Climate risk analysis is currently subject to significant data gaps – Some firms estimated loss rates for the same counterparty are ten times higher than other firms. The BoE has clarified that the results represent a first step towards quantifying the financial risks of climate change. Firms need to continue to develop and refine their approaches to climate risk management beyond this exercise, as they embed climate risk processes and tools across their organisations. Firms used data items such as addresses, postcodes, energy Performance certificates (EPCs), industry codes, countries of incorporation and operation, and insurance flags for the first time in a regulatory exercise, which left a challenge to align data. Firms should continue to build strategic solutions to integrate this data, as successfully sourcing and mapping these to traditional and climate data is crucial to climate risk modelling. As firms consider expanding the reach of their modelling, they should enhance collection to include data for calculating secondary risk channels and other aspects of ESG.
The CBES required banking and life insurance participants to engage with their corporate counterparties to obtain a greater accuracy and coverage of information. Firms struggled with this aspect of the exercise. While many larger counterparties could answer the questions asked of them, some smaller counterparties were unable to submit meaningful data. Since the exercise, counterparties’ consideration of climate risk has expanded and requirements to make TCFD-aligned disclosures are being introduced across the UK. Client outreach is a valuable strategy for data acquisition in the early stages of embedding climate risks and beyond. Firms should consider engaging in periodic client outreach to harness counterparties’ sector knowledge and alleviate the difficulty, should they need to repeat the exercise.
While the exercise provided scope for participants to determine their own approaches to the modelling, there was also a requirement to calibrate climate models to the dataset provided by the BoE. Some participants found difficulty in aligning to the data provided, which is likely indicative of the uncertainty inherent in these models at this early stage in their adoption and so the wide range of possible results. As this area of modelling continues to develop and to drive disclosures, there will need to be an increasing focus on validating the output of climate models against both internal and external information.