stress testing framework. In addition to the risk factors that one needs to support the sector mapping, multiple additional ones would have to be defined for every LE, with a dimensional explosion of model parameters and hard functional limitations. In short, the simplified approach is inadequate to obtain the desired precision.
The good news is that this problem can be solved by focusing on the simulation and pricing stage, where instrument-level details can be utilized without losing flexibility. The process is analogous to including betas and idiosyncratic risk when projecting a portfolio of financial securities based on a limited number of market risk factors. When the climate-contingent price adjustment is performed within a powerful risk management calculation engine, market shocks and time evolution can be applied in real-time.
SS&C Algorithmics and CLIMAFIN have developed such an integrated solution, whose benefits include the following:
Clients adopting Algorithmics as their primary solution for market (and credit) risk management benefit from an unparalleled level of accuracy in their climate risk assessments. The software incorporates climate scenario libraries powered by CLIMAFIN, enabling climate-impacted adjustments at LE and instrument level, with consistent projections to any future time point. The solution comes with a set of databases to automate data sourcing and facilitate the mapping.
To learn more, please contact us.
[1] https://www.ngfs.net/en
[2] Battiston, S., Monasterolo, I., van Ruijven, B., & Krey, V. (2022). The NACE – CPRS – IAM mapping: A tool to support climate risk analysis of financial portfolio using NGFS scenarios.