Integrating models into risk assessment under uncertainty and the cost of delayed action

On 6 July 2022, the OECD held a workshop on Climate Transition Scenarios: Integrating models into risk assessment under uncertainty and the cost of delayed action. The workshop gathered over 260 participants, including over 200 joining virtually and around 60 in-person at the OECD Headquarters in Paris. It was held under the Chatham House rule, with participants speaking on their personal capacity, in order to encourage an open and lively discussion. The workshop was hosted thanks to support from ADEME and Finance for Tomorrow in the context of the Finance ClimAct initiative, supported by a grant from the LIFE program. The workshop was held back-to-back with the OECD Committee on Financial Markets (CMF), and participants included representatives from central banks, financial supervisors, treasuries, climate modellers and researchers, climate policy makers and academia.

This workshop was the second of a series of OECD workshops on Climate Science, Policy, Regulation and Practice, to foster an outcome-oriented dialogue between different communities, including climate modellers, scientists, policy makers, financial regulators, investors and other financial market practitioners and civil society. As part of this series, the OECD held a first virtual Workshop on Climate Change: Assumptions, Uncertainties and Surprises in 2020, which highlighted key issues relevant to the design and use of climate research and science within policy, regulation and practice. It framed key issues to support economic and financial decision actors to interpret and use scientific evidence and modelling in their approaches to climate risks.

This workshop was hosted thanks to support from ADEME and Finance for Tomorrow in the context of the Finance Climact initiative, supported by a grant from the LIFE program.

The workshop aimed to foster knowledge sharing and dialogue on climate transition scenarios more specifically. In a context of deep uncertainty, the use of transition scenarios can be used by financial actors to understand the mid- to long-run challenges associated with the low-carbon transition and better manage climate transition risk. Great progress has been made with the Central Banks and Supervisors Network for Greening the Financial System (NGFS) Climate Scenarios, as well as with the IEA and other models. However, there are still fundamental limitations and gaps that need to be bridged to adequately incorporate climate risks within financial institutions’ risk management practices.

The first session discussed the outcome of the exercise undertaken by the four modelling teams to apply the NGFS scenarios to the French context in the short-term, and the limitations from the necessary assumptions required to apply the scenarios within the near-term context. Available scenarios offer high-level narratives to illustrate the potential trajectories of the economy under different scenarios, but they are difficult to translate into near-term implications of climate change and the transition for private and public decision makers. Progress needs to be made to understand the how to overcome to unmodelable risk factors, including policy uncertainty, and the implications of using various assumptions and interpretations of the scenarios in the short-term. As stressed provocatively during the session, we may fear more adverse scenarios because we fear higher heterogeneous carbon prices, and because we fear more frictions in the multiple transition process. The war in Ukraine also illustrates the type of energy-related challenges that might impact and determine the transition pathway towards a net-zero economy. The current high inflation environment is also changing the approach to climate change. As discussed, central banks entered climate change from financial stability but now there is a growing need to also focus on price stability. Hence, the discussions seem to point towards a potential trade-off for central banks, between accepting higher price volatility in the short term, or the materialisation of greater financial risks over the medium to long term.

The second session discussed how stranded assets need to be included within our analysis of transition scenarios and the wider impact on the economy, particularly with regards to the different between and orderly and disorderly transition. Stranded assets also need to be integrated into the prudential tools of central banks, such as stress tests. While great progress has been made on this topic, there are still areas of future research, including to better assess climate uncertainties and expectations and assess networks of stranded assets.

The last session discussed how to integrate the results of climate risk modelling within financial participants and supervisor’s risk assessments. Although quantifying uncertainty in climate change creates challenges that are missed by commonly-used risk-based methods, several methodologies can help better identify assets that may be subject to climate transition risks. Due to the uncertainty inherent within climate risk modelling, sometimes a more simplified model may offer greater insight, with ‘less is better’. Consequently, shorter, less ambitious scenarios, may offer greater insights for financial market participants and regulators. Participants also debated on the value of undertaking climate stress tests with a shorter time horizon versus a 30-year time horizon. Participants also stressed that we do not need to wait for the perfect model and scenario to act.