Published Date: 20 October 2022

Keynote Speech by Mr Daniel Wang, Executive Director, Monetary Authority of Singapore, at the Bank of England and Prudential Regulation Authority Climate and Capital Conference on 20 October 2022

Distinguished guests,

Ladies and gentlemen,

1. Good morning. I would like to thank the Bank of England (BOE) and the Prudential Regulation Authority for your kind invitation to speak at your inaugural climate and capital conference. It is a privilege and pleasure to join my fellow regulators, academia, and members of the financial services industry here in London, as well as those of you attending virtually.

2. Being a research conference, we are here to learn, debate, challenge and importantly, to listen. In this regard, I enjoyed and benefited much from the first day of the conference. It is a most unenviable task to speak following yesterday’s insightful keynote speeches from Deputy Governor Woods and Professor Lord Stern. That said, I intend to contribute perspectives on environmental risks and climate risk assessment through the lens of an insurance regulator, coming from a part of the world that is deeply grappling with the effects of climate change.

A growing climate crisis

3. Open a newspaper today, and one can be forgiven for thinking the Horsemen of the Apocalypse are upon us. Pestilence and disease, war and conflict, food and energy supply disruptions, geopolitical risks and inflation are but some of the pressing conjunctural challenges the financial sector, central banks and regulators have to deal with. But the greatest of all these challenges has already come – we are already dealing with a growing climate crisis.

a) According to the World Meteorological Organization, the number of natural disasters worldwide has increased fivefold over the last fifty years. This amounts to more than 11,000 reported disasters worldwide since 1970, US$ 3.64 trillion in losses, and over 2 million deaths.

b) Climate change has brought about extreme events including severe tropical cyclones, which have in turn raised the intensity of other events such as flooding and associated impacts.

i. Low-lying megacities, deltas, coasts, and islands in many parts of the world are now more vulnerable. Just this year alone, floods triggered by torrential monsoon rains devastated Pakistan, while floodwaters also rose in Malaysia, Indonesia, and Thailand.

ii. Research has found clear human influence on extreme rainfall events, sometimes in conjunction with major climate influences like the El Nino-Southern Oscillation. There has been record rainfall in Japan and India in 2022. At the other end of the spectrum, we read about droughts in China and rising desertification in Mongolia.

4. It is evident that while climate change is a global phenomenon, its effects are local, with very real impacts on people and communities.

5. In the insurance sector, where risk management is the focal objective of the industry, the sophisticated management of natural catastrophe and climate risks has been a decades-long endeavour. Nonetheless, the gradual evolution of climate change brings about heterogeneous and highly uncertain outcomes when considered over a long-term horizon. Dealing with multi-decade effects of global climate change, while needing to perform effective projections and make real-time investment, underwriting and pricing decisions amidst data scarcity and technological limitations represent a whole new and altogether more complex challenge.

6. In my address today, I will explore how climate scenario analysis can aid regulators and the financial services industry make effective climate risk assessments.

Scenario-based climate risk assessments

7. From the insurance industry’s perspective, climate change is not typically viewed as a distinct risk. Rather, it comprises a combination of various and diverse risk drivers, which may result in increased interconnectedness of different risks. In addition, it is often seen as a change factor affecting the frequency and severity of other risks to an extent which may not have been seen in the past few decades, while impacting both sides of an insurer’s balance sheet. The still evolving, multi-faceted nature of climate risk is also what makes this particular risk so uncertain and challenging to assess and manage.

8. Regulators and insurers need to consider how to assess and address issues within risk management and governance, valuation of assets and liabilities, and conduct of business, in light of such uncertainty. As such, the analysis of climate risks will benefit from a forward-looking view and a combination of qualitative and quantitative approaches.

9. Scenario-based climate risk assessments can help regulators and insurers better analyse the impact of environmental and climate change risks, by allowing the testing of various narratives and parameters. When complemented with quantitative assessments, these help with exploring the effects of transition risk on insurers’ investment portfolios, while facilitating discussions on risk mitigation actions. Physical risks can also be measured through this tool, such as assessing the impact on insurers’ real-estate and infrastructure holdings in a flood or typhoon-prone area stemming from higher frequency and intensity of these natural disasters.

10. Meanwhile, qualitative scenario assessments permit the user to discern and consider changes in the socio-economic environment under different climate pathways, facilitating a better appreciation of inter-relationships, dependencies, and feedback loops of the key drivers of climate risks. An insurer conducting such assessments can glean climate change-related insights from supposedly second-order variables like population migration, and geopolitical conflict.

11. For insurance regulators, the use of quantitative and qualitative assessments helps us better understand the financial stability implications of climate change, which in turn informs reviews of our prudential toolkit and supervisory interventions at the firm level.

12. For insurers, other financial institutions and their counterparties, scenario analysis can provide insights on how their business models may need to evolve alongside a transition to net zero. As mentioned, scenario building and evaluation can reveal key change agents and feedback loops, supporting early strategic perspectives on the positioning of an insurance portfolio and related investment decisions.

13. Findings from climate scenario analysis can then be used to formulate credible and feasible transition plans. These analyses can contribute to companies’ strategy-related disclosures, in line with recommendations by the Task Force on Climate Related Financial Disclosures (TCFD) as well as proposed disclosure requirements by the International Sustainability Standards Board (ISSB).

14. I have highlighted the merits and use cases of climate scenario analyses. I must first acknowledge that the work is not easy and drawing firm conclusions from a single exercise is not only difficult but imprudent. Indeed, quantifying the impacts of climate change over the long term is challenging given the inherent uncertainty of key drivers both internally and externally.

Challenges in Climate Risk Assessments

15. There are three key challenges – data paucity, technical gaps, and operational difficulties.

Data Paucity

16. The first issue is data, specifically the lack thereof. More work is required by regulators, insurers, and the broader ecosystem to address data gaps needed for robust climate risk assessments.

a) As we are aware, climate risk manifests primarily through transition and physical risks, and increasingly via litigation risks. Assessing transition risk would require sectoral classification of insurers’ assets. With this data, we would be able to project market and credit risks by applying equity shocks and credit spreads to each sector class. However, this is not enough as each individual asset counterparty would experience a different level of transition risk based on its own intended net zero transition pathway. Individual counterparty assessment would require extensive data in the form of the counterparty’s own carbon emissions and from upstream and downstream stakeholders. The availability and accessibility of such data is substantially lacking at this juncture.

b) To assess physical risks, insurers will need granular exposure location information. Property and casualty (P&C) insurers possess such data, being in the business of insuring property assets from natural disasters and other perils. However, the manifestation of physical risks may happen over areas involving data that insurers currently may not collect or closely monitor. Examples include physical risk parameters like changes in soil moisture and the land area exposed to crop failure.

Technical gaps

17. The next challenge is about technical gaps. While there are many climate models to choose from, it is not always apparent which ones are most relevant. Indeed, climate modelling is a highly technical field and can be difficult to understand and penetrate.

a) An example is the gap between current scientific work and definitions versus how insurers consider natural catastrophe risks.

b) Assessment of projections of physical risk models and their calibration require specific scientific expertise. History provides little guidance and extreme events cannot be ruled out over the near to medium term as the distribution of such events is likely shifting over time, with nonlinear threshold effects and tipping points of uncertain timing and possibly systemic impact.

c) This is complicated further by the lack of a clear methodological framework for translating climate scenarios into macro-financial analysis.

18. This is linked to what I said earlier about data. Given the technical challenges inherent to climate risk modelling, the type, range, and granularity of data needed will change, adding yet another dimension of uncertainty.

19. In addition, there is clear scope for climate modelling to be enhanced.

a) Macroeconomic models typically used by central banks have certain limitations that make them less suited to studying climate risks. For instance, they are often used to assess short-run divergences from long run equilibria rather than investigate structural changes in the economySource: (2020) NGFS Guide to climate scenario analysis.

b) In the industry, P&C insurers were able to withstand losses from several major perilsSuch as Hurricanes Katrina, Rita, and Wilma in 2005, the series of tropical cyclones in 2017, 2018 and 2019 in the U.S. and Japan , which illustrate successful risk management practices on the back of probabilistic NatCat modelling and quantitative exposure monitoring. However, the models useful for today's and next year’s portfolio management do not lend themselves well to making projections over multiple decades. They are limited in providing decision-useful quantitative information over the longer time horizon needed to assess the exposures and vulnerabilities associated with climate change.

c) Meanwhile, each stakeholder can differ in degrees of sophistication that they exhibit in modelling. This can bring about inconsistent interpretation of scenarios and the assessment of risks.

Operational challenges

20. The third challenge is operational in nature.

a) Substantive resources are required to conduct a climate risk assessment. Effort and expertise are required to gather data, create customised models, run them, and most importantly, understand and review the results. As a regulator, I am cognisant of the amount of resources required. In the case of Singapore, the Monetary Authority of Singapore (MAS) gave our insurance industry more than 12 months to get ready and conduct a bottom-up quantitative assessment of climate risks under a range of scenarios.

b) A long time horizon introduces uncertainty, and significant effort is required to adjust existing models to perform projections over such time horizons. Some models are also not able to utilise all the parameters provided by the regulator. For the MAS, we emphasized to the insurers that our scenario analysis exercise is meant to facilitate learning for both MAS and the industry, and forms part of a multi-year, iterative journey to build capabilities in climate risk assessment.

c) Some insurers also provided feedback that the parameters provided by regulators for climate scenario analysis exercises were not sufficient for their risk assessment purposes. To address this, these insurers have had to undertake additional scenario expansion efforts, either through their vendors or through proprietary modelling frameworks, to generate additional parameters for their use.

What we can do together

21. Regulators do not view scenario analysis through rose-tinted lenses. While it is an essential tool for making climate risk assessments, more work and development is needed given the uncertainties associated with climate change. This is also why climate risk assessment and scenario analysis needs to be an iterative learning process. The initial results will identify new pockets of risks and key sensitivities of the scenario that were not initially included. These aspects can then form the basis for follow-up analysis and research.

22. The data shortage issue is being addressed at various platforms, including the work by TCFD and ISSB. The International Association of Insurance Supervisors (IAIS) has been collecting climate risk data from its members for several years through its Global Monitoring Exercise. Data depth and familiarity will continue to grow as regulators and the financial services industry refine our parameters and definitions. Nonetheless, close cooperation and understanding between the public sector and real economy players are necessary to ensure all of us can obtain sufficient decision-useful data to assess the risks arising from climate change.

23. Technical gaps are progressively being addressed. The Geneva Association has established a ‘Task Force on Climate Change Risk Assessment for the Insurance Industry’, involving global P&C and life reinsurance and insurance companies. The Taskforce aims to develop methodologies and tools for scenario analysis and stress testing for the insurance industry. A key part of the process will involve collaborating with different stakeholders to converge on decision-useful solutions. I anticipate there will be increased cross-group and cross-company engagement and collaboration to pool climate risk management expertise and enhance firm-wide understanding of these risks.

24. The Network for Greening the Financial System (NGFS) has continually developed and refined its climate scenarios to make them relevant to a broader audience. The third vintage of the NGFS climate scenarios was released last month, featuring updated input assumptions and modelling enhancements. In addition, the NGFS will publish a joint report with the Financial Stability Board next month on climate scenario analysis. The report will bring together a global perspective from the various national and regional climate scenario analysis exercises conducted thus far, and draw lessons for effective scenario analysis.

25. The IAIS Climate Risk Steering Group (CRSG) has been actively supporting insurance regulators in developing capabilities to assess climate risks. In the past year, we have conducted a series of workshops to build capacity on scenario analysis, attended by more than 200 insurance regulators. The workshops covered the basics like understanding the micro and macro prudential objectives of scenario analysis to more advanced discussions on technical design specifications. Special mention goes to colleagues at the BOE for leading this work within the CRSG. The CRSG will provide additional guidance on climate scenario analysis to IAIS members in the coming months.

26. Last but not least, there are many initiatives tapping on public-private and/or academic collaboration to develop robust approaches for climate risk assessment and management. In Asia, this includes the Global Asia Insurance Partnership (GAIP), a tripartite partnership between the insurance industry, regulators and policymakers, and academia to address the future development and needs of the insurance sector. Climate risk is a focus area of the GAIP. The GFANZ Asia-Pacific Network was just launched in June 2022 and will support engagement between financial institutions and policymakers across the Asia-Pacific to accelerate the region’s net-zero transition.


27. In summary, while much remains to be done, I am optimistic that the insurance and financial services industry, as well as my peer regulators, are on the right track in assessing and managing climate risks. Close collaboration between the regulatory community, industry and academics is vital. When it comes to scenario analysis and climate risk assessments, simply put, we are in the same boat and learning together. Only by rowing and working together can we make clear progress towards robust climate risk assessments. And I am confident the panel discussion and papers presented following this address will further contribute to our learning.

28. A remarkable lady, deeply respected for her inner strength and quiet leadership, once said – “It is worth remembering that it is often the small steps, not the giant leaps, that bring about the most lasting change.” Queen Elizabeth II was laid to rest last month. But her words will continue to echo, and I hope they inspire us to work together to bring about a sustainable future for our planet.

29. Thank you.