Financial institutions need to start building a climate dashboard
Regulators are sending a message by doing climate stress tests that using incomplete data and simple methodologies is preferable to doing nothing. Financial institutions that acknowledge that climate change is having a financial impact on them and that this will grow in the future should take the same approach.
- Climate stress testing is going mainstream, with 31 regulators having completed or begun one, so banks need to keep up their work on climate-related risks
- Financial institution (micro-level) climate risk metrics measure different things than regulatory (macro) climate stress tests, and banks should look at the results of both types together
- Financial institutions should connect their climate risk metrics to see how they are impacted when adopting short- and medium-term Net Zero targets
The process of stress-testing the financial system for climate risk is in its earliest days, and data and methodologies are likely to continue to improve over time as banks and regulators gain more experience. With stress tests considered, being planned or having been announced by several Islamic markets including Malaysia, the UAE and Bahrain, financial institutions will need to prepare to develop experience on climate risk.
The experience financial institutions will need includes working to improve their own climate risk approach in parallel with regulatory stress tests, as well as understanding how Net Zero commitments will factor into that picture. In the process, they’ll need to continually find ways to validate or challenge their analysis as a way towards continual improvement.
The French central bank, Banque de France, has released a summary of its experience with climate stress tests, including the challenges it identified during the process. The summary highlights several points that are relevant to any financial institution working to improve its climate stress test readiness. These can be broken down into three main issues:
1) Most regulators have used scenarios from the Network for Greening the Financial System (NGFS), although some tailor the stress tests to specific nation-relevant issues;
2) Many of the stress tests have been challenging for banks because the scenarios used unfold over a longer period of time, rather than a ‘shock’ followed by a response;
3) Data is a persistent challenge in undertaking a stress test because different sectors (and sub-sectors) will respond differently to different types of climate stress.
Climate stress testing is just one part of the financial system’s response to climate change. Financial institutions shouldn’t just view climate risk as a regulatory compliance issue for biennial or less frequent stress testing. The methodologies used by regulators are not perfect, but neither are banks’ own analytics for measuring financed emissions. Each shows one dimension of climate risk and is best used as part of the overall picture guiding financial institutions’ strategy, compliance and risk management processes.
Ultimately, different types of metrics for climate risk or exposures are serving different uses, which is why many banks opt to track climate metrics using a ‘dashboard’ rather than a single indicator. For financing of specific assets, banks would want to know how much of the physical assets are susceptible to physical risks such as flooding, storms, heatwaves and windstorms. The risk to the banks may be limited (directly) if the assets are insured, and may be of more consequence to the insurers. However, banks would have an interest in the exposure to physically at-risk assets and the financial strength of the insurers their customers are using because rising climate risk could increase the financial vulnerability of the insurers.
Even if the insurers are themselves financially strong, if their coverage is renewed annually, climate risk could lead to rising premiums or loss of coverage, which would weaken the repayment capacity of the borrowers and push more risk onto the banks. For this metric, banks might want to know the share of their financing that is going to geographical locations where the acute physical risk is higher, and to know whether the higher-risk places are becoming more or less vulnerable. The concern is mostly related to changing risk for individual counterparties that could increase their chances of defaulting and the amount of the bank’s exposure if a default occurred.
This contrasts with the climate stress tests that regulators are doing, which include physical risk elements, but often focus more on the financial impacts of macroeconomic changes triggered by a faster or slower pace of decarbonization and a faster or slower pace of physical risks materializing. Because climate stress tests are typically done at the national level, they are looking at whether the size of the risks could be destabilizing for one or many banks, which would have spillovers onto the ability of the financial system to function smoothly.
One of the challenges that Banque de France found in its stress testing was that banks’ models “to quantify risks are not adapted to incorporating extremely smooth developments in macroeconomic and financial variables over a long period”. Returning to the bank-level concerns about climate risks, these models are trying to measure risk in terms of default probabilities and loss once a default occurs.
The bank-level concerns about climate risks at the counterparty or sub-portfolio level contrast with the macroeconomic scenarios for climate risks that assume gradual impacts from different decarbonization paths and that “none of the analysed scenarios results in an economic recession by 2050”. It is likely that the climate scenarios find just a small incremental increase in default probabilities or loss amounts, and this compounds each year over the long time frame studied through 2050.
In light of the climate-related disruptions that have grown in severity and scale in recent years, this approach to climate stress testing probably dramatically understates how a ‘severe shock’ would impact financial stability. Banque de France specifically highlights how the NGFS is working to include many characteristics that could create a much more severe shock, including “non-linear dynamics; the existence of tipping points; irreversibilities; and materialisation over short-, medium- and long-term horizons.”
After adding elements measuring the top-down financial stability metrics to bank counterparty or sub-portfolio risk characteristics, financial institutions have a better view of how their risk looks, but very little vision for what they’re doing about it. Net Zero commitments by financial institutions have become much more commonplace since 2021, and they would also need to be reflected in a dashboard of climate risks and opportunities.
The bottoms-up (counterparty or locational) risk and top-down (financial stability) risks almost always assume continuation of business as usual. Net Zero commitments outline a different view on the future of the bank’s customers and their emissions risk profiles. Merely making a commitment by 2050 doesn’t change the future path of the bank or its customers, and wouldn’t impact the risk (micro- or macro-), which is why intermediate targets are so important.
Most financial institutions won’t approach Net Zero targets as immediate cause to implement ‘portfolio decarbonization’ and the implementation of such an approach would be ineffective and disruptive, moving financed emissions from one part of the financial sector to another without reducing them. At the same time, a long-term target with nothing more would be an invitation to wait until the last minute, or as long as possible, which would be as effective as doing nothing.
The pace of movement towards Net Zero targets will also need a place on the climate risk dashboard. Depending on what a financial institution decides is its short- and medium-term targets, its climate risk (on a micro and macro level) will change. These changes are rarely incorporated into regulatory climate stress tests, although some of the stress tests have added consideration of the medium-term (3–5 years and beyond) response by banks to the realization of climate risk in their starting (static) portfolio.
Underpinning all of these types of climate risk dashboard items is a huge amount of methodological uncertainty. We know that climate risks (both physical and transition) will be too large to ignore, and likely will imply seismic changes in the structure of economies. Lack of discrete bottoms-up data and inability to know how much physical risks will accelerate the pace of change, rather than the smooth pace implied by the scenarios in use, means that each dashboard item will show slightly different outcomes for climate risk.
The overwhelming message regulators are sending by doing climate stress tests is that incomplete data and simple methodologies are an ok, and are preferable to doing nothing. Financial institutions that acknowledge that climate change is having a financial impact on them that will grow in the future should take the same approach. Start defining what you can with the data on hand, and commit to a process of continual improvement over time as data and analytical tools improve.
Want to learn more about responsible finance in Islamic markets & Islamic finance? Subscribe to RFI’s weekly email newsletter today!