At the Federal Reserve’s October 2021 Stress Testing Research Conference, Governor Lael Brainard commented on the Fed’s near-term intent to thoughtfully define a first-generation climate scenario analysis for US financial institutions.
Governor Brainard acknowledged the complexity of modeling novel combinations of risks, absence of historical precedents, need to consider financial interdependencies and regional differentiation, and difficulty predicting policy, technology, and behavioral changes affecting economic transition. But she nonetheless framed expectations by indicating that “Even … a rudimentary first attempt…will help with risk identification and suggest useful lessons to inform subsequent improvements in modeling, data, and financial disclosures.”
The Fed is specifying scenario analysis, rather than traditional stress-testing, as the current approach to estimating the potential effects of climate change on individual institutions and financial markets, avoiding mention of capital charges. The Basel Committee on Banking Supervision, responsible for setting the global standards for the prudential regulation of banks, provides a helpful distinction between the two approaches. Scenario analysis is “a forward-looking projection of risk outcomes” that identifies a range of plausible, unified climate physical and transition risk scenarios and economic outcomes; links their impacts to financial risks; assesses sector and/or counterparty sensitivities to those risks; and, based on these impacts, calculates an aggregate measure of exposure and potential losses. In comparison stress testing is a “specific subset of scenario analysis, typically used to evaluate a financial institution’s near-term resiliency to economic shocks, often through a capital adequacy target”.
The Fed will build on the different objectives and approaches to conduct climate stress tests specified by regulators for financial institutions in the UK, France, and greater Europe. Ideally, investors can tap into scenario analysis models, incorporate the insights into mainstream risk management and investment decision-making, and build greater portfolio resilience. Scenario analysis enables an investor to be more informed and therefore better able to plan for potential climate outcomes and monitor early warning signs.
Most of the regulatory exercises looking at systemic climate risk are based on the Network for Greening the Financial System’s (NGFS) reference scenarios that integrate climate science, economic trajectories, societal choices, and human development scenarios. These scenarios provide anchoring information on key demographic, macroeconomic, and energy mix and carbon pricing variables. Multiple scenarios explore:
- Orderly outcomes (from the early implementation of progressively more stringent climate policies, with a higher likelihood of less than 2°C global warming),
- Disorderly outcomes (from later or varying policy implementation, which have the potential to limit global warming to 2°C but with significantly higher transition risks), and
- Scenarios where more severe global warming of 3°C and higher ensues.
A range of methodologies exists for using scenarios to evaluate the climate risk of portfolios, including:
- Top-down approaches that model at the asset class or sector level, and
- Bottom-up approaches that assess entities’ individual assets and transition risk exposures and aggregate to the portfolio level.
There is a general tradeoff in accuracy and complexity using either approach. At The Climate Service, we model at the asset level and then aggregate losses to the portfolio level. This modeling is complex as it uses locational climate data and performs specific vulnerability assessments for all assets. In doing this, we have found significant variation in climate risk among entities within a sector, which suggests possible inaccuracies in a top-down approach. We are also actively exploring how to incorporate insights from top-down approaches to supplement our bottom-up analysis.
In general, physical risks will reduce the value of a financial portfolio and increase its credit risk, with the risk growing over time. Transition risks are more sensitive to the scenarios for individual sectors and entities, but generally peak in the nearer decades and taper off.
While larger banks are exploring both NGFS and bespoke scenario analysis, simplified or qualitative stress-testing approaches can also provide valuable insights, as the capacity to run comprehensive climate scenarios is developed. This can include, for example, identifying portfolio segments that are based on high carbon-emission entities, and evaluating the impact of carbon price shocks. Alternatively, insights from a bespoke approach, one that uses basic climate scenarios to evaluate assumptions and prospects for a key sector or portfolio segment, can provide the basis for quantitative assessments later.
As climate stress-testing evolves, TCS will provide updates and commentary, to help guide entities looking to understand potential climate risks in their portfolios, incorporate insights into risk management, and define strategies for more resilient outcomes.