Europe's recent regulatory climate stress tests aim to reveal the vulnerabilities for the banking sector. Beyond some limitations, not least in data quality, S&P Global Ratings sees them as a valuable way to start to gauge environmental and climate-related risk exposure. They should also indicate whether the institutional framework can be effective in addressing systemwide risks. We have been increasingly incorporating environmental considerations and data into our credit rating analyses. As testing becomes more sophisticated over time, it should suggest whether banks' management teams need to adjust risk appetites, develop tools to better monitor and manage climate-related risks, and even consider changes to their business models.

Increasingly brutal and frequent extreme weather events that have lashed Europe in recent years mean policymakers need no reminder of the potential severe macroeconomic and social implications of climate change. While eyes are focused on what may emerge from the United Nations Climate Change Conference (COP 26) in the fourth quarter, in July the European Commission released its "Fit for 55" legislative proposal to underpin the EU's commitment to comprehensively address climate change. It targets a reduction in net greenhouse gas emissions by at least 55% by 2030, in line with a broader ambition to become the world's first climate-neutral continent by 2050.

Catalyzed by a strong political consensus, a mandate to ensure financial stability, and increased investor interest, European regulatory authorities have accelerated their efforts to identify climate-related risks (namely, physical and transition risks) and to assess the financial sector's exposure to these risks (see chart 1). The translation of climate-related risks into quantifiable financial risk has been one key output from this work.

Supervisors' climate stress tests were able to identify the main sectors and geographies generating transition and physical risks for banks' assets and sought to quantify their exposure to these risks. As such, they provide a forward-looking view on how different climate scenarios could affect banks' assets. Supervisors are not yet penalizing banks with higher capital requirements for such long-range risks, but they are ratcheting up the pressure on bank management to proactively tackle these risks.

Chart 1

Various Regulatory Initiatives Go In The Same Direction

Over the past few months, the European Central Bank (ECB), Bank of England (BoE), and the French Autorité de Contrôle Prudentiel et de Resolution (ACPR; the supervisory arm of the Banque de France) have started to conduct climate risk-related stress test exercises on the financial institutions they supervise. The ECB and ACPR reported high level results in the past two months; the BOE will report in the first half of 2022. We expect these initiatives to be performed on a regular basis and to complement the traditional solvency stress tests that supervisors have conducted for many years.

The nature of these climate exercises differs in many ways from the traditional stress-testing practices, like those the European Banking Authority (EBA) and ECB released on European banks on July 30, 2021 (see "2021 EU Bank Stress Test: More Demanding, Better Resilience," published Aug. 2, 2021).

The time-horizon of the climate stress tests is far longer.  The climate scenarios extend over 30 years, compared with three years generally covered by traditional exercises.

The scope is wider.   Both banks and insurers are involved and assessed under the same scenarios.

The scenarios are very different.   Unlike the traditional exercises, none of climate stress tests includes a GDP contraction in their macroeconomic assumptions. They rather look at transition scenarios, with different types of climate stress to assess vulnerabilities to physical and transition risks related to climate change. As a result, the magnitude of the modelled impact on bank capitalization is typically smaller than that observed in the traditional stress tests.

The objective is different.   Climate stress tests do not check that banks have a sufficient capital buffer to withstand adverse scenarios, but rather to evaluate those institutions' exposures to the risks induced by climate change and, ultimately, to raise their awareness of the implications of these risks on their business and strategy.

The climate stress tests are used differently.   They are not used to set capital requirements, nor to require individual entities' actions to adjust loan loss coverage. Instead, they are exploratory exercises that may inform the authorities' approach to system-wide supervisory policy and spur further collaboration and work between regulated entities and supervisors to address any issues highlighted.

A Consistent Approach To Scenario Analysis

The three regulators built their stress test scenarios on those published by the Network for Greening the Financial System (NGFS) in June 2020. These simulate different paths of government climate policy towards net zero greenhouse gas emissions.

• The "orderly transition" scenario considers a timely and effective implementation of climate policies, successfully limiting global warming, with therefore relatively contained costs from both transition and physical risks.
• The "disorderly transition" scenario implies a delayed and abrupt implementation of transition policies, with high costs of implementing green policies, but limited costs from damages generated by physical risk.
• The "hot house world" scenario implies no new climate policies are implemented, with consequently very limited costs related to transition risk, but extremely high costs associated to damage from physical risk.

In addition, the ACPR added a second adverse scenario for a disorderly transition, called the "sudden transition" scenario, which combines a sharp increase in the price of carbon and a less favorable evolution of productivity from 2025 onwards than in the baseline scenario. This increased the variability between the different scenarios, to further stress the macroeconomic and financial impacts.

Emerging Lessons From The ECB And ACPR Stress Test Results

One should not read too much into the conclusions of these exercises, not only because of significant uncertainty about the pace and impact of climate change, but also because these exercises did not consider spillover effects, like supply chain disruptions, or amplifications that are typically observed during financial stress or crises. Still, there are some notable takeaways, in our view (see chart 2):

(1) Further delays in addressing climate change will only increase the likelihood of environmental events.   This could severely affect the broader economy and ultimately destabilize the financial system.

(2) The impact of the climate scenarios on banks' balance sheets--for example in terms of expected losses--varies significantly.   This depends on the severity of the scenario.

(3) High correlation exists between banks' vulnerability and the concentration of their assets in geographies and economic sectors more sensitive to physical risks.   Regarding physical risk, firms' vulnerability varies across Europe, depending on the hazard. Floods are a relevant risk driver in many countries, especially in Central or Northern Europe. Heat stress, water stress and wildfires predominantly affect Southern Europe. Specifically, about 12% of sampled eurozone bank credit exposures to non-financial companies are subject to water stress, 11% to heat stress, 11% to flood risk, 5% to wildfires, 1% to coastal floods/sea level rise, with almost 10% of loans to non-financial companies being subject to multiple physical risk drivers. The most common combination of risk drivers consists of water stress and wildfires, complemented by heat stress, which means that, consequently, the portion of banks' exposures to firms operating in areas that suffer from multiple risks is particularly relevant for banks located in Greece, Spain and Portugal. Physical risk also generally occurs in specific economic sectors, like in accommodation and food, in transportation and in storage sectors, followed by manufacturing, and construction and real estate.

(4) Banks more exposed to physical risk tend to be less strongly capitalized and less profitable.   This suggests that physical risk may amplify other types of bank vulnerabilities, exacerbating the potential implications for financial stability. For example, the ECB analysis showed that among the 25% least well capitalized banks, by common equity tier 1 (CET1) ratio, the median exposure is six times higher relative to the 25% most well-capitalized banks. While more than 70% of the banking system credit exposures to the identified high-risk firms are held by 25 relatively large banks, these banks tend to be well-diversified by asset mix and geography, have sizable capital buffers, and generally exposure to physical risks lower than 7% of their total assets.

(5) Collateralization plays an important role in limiting credit losses related to physical risks, with more than 60% of banks' loan exposures to firms that are subject to physical risks secured by collateral.   However, half of this collateral consists of physical assets, which suggest risks around their potential devaluation.

(6) Banks may struggle to take physical risks into consideration.   The ACPR noted that banks generally do not pay lot of attention to the physical damages of their own facilities, as the latter is mostly handled by insurers. Banks therefore struggled to model the possible impact of changes in insurance premiums and coverage policies on their credit risk parameters, in particular loss given default and probability of default.

(7) Transition risk is concentrated in key sectors.   Bank loan exposures to sectors vulnerable to climate change are mainly concentrated in the high emission-intensive sectors, like manufacturing, electricity, transportation and construction sectors. These account for a limited 11% of loans to non-financial companies. Conversely, the majority of loan exposures to non-financial companies are to low or moderate emissions-intensive sectors, which could indicate that risks to financial stability is broadly manageable.

(8) French banks and insurers appear to have an overall moderate exposure to transition risks.   In the orderly transition scenario, institutions' aggregated risk costs rose by 22% between 2025 and 2050, with the corporate portfolio accounting for the bulk of this growth. Unsurprisingly, under the more adverse scenario, cost of risk was about 30% higher than that under the orderly scenario.

(9) Non-EU exposures could generate a disproportionate rise in cost of risk.   In the ACPR exercise, Europe accounted for about 75% of banks' exposures, with exposures to the U.S. accounting for a modest 9%. Although the majority of credit losses comes from the European portfolio, the non-EU portfolios suffer from a larger increase in the cost of risk under the more adverse scenarios. This mainly reflects a higher concentration to sensitive sectors in the U.S. corporate portfolio. That said, the exposure of French institutions to the sectors most affected by transition risk, as identified in this exercise (for example, mining, coking and refining, oil, agriculture, construction), is relatively low.

(10) Physical and transition risks may be dependent on each other.  Greater policy action could reduce physical risks in later decades, but increase the impact of transition risks in the meantime.

Chart 2

Estimating The Greenness Of Banks' Activities Remains A Key Challenge

European supervisors and banks are making remarkable progresses in quantifying the financial system's vulnerability to environmental risks. These efforts will likely help reduce information asymmetries, ease transparency, and improve comparative analysis on environmental data, in our view. That said, the journey is still long, given the challenges they face to identify the greenness of activities, and to classify and measure climate-related risks.

These challenges were clearly explained by the EBA in its Advice to the European Commission, published in March 2021. Not surprisingly, the main obstacles for designing a climate risk stress test framework, according to EBA, are: the lack of a comparable and consistent dataset of climate risk indicators, and the challenge to project the impact of climate change over a much longer time horizon (generally 30 years) than the one- to five-year periods that banks and supervisors typically use for business planning and stress testing exercises.

The EBA used two approaches to assess greenness, with divergent results:

• Under the sector-based approach (which mapped the standard EU classification of economic activities into categories that are relevant for climate transition risks): more than half of banks' exposures (58% of total non-SME corporate exposures to EU obligors) were allocated to sectors that might be sensitive to transition risk; and
• Under the emission-based approach (which used the carbon emission intensity of banks' exposures): about 35% of the total non-SME corporate exposures were to EU obligors with greenhouse gas emissions above the median of distribution.

These discrepancies emphasize the complexity of defining a universal methodology to measure the impact of climate risks on banks' balance sheets. However, both approaches shared one conclusion: the manufacturing, utilities, construction, transportation and real estate sectors are highly exposed to transition risks. They amounted to 50% of total exposures submitted in the exercise.

Banks Are Not Yet Ready To Report In Line With The EU Green Taxonomy

The EBA tested banks' readiness to apply EU green taxonomy--from 2022, EU banks must disclose the alignment of their exposures with taxonomy criteria. Some 26 of the 29 banks also provided an estimate of the greenness of their exposures, using the EBA's suggested key performance indicators, notably the green asset ratio (GAR), which is constructed for each bank by dividing taxonomy-aligned exposures to taxonomy-eligible exposures.

The exercise revealed that only 33% of banks' exposure are to sectors in the scope of EU taxonomy, with the bulk located in the real estate and construction sectors, but only part of this exposure is compliant with the taxonomy criteria and hence, taxonomy-aligned. This explains why the average GAR for the banks in the sample was a low 8%.

Questionnaire responses revealed that banks are at different stages in applying the EU taxonomy: while one-third of them were able to classify almost all of the exposures, one-third could not classify even 10% of their exposures. Availability of client data appears to be a common deficiency: 80% of the sampled banks doubted that even 30% of their clients would be able to provide taxonomy-based information in 2022. It therefore looks highly challenging for banks to leverage on the EU taxonomy to measure their exposure and vulnerability to climate risk.

Banks are likely to enhance their climate risk management in coming years

We expect that the European regulators' initiatives will generate benefits far beyond the insights derived from the stress test results. Over time, we anticipate that a growing cadre of banks will develop an effective climate risk management framework, which, according to Bank for International Settlements' April 2021 guidance, should have three goals:

• Identify material climate risk drivers and their transmission channels;
• Map and measure climate-related exposures and any area of risk concentration; and
• Translate climate-related risks into quantifiable financial risk metrics.

In our view, financial institutions will leverage the regulatory scenario assumptions to develop their own scenario analysis to measure their vulnerability to climate risks. This is helped by the fact that banks and regulators almost agree on the metrics to use in identifying, mapping and measuring the climate-risk exposures: the most common parameters to assess transition risk are carbon footprint of assets, or indicators related to "greenness" of financial assets and real estate exposures. They similarly agree to some extent that metrics to assess physical risk focus on the type of hazard associated with the banks' geographical exposures, and the related probability and potential severity. Increasingly, we expect this rising sophistication in data analysis and risk management to influence risk appetite and banks' long-term strategic plans.

Greater Disclosure On Banks' Exposure To Climate Risks Will Inform Our Ratings

We already incorporate in our ratings environmental factors and risks associated with climate change, when material from a credit perspective and sufficiently measurable/predictable--even beyond our outlook horizon.

We consider in our ratings only the environmental (and other ESG) factors that can materially influence the creditworthiness of a rated entity or issue and for which we have sufficient visibility and certainty to include in our credit rating analysis. We define them as "ESG credit factors." These credit factors can have a negative or positive impact on creditworthiness, depending on whether they represent a material risk or opportunity (see chart 3).

Chart 3

Although only a few rating actions on banks have stemmed from climate and environmental risks so far, we acknowledge that the relevance of these risks and opportunities is rising, as are the effects they might have on financial systems. We are therefore increasingly incorporating environmental considerations and data into our credit rating analysis. As supervisors and banks provide greater transparency on the financial sector's vulnerability to these risks, this will likely increase the quality and the quantity of data we could leverage in our credit analysis; for instance in our credit loss estimates at both system and individual bank levels. Fuller and more comparable disclosures of banks' exposures and vulnerabilities to climate and environmental risks would help us further differentiate between banks.

Related Research (S&P Global Ratings)

• 2021 EU Bank Stress Test: More Demanding, Better Resilience, Aug. 2, 2021
• ESG In Credit Ratings Newsletter, July 22, 2021
• Embedding Environmental Factors In Strategy And Risk Management: For Banks, A Long Journey Just Begun, Sept. 28, 2020
• Standardized Non-Financial Disclosures Will Clear The Path For Better ESG Analysis, June 19, 2020
• Delays In Addressing Global Warming And The Longer-Term Ratings Implications, Dec. 3, 2019
• The Role Of Environmental, Social, And Governance Credit Factors In Our Ratings Analysis, Sept. 12, 2019
• The EU Green Taxonomy: What's In A Name?, Sept. 11, 2019

Related Research (External)

• ECB Climate-Related Risk And Financial Stability Report, July 2021
• Bank of England: Key Elements Of The 2021 Biennial Exploratory Scenario: Financial Risks From Climate Change, June 2021
• ACPR: A First Assessment Of Financial Risks Stemming From Climate Change: The Main Results Of The 2020 Climate Pilot Exercise, April 2021
• Bank for International Settlements: Climate-Related Financial Risks--Measurement Methodologies, April 2021
• Bank for International Settlements: Climate-Related Risk Drivers And Their Transmission Channels, April 2021
• European Banking Authority: Advice to the European Commission, March 2021
• NGFS Climate Scenarios For Central Banks And Supervisors, June 2020