For insurers and financial institutions, incorporating climate risks in stress testing will not only be a regulatory expectation in the short-term but will also be a long-term endeavour – so it’s important to get started now.
Share this article
Insurance companies understand probability better than anyone. Every day, their organisations make strategic decisions that are informed by the likelihood and severity of future events.
This might explain why more and more insurers in Australia have begun factoring a broader range of climate-related risks into their stress testing. While they don’t know exactly what the future holds (details of local regulation is not yet confirmed), Australia’s insurers do know that the Australian Prudential Regulation Authority (APRA) is already assessing how banks analyse climate risk. Furthermore, Treasury has released its consultation on Climate Related Disclosures, which will likely be effective from 1 July 2024, and will require disclosure of the impact of climate risks and opportunities, including quantitative climate scenarios to be rolled out in the next 3 years. Local insurers also know that overseas regulators are now including insurance companies in their climate testing exercises.
We’re aware that overseas insurers are already busy trying to navigate their new regulatory requirements. In Australia, local banks are working to understand and adapt to APRA’s assessments. Informed by this, we’re also guiding insurers in Australia on how to prepare for the likely regulation of tomorrow. Because while we don’t know the finer details of what’s coming, we can anticipate the probable direction it is heading based on international experience.
In its 2021 Climate Vulnerability Assessment paper, APRA said that stress tests “provide forward looking analysis and insight on the resilience of entities, industries and the financial system to current and emerging risks.” It went on to say that: “Climate stress tests, much like traditional prudential stress tests, seek to determine the size of, and assist in the preparation for, potential financial system risks.” Furthermore, the ISSB as part of the IFRS S2 Exposure Draft has also stated that an entity should use climate-related scenario analysis to assess its climate resilience.
Given the potentially material and systemic nature of climate-related financial risk in Australia, as we've seen last year with the record breaking NSW and South East Queensland Floods which became the most costly extreme weather event in Australian history, the value of climate-risk stress testing cannot be overstated. When done well, climate stress testing helps boards and management to understand emerging risks. It involves modelling the impact that physical and transition risks will have under different climate scenarios on the balance sheet and on financial risks relating to claims costs and an insurer’s asset portfolio.
In 2021-22, APRA conducted its inaugural assessment of the climate risks faced by Australia’s five largest banks, with plans to expand future assessments to include insurance companies and superannuation funds.
Informed by APRA’s findings, and the work done by overseas insurers who are already being assessed by their regulators, we’ve identified some proactive steps that Australia’s insurers can take now. By focusing on these, organisations will set themselves on the path towards robust climate stress testing.
Insurers require reliable data to find a relationship between climate-risk factors and financial impact (for example, past earthquakes, bushfires or floods linked to claims costs, profit impacts, capital requirements and reinsurance recoveries and cost). The sooner you start collecting and linking such data, the sooner you’ll have a baseline to work from.
However, insurers need to focus on sourcing good quality data at the right level of granularity as this will be critical to deriving rich insights, drawing the correct conclusions and accurate disclosures.
A good place to start is to identify possible climate-risk factors and consider how data can be collected. Past, present and forecast data will be required from internal and external sources, so it’s important to clarify what you do and don’t currently have access to – and to check the legality of collecting each type of data (for example, can information not directly linked to insurance risk and pricing be captured and retained?). Potential sources of data could include reinsurers (who can provide market aggregate data), government agencies, existing large data providers as well as new entrants with a specific focus on ESG data (which can provide proxy data).
Technology will also require an uplift. Insurers require technology that can provide both ‘big picture’ and ‘granular’ modelling and calculations. For example, if you’re insuring properties, how far is each site from the sea? What is the likelihood of the ocean rising above a certain level and damaging the property? To answer such questions, your tech may need to combine in-house data with satellite images, AI, and other external sources.
Additionally, data which can serve multiple purposes and deliver business value should be prioritised. Combining data enrichment activities with existing business and technology initiatives can speed-up the data collection and business buy-in process, while at the same time benefiting other parts of the organisation such as pricing, product development, risk management,customer experience, and financial reporting.
Insurers need to be able to construct, validate and interpret scenarios. While climate stress testing is not a new phenomenon to many insurers, including broader climate risks and transition risks is new. Most insurers therefore have transferable skills and capabilities that are relevant for these developing requirements. For example, existing actuarial teams already perform stress testing for capital purposes. Insurers also excel at perils modelling, a practice that can gauge the probability and severity of natural disasters and factor that into pricing.
Current inhouse expertise will need to be complemented with additional knowledge, such as specialised climate scientists, to provide sophisticated projections of climate scenarios. The actuarial team can then link that to capital or reserving calculations.
A process of upskilling will also be necessary. For example, boards need to be able to interpret climate stress testing so they can translate that into strategy and management action (discussed below), as well as understand how these stress tests may interact with broader climate and financial reporting assumptions.
Upskilling can begin immediately. In the short term, external specialists might be engaged to help train up teams, build in-house capabilities as well as help form scenarios. With the rapid regulatory developments in Europe, overseas talent might be recruited to share their experience of adapting to insurance regulation in other jurisdictions.
Insurance leaders will need to incorporate climate risk and stress testing results into their strategy and action plans. In doing so, they must balance traditional requirements (e.g. shareholder value, growth, managing volatility) with emerging requirements (e.g. purpose beyond profit, new regulation, evolving community expectations, social license).
Financial modelling can reveal the anticipated returns for different levels of natural peril risks. This then informs decisions about capital allocation between different portfolios, as certain insurance products will be more ‘peril exposed’ than others (e.g. motorcars and homes that can be damaged from natural perils; whereas products like professional indemnity of directors carries no perils exposure).
Stress testing results allow management to compare the short-term and long-term financial impacts (e.g. captive positions, balance sheet positions) and assess their exposure to climate related risks. Leaders can then make informed decisions about risk appetite and risk tolerance levels. And those decisions will determine any action/inaction to control risk exposure and/or to mitigate any potential impact from those climate related risks.
It’s vital that the outputs and learnings from climate-risk stress testing are widely communicated (and understood) across the organisation. Rather than operating in silos, insurers need to coordinate efforts across multiple teams to maximise the value of their work.
For example, information gleaned from perils modelling and reinsurance catastrophe modelling can assist underwriters and the pricing team to better manage climate risk. In addition, insurers can utilise learnings from climate-risk stress tests and scenario analysis and take a more proactive approach in risk-based pricing that's based on future forecasts in frequency and severity of disasters rather than the current approach which is based on past disasters. This can help answer crucial questions, such as
We believe that climate stress testing can contribute hugely to the way Australia’s insurers approach risk and growth opportunities. Over time, it can play a bigger and bigger part in how insurance and financial institutions calculate the probability of future events, and the potential impacts of these.
A practical first step would be to work from the findings highlighted in APRA’s Climate Vulnerability Assessment (CVA) report. In this report, APRA found that the banks need to develop new tools, techniques and data sources to conduct climate risk analysis. Building on this, insurers can make a start by considering these findings and taking the following steps:
That will give investors confidence that you’re taking measured, progressive steps towards managing climate risk and capturing future growth opportunities. Climate-risk stress testing is a journey not a destination. The sooner you put a framework in place, the sooner you can identify any deficiencies and begin the process of continual improvement. Starting now rather than waiting for regulatory requirements and clear guidance from regulators means that not only do you take a leading role, you will also find it easier to comply with requirements once they do come into place.