N° 122 : The main results of the 2020 climate pilot exercise

Its unprecedented and ambitious nature lies in the time horizon over which the risks were assessed (30 years), the methodologies used (analysis of scenarios broken down by economic sector), its innovative hypotheses (notably the dynamic balance sheet), its coverage of physical and transition risks, and the fact that the participating institutions directly assessed their risks on the basis of common hypotheses. It illustrates the leading role played by the French financial authorities and the Paris financial centre and the progress made in the fight against climate change since the adoption of the Law on Energy Transition and Green Growth and the signing of the Paris Agreement in 2015.

This exercise, conducted from July 2020 to April 2021, achieved its objectives of:

• Mobilising French banks and insurers: Virtually all the banks and insurers that contributed to the preparation of the exercise were present, as the exercise was carried out on a voluntary basis, and others joined them even though they had not taken part in the preparatory phase: in total, 9 banking groups and 15 insurance groups got involved over three quarters without accounting for the preparatory phase, despite the context of the Covid-19 crisis, to carry out this pilot exercise. This very high level of mobilisation enables us to present representative results (85% of the total balance sheet of banks and 75% of the total balance sheet of insurers) with high added value for these two sectors.

• Raising awareness about climate risks: despite the methodological difficulties and the absence of certain key data, the participating institutions welcomed the relevance of this pilot exercise and the many advances it led to in terms of cross-disciplinary team mobilisation, internal discussions on risk analysis and the limits of the models currently used, but also in terms of strategic orientation and a better understanding of the issues and the impact of climate change on business models. Financial institutions became aware of the fact that this type of exercise was not only feasible but also extremely useful for making headway in taking climate risk into consideration. The pilot exercise therefore served as a catalyst for spurring debate and, for some, sped up the mobilisation of teams and resources.

• Quantifying and assessing complex transition or physical risk scenarios, drawing in particular on the work of the NGFS, which is the network of central banks and supervisors for greening the financial sector. The ACPR, with the help of Banque de France staff, prepared this exercise in accordance with the NGFS guidelines on the construction of climate change scenarios and based on two of the scenarios published by the NGFS in June 2020. These scenarios will also serve as a basis for other exercises currently being prepared, such as those of the Bank of England in June 2021 and the European Central Bank in 2022. It is important that a growing number of supervisors take up this work in order to launch their own exercises and thus contribute to the development of a common base of knowledge and climate risk assessment.

• Providing a first measurement of risks and vulnerabilities to which French financial institutions are exposed: the pilot exercise thus usefully complements the ACPR's previous analyses, published in April 2019, which were based on questionnaires. In addition to this snapshot, the pilot exercise adds a forward-looking view of risks over a long-term horizon, conditional on the implementation of several alternative scenarios. The exercise thus offered financial institutions the possibility of assessing their corrective actions (e.g. exit from certain sectors), thanks to the dynamic balance sheet hypothesis, and of taking new risks into consideration: potential hiatus between strategies for exiting certain greenhouse gas emitting activities and the objective of maintaining market shares, the willingness to finance the economy or to preserve a client relationship, which could result in a more lasting exposure to transition or physical risks than expected.

The pilot exercise revealed an overall "moderate" exposure of French banks and insurers to climate risks. However, this conclusion must be put into perspective in view of the uncertainties concerning both the speed and the impact of climate change. It also crucially depends on the assumptions, the scenarios analysed and the methodological difficulties raised by the exercise. Based on the current balance sheet structures, it nevertheless appears that considerable efforts must be made to help significantly reduce greenhouse gas emissions by 2050 and to contain the rise in temperature by the end of the century.

• The pilot exercise exhibits, conditional on the retained scenarios and assumptions, an overall "moderate" exposure and vulnerabilities as highlighted in the ACPR's previous work. According to the projections of the Intergovernmental Panel on Climate Change (IPCC) used in this exercise, France, which accounts for about 50% of the exposures of French financial institutions, and Europe, which accounts for about 75% of exposures, are relatively less affected than other geographical areas. France also produces less than 2% of global greenhouse gas emissions1 . On the other hand, exposures to geographical areas such as the United States (which accounts for around 9% of exposures) appears to be sensitive to transition risk.

• The exposure of French institutions to the sectors most impacted by transition risk, as identified in this exercise (e.g. mining, coking and refining, oil, agriculture, construction, etc.), is relatively low. In addition, institutions tend to reduce their exposures to these sectors by 2050. However, these sectors post the highest increase in the cost of risk and probabilities of default. The cost of risk rises threefold in these sensitive sectors. By way of comparison, the Covid-19 crisis led in 2020 to a twofold increase in the cost of risk for French banks in a context of heavy business losses. The contribution of these sectors to the rise in the cost of risk (e.g. provision for expected losses) appears to be greater than their share of banks' balance sheet. Relative portfolio losses for banks and insurers are also concentrated in these sectors, albeit with significant dispersions depending on individual exposures. When interpreting these results, it should be borne in mind that none of the scenarios analysed implies an economic recession by 2050, contrary to the usual practice of stress tests, but, for the adverse scenarios, a lower trend in activity. In this context, the increase in the cost of risk indicates that the energy transition, which is necessary in order to comply with the Paris Agreement, requires significant efforts to adjust the system and economic structures.

• Even though France is relatively spared in the Intergovernmental Panel on Climate Change (IPCC) scenarios, the pilot exercise shows that the vulnerabilities associated with physical risk are far from negligible. Thus, on the basis of the information provided by insurers, the cost of claims could rise by a factor of 5 to 6 in certain French departments between 2020 and 2050. The main hazards contributing to this increase in claims are related to the risk of "drought" on the one hand and that of "flooding" on the other, as well as to the rise in the risk of cyclonic storms in the overseas territories (French Caribbean). This increase in claims highlights an insurability risk in certain parts of the country, a risk, which insurers felt could be fully offset by an increase in contributions. As regards banks, the exercise shows that the progress made in taking physical risk into consideration is very limited in view of the findings of the ACPR in 2019 which already underlined that the assessment of this risk was lagging significantly transition risk analysis. Only two institutions were able to quantify the impact of an increase in the lack of insurance coverage on its credit risk parameters. This situation is primarily linked to the difficulty encountered by institutions in obtaining a precise view of the geographical location of their exposures (real estate, corporate) at the group consolidated level.

• Banks and insurers must therefore step up their efforts to combat climate change today by integrating climate risks into their financial risk assessment process, as these efforts will contribute to the changes that will be observed in the medium and long-term. Taking better account of climate risks is indeed necessary to promote a better allocation of resources and ensure the financing of the transition. While banks and insurers seem to be generally aware of this issue, their degree of maturity remains heterogeneous and some institutions have not necessarily yet integrated the proper degree of urgency to act.

This exercise also brought to light a number of methodological limitations on which progress needs to be made. It therefore marks the starting point for further work to improve the methodology of climate stress tests. The main points for improvement identified by the ACPR concern:

• The hypotheses used to create the scenarios and identify sensitive sectors: one difficulty encountered by the institutions that took part in the exercise was the low variability between the different scenarios put forward by the ACPR. This point had already been identified by the ACPR, which had led it to add a sudden transition scenario to the scenarios published by the NGFS. Moreover, the models used by banks to quantify risks are not able to integrate very smooth evolutions of macroeconomic and financial variables over a long period. The same applies to insurers, which are used to dealing with extreme climate shocks but not with smooth and deterministic impacts over a long period. The very long time horizon also implies costly work in projecting transition matrices. Finally, the absence of feedback effects between the sectoral structure of the balance-sheet of the financial sector and financial risks (generated by climate change) does not necessarily encourage institutions to implement an active risk reduction policy, as most of the transition scenarios considered reach the objective of carbon neutrality in 2050. A second difficulty lies in the identification of sectors that are sensitive or exposed to climate risk: this identification first of all depends on the method used. It then implies assumptions on the evolution of the energy mix, the intensity and the energy efficiency of production, which were not properly integrated in this exercise. Finally, there is the question of sectoral granularity and the linking of exposures or counterparties to a given nomenclature, in particular when those counterparties are active in several economic sectors.

• Taking into account the "physical risk" is a notable area for improvement on which collective work is also needed because it also implies taking into account interdependencies and a sound knowledge of the value chains, which remains largely insufficient. One of the main reasons for this is the absence or incomplete nature of the information published by companies. This obstacle may be gradually reduced with future ESG disclosure requirements for companies (at least at the European level). As regards the insurance sector, further work should be conducted on the insurance protection gap.

• The improvement of the models used by banks and insurers and the data sources is needed to take better account of climate risk (in particular at the sectoral or company and counterparty level). Several interesting methodological avenues implemented by financial institutions in the context of this exercise should be explored in greater depth.

Next steps: the results obtained will be followed up by the setting-up of new working groups with the Paris financial centre and with external counterparties. In addition, ACPR and Banque de France experts are actively contributing to the preparation of the exercise to be conducted by the ECB in 2022 and to the European and international work conducted in several fora such as the European Insurance and Occupational Pensions Authority, the Basel Committee on Banking Supervision, the International Association of Insurance Supervisors and the Financial Stability Board. This financial risks assessment exercise induced by climate change will be repeated regularly. The next ACPR exercise may take place 2023/2024.