Stress testing often involves resilience to a single event, comparing projected losses with capital levels. In climate stress tests, the dollar value of projected loss is often much less informative than the qualitative gaps of internal capacity, governance and data quality they can expose, which are harder to reinforce than financial capital buffers.
- Climate stress testing has moved up regulators’ agendas but shouldn’t be seen as just another stress test
- Assuming that climate change is a ‘stress event’ that just has to be weathered understates the permanence of climate risk for financial institutions
- Regulatory stress testing is valuable where it highlights weaknesses in capacity, governance and data, which take time to fix in order to improve the financial sector’s resilience
Stress testing the financial system has become an important tool for regulators who want to understand whether individual financial institutions are contributors to systemic stability or instability should there be a future shock. In evaluating this question for ‘traditional’ financial risks, history provides a good guide to what an adverse event would look like, and how macroeconomic stresses would translate into financial institutions’ losses. It’s different for climate change.
That’s not to say that regulators shouldn’t pursue climate stress testing — this is more important than ever — but the desired results should be different than for other types of stress testing. What we’ve learned this year from the results of climate stress tests conducted by the Bank of England and the European Central Bank is that banks often don’t have enough internal capacity to explain their climate risks.
The Bank of England’s Anil Kashyap, a member of the Financial Policy Committee, included among his take-aways from the bank’s Climate Biennial Exploratory Scenario (CBES) the question of whether banks’ staff would have “the needed data and expertise to effectively utilize and eventually challenge the consultants” they relied on in the stress tests.
The question he raised came from the ability to observe how different banks approached climate modeling and how different responses from individual banks highlighted gaps in the data or the expertise they had to integrate scenario-specific inputs into their responses. That’s a very different output than what would come from a traditional stress test, where individual banks’ assumptions may vary from the banks or from what the regulator would expect, but in many cases, the differences would be driven by the bank having more information about its own sensitivity to the shock, not less.
In bank stress testing, the range of inputs will be much more diverse than just macroeconomic assumptions and neither banks nor regulators will fully know how they will impact the economy. The degree of uncertainty in a climate stress test will be orders of magnitude greater than a traditional stress test. Furthermore, the time period in which the bank will be stressed is different.
Rather than testing the reaction function to a single shock, climate stress tests focus on both physical and transition risks, which introduces a new feedback loop. If the stress test were narrowly focused only on the ability to withstand the impact of a physical risk event, such as widespread flooding, it would more closely resemble a traditional stress test. There would be a ‘stress event’ and then regulators would get information about whether there would be enough remaining strength among financial institutions to weather the stress to come out the other side strong enough to finance recovery.
A climate stress test overlays that type of stress test for a single stress event and recovery with an ongoing source of endogenous shocks to the economy based on changing access to financing in specific sectors. The result of the second type of stress requires banks and investors operating in competition with one another to coordinate the reallocation of their financing at just the right pace, where too rapid shifts can destabilize the economy or and too slow can undermine the transition and increase physical risks.
The output of a stress test will combine all of these distinct elements together with huge data gaps because the impact of climate change will change the correlations that prevailed in the past. Looked at in another way, the output of climate stress tests, which is often measured by the financial losses that result, is not the primary metric of success at the end.
Most stress tests take the financial loss and see whether it causes capital deficiencies across the financial sector. Climate stress tests are different because the quantitative test is less important than what they reveal about a financial institution’s internal capacity, governance and data quality. We shouldn’t take solace when climate stress tests show that there’s enough capital in the system to weather any foreseeable climate risk.
Climate risks are even less likely than other risks used in stress testing to portray an accurate view of the future. The qualitative results from the stress test measuring internal capacity, governance and data quality — as well as how it is all put together in the response to the specific stress scenario — will hold far greater value to point regulators and industry stakeholders towards the source of the weaknesses that need to be remedied. And those weaknesses are likely to take the longest to fix, which is why the pace of climate stress testing is moving so quickly up the agenda of regulators.
Want to learn more about responsible finance in Islamic markets & Islamic finance? Subscribe to RFI’s weekly email newsletter today!