Step 3: Defining scenarios
A forward-looking, risk-based approach to the ORSA requires that risks identified are subjected to a wide range of stress tests or scenario analyses. This helps to assess the resilience and robustness of the business strategies under different developments of climate change risks over time and assists management in deciding on mitigating actions for excessive risks.
The first step in conducting a scenario analysis is to identify the range of relevant climate change scenarios including the main assumptions, the macroeconomic parameters, and the level of granularity at which the risks are assessed for each scenario. This step is vital in ensuring interesting and challenging scenarios that facilitate internal communication and discussion. The level of granularity applied will depend on the risk exposures identified and the modelling techniques available.
EIOPA specifies two long-term climate change risk scenarios that should be used as a minimum:
Global temperature increase remains below 2°C, preferably no more than 1.5°C, in line with Paris Agreement
Global temperature increase exceeds 2°C.
The scenarios will depend on the assessment of the materiality of climate change risk exposures and will differ from one insurer to another. Insurers can either consider publicly available scenarios or can develop the scenarios themselves or build on existing scenarios, which will require expertise and resources. There are numerous publicly available climate change scenarios that have been released, including publications by the Network for Greening the Financial System (NGFS), the Prudential Regulation Authority (PRA) in the UK and commercial data providers such as Twenty Four Seven.
Step 4: Modelling of risks
Modelling transition and physical impacts at granular level
When modelling the climate change risks, (re)insurers have to translate the transition and physical impacts on asset prices and underwriting activities, since the publicly available climate scenarios do not always provide this information at high resolutions:
Transition impacts on assets
Physical impacts on assets
Physical impacts on underwriting activities
- Several methodologies are available to estimate the impact of transition pathways on asset categories.
- These generally assign a carbon-sensitivity to economic sectors or activities, which are used to differentiate the impact of a specific transition pathway on the overall economy and asset returns to the different sectors.
- Physical risks can be difficult to translate to financial impacts on investments in companies - large companies often have activities in multiple locations and countries and extreme weather events / natural disasters may affect financial losses in complex ways due to supply chain effects.
- Scoring models are available for companies and real estate investment trusts, ranking their sensitivity to physical risks.
- Physical impacts need to be translated to changes in the frequency & severity of acute perils and chronic effects.
- These changes are then converted into financial impacts on insurer’s underwriting portfolio in the relevant geographical areas.
Practical example for assessing the impact of physical risks (floods) on a non-life insurance portfolio
The potential impact of climate change on a non-life insurer covering flood risks can be assessed by using the JRC Peseta IV study on rising river flood risk in the EU.
The study assessed the impact of projected changes in frequency and severity of river floods on the expected annual damage (EAD) under various climate scenarios (1.5°C, 2°C and 3°C warming) and future socioeconomic conditions (2050 and 2100 economy), which were then transformed into financial losses for the entire economy. The study summarises the EAD for all EU countries under present conditions and under different combinations of the future socioeconomic conditions and climate scenarios.
The study can be used to assess exposure to regions that are strongly impacted by rising river flood and to estimate the impact of climate change on the insured losses in its underwriting portfolio, based on the projected change of overall economic losses under the various scenarios.
Changes to estimated losses under various scenarios can be derived by considering the changes in EAD presented in the study. The average annual loss (AAL) for a country under a specific scenario for temperature and socioeconomic conditions can be estimated by multiplying the AAL by the ratio of the EAD for that scenario divided by the base EAD.
For a more precise approach, insurers can consider regional rather than national level information and apply the same method to obtain the projected changes in EAD at regional level.
Insurers need to decide on the projection period to consider in performing the scenario analysis. For a short-term risk analysis, the scenarios or stresses can be applied to the current balance sheet without the need to project the balance sheet. Simplification techniques can also be applied whereby transition and physical events occurring in the future are assessed against the current balance sheet.
Alternatively, a partial scenario analysis can be performed that preserves the long-term character of climate change scenarios without projecting the full balance sheet by projecting simple ratios for different perils or geographic areas.
There are however important advantages of performing a full balance sheet projection, such as ensuring internal consistency, enhancing insights about the sustainability of business models and strategy, and ensuring feasibility of management decisions intended to mitigate asymmetric balance sheet shocks. The goal of long-term scenario analysis is not to provide detailed projections of all financial components, but to evaluate the business strategy across a range of scenarios and material climate change risk drivers.