Risk Management

Climate Scenario Analysis: Stress Testing the Future

The systemic risks of climate change are becoming more apparent to regulators. As a result, they are increasingly looking to stress testing and scenario analysis to ward off potential ‘green swan’ events. The learning curve for firms and supervisors is steep, but collaboration between them may be the path to addressing the biggest risk we face today.

Thursday, July 2, 2020

By Jo Paisley and Maxine Nelson

Imagine you are the CRO of a bank. But it is 1980, and you’re being asked to undertake scenario analysis, looking at how you expect your business will evolve over the next 30 years. Would you have any inkling about the growth of personal computers, the Internet, big data or even mobile devices? You might be hard pressed to imagine the growth of social media, and the ways in which it would connect people across the globe, creating platforms for consumers to share views on your firm, changing market dynamics and the sizable impact on firms’ reputations.

However, the same is being asked of banks today. Regulators are increasingly demanding that banks perform scenario analysis over a similar time horizon. Yet today the focus is on how climate change might impact their balance sheets and business models.

The UK’s Prudential Regulation Authority (PRA), for example, has released a consultation about testing the resilience of the largest UK banks and insurers to the risks associated with different climate scenarios in its next Biennial Exploratory Scenario (BES) exercise – now scheduled for 2021.  In April 2019 the PRA’s Supervisory Statement set out the regulator’s expectations for how firms should manage the financial risks from climate change.

For Regulators, Systemic Risks Loom Large

Why are regulators becoming so concerned about the risks from climate change? In part, it is because it is expected to have systemic impacts that are far-reaching in breadth and magnitude, affecting households, businesses and governments, across all sectors and geographies. But it’s also because the past is unlikely to be a good guide to the future.

Climate change is expected to feed through to the economy via two channels – physical risk and transition risks – which are both characterized by deep uncertainty and non-linearity.

Physical risks will arise through ‘acute’ weather-related events, such as heatwaves, floods, wildfires and storms, as well as ‘chronic’ or longer-term shifts in climate patterns, such as changes in precipitation and extreme weather variability, rising sea levels and increasing mean temperature. Transition risks arise from the process of adjusting toward a lower-carbon economy and cover the impacts of policy and legal changes, technological changes, and shifting sentiment and societal preferences.

Despite the high degree of uncertainty, the risks are foreseeable, even though the precise outcomes, time horizon and future pathways are unclear.  This uncertainty elevates scenario analysis to be a critical new tool that firms will need to use.  The good news is that scenario analysis – if done well – can both improve risk management and foster better strategic conversations at the board level.

Scenario Analysis Trends: Usage, Types and Results

In GARP’s Second Annual Survey of Climate Risk Management, we found that more financial firms are beginning to undertake this type of analysis. Roughly 60 percent of the 71 firms in our sample now use scenario analysis – though most employ it on an ad hoc basis, rather than as a part of regular risk assessment.

Firms that employ best practices use a range of scenarios, as the outcomes from them will differ significantly. For example, scenarios involving higher greenhouse gas emissions pathways will have more global warming and more physical risks than those scenarios with lower emissions pathways.  On the other hand, scenarios in which emissions are kept lower (for example, through policy changes) will tend to have lower physical risks, but more transition risks.

As Figures 1 indicates, firms are using multiple scenarios, with more than 20 percent of firms using either two or four.


Figure 1: Number of Scenarios Used, Per Firm

The most popular scenario used is the Representative Concentration Pathway 8.5 (RCP 8.5), published by the Intergovernmental Panel on Climate Change (IPCC), representing a high-emission pathway leading to higher physical risk and lower transition risk.  The next most popular scenarios are aligned with the Paris Agreement, namely the International Economic Agency’s Sustainable Development Scenario (‘IEA SDS’) and the IPCC’s RCP 2.6.  In these scenarios, global temperature increases are kept below 2⁰C (degrees Celsius) above pre-industrial levels, with CO2 annual global emissions decreasing over time to be near, or below, zero by the year 2100.


Figure 2: Scenarios Used by Financial Firms

It is encouraging to see that more than half of the firms that have undertaken scenario analysis have acted as a result of it. Many have improved disclosures, with some changing portfolio composition, strategy, lending practices, underwriting and products or services.  Moreover, most of the firms currently not using scenario analysis plan to use it within the next two years.


Disclosures and Stress Tests - Opportunities and Challenges

Scenario analysis plays a prominent role in disclosure frameworks, such as that established by the Task-Force on Climate-Related Financial Disclosures.  Regulators are also increasingly focused on running stress tests for the firms that they supervise.

Is there a danger, however, that meeting these growing external demands might undermine the use of scenario analysis for internal risk management?  Will the prospect of needing to publish these analyses tempt firms to present their results in as positive a way as possible?  Without standardized scenarios, firms are likely to be concerned that they will choose overly-harsh scenarios, leaving themselves more vulnerable than their competitors.

Some good news is that supervisors are designing more standardized scenarios.  The Network for Greening the Financial System (NGFS), which now comprises 66 central banks and regulators, recently published a set of global ‘reference’ scenarios and a guide for supervisors, providing practical advice on how scenario analysis may be used to assess climate risks to the economy and financial system. That is helpful, as it should provide a coordinating framework for subsequent supervisory exercises.

Although the NGFS isn’t a standard setter, it is a ‘coalition of the willing’ that is sufficiently concerned about the risks from climate change that they are currently investigating how micro- and macroprudential supervision might need to change to protect firms and the financial system.

However, supervisory stress tests do bring their own challenges that risk managers need to guard against. First, the tests carry the  risk of becoming a beauty parade, where  “banks want to look ‘the prettiest’ by trying to anticipate the opinion of supervisors and market participants.

Second, supervisors tend to develop their own regimes,  potentially leading to a patchwork of overlapping, disparate tests that cost a lot of money for both supervisors and the firms.[1]  We are, in fact, already seeing some of this behavior: for example, shortly before the NGFS published its guide and scenarios, the ECB and the Monetary Authority of Singapore both launched  public consultations on how they expect banks to safely and prudently manage climate-related and environmental risks.

Third, risk managers will need to consider whether the standardized scenarios properly capture the specific risks in their portfolios; these standardized or ‘reference’ scenarios may aid supervisors in assessing the risks to the financial system in a consistent manner – but they will not necessarily help a CRO understand the risks facing their particular firm.

[1] For more background on some of these challenges, please see the GARP Risk Institute’s Code of Practice for supervisory stress tests, aiming to inspire some degree of coordination and harmonization across jurisdictions.


Still Learning

Supervisors and firms are also cognizant that they are on a steep learning curve.  The formation of the UK’s Climate Financial Risk Forum (CFRF), under the auspices of the PRA and the Financial Conduct Authority (FCA), is an example of an innovative, inclusive approach to learning.  Comprising five banks, five asset managers and five insurers, the CFRF aims to build capacity and share best practices across financial regulators and industry to advance financial sector responses to the financial risks from climate change.

The Forum has published a guide summary, with chapters on risk managementscenario analysisdisclosure, and innovation, providing a range of practical insights for financial firms to help them better understand the risks and opportunities that arise from climate change. Moreover, they offer advice on how to integrate these risks and opportunities into risk, strategy and decision-making processes.

As Secretariat to the CFRF’s Risk Management and Scenario Analysis working groups, we saw first-hand the advantages of collaboration between industry and regulators. In its BES consultation paper, the PRA rightfully noted that measuring the financial risk from climate change is a “complex” challenge that requires innovative ideas and teamwork. “It involves assessing the effect of multiple climate pathways, with different physical and transition effects, over several decades. This requires new tools and approaches to measure and understand the risks,” the regulator elaborated.

The paradigm shift we are now witnessing is noted in a recent Banque de France/BIS book, aptly named ‘The Green Swan.’ So-called ‘green swan’ events are essentially climate-related tail risk events, with potentially catastrophic consequences and involving even greater complexity than black swan events.

Climate financial risk management, and scenario analysis in particular, is still in a relatively early stage. But we can expedite the process by working together and learning from one another.

BylawsCode of ConductPrivacy NoticeTerms of Use © 2024 Global Association of Risk Professionals