Sélectionner une page


– février 2024

Following the past financial crises, stress testing and scenario analysis in recent years has become an increasingly important part of risk management, for banks and other institutions, within the balance sheets of an individual organization. In a rapidly changing environment, climate scenario analysis evolved from being simply a ‘nice to have’ tool (or recommended as best practice) to being increasingly seen by regulators as a practice to be developed, improved, and to be rendered mandatory.


Stress test and climate scenario analysis

According to the TCFD’s recommendations (Task Force on Climate-Related Financial Disclosures), scenario analysis is a complementary tool to critical strategic thinking, which, through the combination ofseveral possible characteristics, challenges the traditional knowledge of prospective events, proposing more risk-inclusive alternatives.

The characteristics of a valuable scenario analysis are:

1. distinctive; each scenario must be well delimited and different from others,

2. plausible, the features must be possible and with a straightforward narrative,

3. consistent, each scenario must have a strong internal logical rigor,

4. relevant, each scenario must clearly bring a concrete contribution to the future strategic and/ or financial implications of climate-related risks and opportunities.

NGFS — The Network for Greening the Financial System

Often, the scenarios are climate-based and designed by The Network for Greening the Financial System (NGFS). Indeed, the NGFS released a framework for scenario analysis to Central Banks, Supervisors, and Policy Makers, in April 2019, to facilitate the success of the Paris Agreement and the analysis of the physical and transitional risks faced by financial systems.

The list of climate scenarios was published in late June 2020 and revised a year later within the “NGFS Climate Scenarios for central banks and supervisors” report. The development of a set of climate scenarios was created in cooperation with an academic consortium comprised of the Potsdam Institute for Climate Impact Research (PIK), the International Institute for Applied Systems Analysis (IIASA), the University of Maryland (UMD), Climate Analytics (CA) and the

Swiss Federal Institute of Technology in Zurich (ETHZ).

The framework includes several possible scenarios. 3 of the main ones include the achievement of climate targets (met / not met) as variables as well as the transition path used (orderly/disorderly). (See below figure from NGFS).

Orderly: Early and ambitious action for a net zero-carbon economy. In this scenario, it is assumed that climate policies are introduced early and gradually become more stringent. Zero net CO2 emissions are achieved before 2070, giving a 67% probability of limiting global warming to less than 2°C. The physical and transitional risks are both relatively low.

Disorderly: Late and/or unplanned action. In this case, climate policies are not introduced until 2030. As actions are taken relatively late and are limited by available technologies, the result is a higher transitional risk.

Hot-house world: limited action with significant global warming and a sharp increase in exposure to physical risks. These conditions assume that the policies used today will remain unchanged in the coming years, causing emissions to rise continuously, seeing an increase in global temperatures of at least +3°C in 2080 as well as severe physical risks which would include irreversible changes such as sea level rise.

Finally, in the top right quadrant of the image below, there is a category called ‘Too little, too late with high physical and transition risks. This category is not included among the scenarios deemed possible by the NGFS because this quadrant is currently less studied and requires the development of additional scenarios.

It is well known by now how climate risks, if not acted upon in a timely manner, can translate into financial risks (operational, credit, liquidity…) and evolve into systemic risks. Many central banks have implemented this tool using the scenarios provided by the NGFS, sometimes with additional risks and factors, in order to create a total ad hoc model that fits the case. As examples, you will find below the cases of Great Britain and Europe.

Great Britain

In Great Britain, a first attempt to include climate in stress tests was successfully done already in 2019, when climate criteria were used to analyze insurance risks. Today, the Bank of England (BoE) is reintroducing a similar model, with the launch of a
climate-related stress test program in 2021, called the Climate Biennial Exploratory Scenario (CBES for short).

The CBES gives precise guidance to the financial participating institutions, specifically detailing the variables to be used and how to run the model. In any case, the applicable possible scenarios to explore the two key risks of climate change, the transition risks and the physical risks, will be an early policy action scenario, a late policy action scenario, and a no additional policy scenario. Each scenario assumes a timeframe of 30 years.

The desired 2021 outcomes for the CBES were to:

- Assess the financial exposures of individual participants and the financial system in general to climate-related risks.

- Substantiate the participants’ business model challenges from these risks, and gauge their likely responses as well as the implications of financial services provisions.

- Assist participants in enhancing their management of climate-related financial risks. This includes engaging counterparties to understand their vulnerability to climate change.

The early findings of this exercise have confirmed what was previously assumed: climate change, and the transition to net zero, not only will have a strong impact on business and households — no matter the adopted scenario but more importantly, the under-
estimation of climate risks may lead to systemic risks, strongly impacting the profitability of banks and insurers. In any case, projections of expected losses are still uncertain and difficult to quantify due to large data gaps. This is where the BoE will focus in the upcoming CBES, in order to foster improvements in this area.


In Europe, however, the situation seems to still be developing.

The ECB Banking Supervision launched the first climate risk stress test study plan in January 2022.

The goal was to test how prepared Europe’s leading banks were for an economic shock caused by a list of expected and unexpected climate changes. The stress test serves both banks and supervisors as a learning process and a tool to develop climate-related policies, without having any direct impact on capital requirements.

This exercise was carried out in the first half of 2022 and at the end of July and the results were finally published in the “2022 climate risk stress test” report, prepared by the European Central Bank Banking Supervisor.

This latest report demonstrates that some progress has been made with respect to 2020, but it is not yet satisfactory, given the results of the European Central Bank’s first climate stress test, which warns that for some large institutions, there is still a lot to be done.

For example, out of 104 European banks that were considered, 60% still do not have climate risk stress test models; furthermore, only one bank out of four (i.e. about 20%) considers climate as a variable risk in lending. An alarming finding of the study was that if a disorderly transition were to occur over a short-term period of three years, the 41 largest banks would see combined credit and market risk losses of EUR 70 billion.

Further to this announcement, the ECB expects banks to make substantial progress in the coming years. In order to help banks on this path, the ECB will carry out follow-up activities to provide suggestions on how to address this process. This work is expected to be completed by the end of 2022.

Challenges for the years to come

In conclusion, we can certainly state that in recent years, much has been done in favor of the emerging climate-related risk policies. The first results are slowly being achieved, as can be seen from the attention the topic has received and the number of financial and non-financial institutions engaged in it. In any case, there is still a long way to go, and much remains to be done.

In this regard, a persistent problem in the development of models and policies is the lack of data and their quality. Central banks and the supervisory authorities must encourage all stakeholders to develop data extraction and modeling techniques that will allow for consistent testing to understand how to tackle climate risks.

World-leading banks must be prepared to include climate stress tests within their risk management systems. Increasingly, large financial institutions are relying on experienced consulting firms that can gather the necessary data for the creation of climate-related risk assessment models, both for internal management purposes and in anticipation of future policy developments.

By Alberto SALICE, Senior Consultant at Square Management

Share This