How are financial institutions updating their data strategy to prepare for climate-related requirements?

Just as the world has undertaken the first Global Stocktake since adopting the Paris Agreement, financial services organisations are under increasing pressure to measure and manage emerging climate risks. Specifically, financial institutions are likely facing:

1. New data requirements across many areas of operations. For example, requirements around data for risk management, strategic decision-making, and reporting. This shift is due to rapid developments in climate-related regulatory and reporting requirements, as well as increasing expectations from stakeholders.
2. Increasing regulatory focus on quantitative measurement and management of emerging climate risks. For example, initial regulatory climate stress-testing exercises have already been conducted in Australia¹, UK, France, and Singapore.

Use case 1: Emissions calculations

Calculating emissions (including baselining and projections) is key to meeting strategic and regulatory requirements. At the same time, companies can use this higher quality data to develop deeper insights and form new initiatives to gain a competitive advantage in the market. Achieving both things requires a well-planned data strategy. 

Specifically, an organisation’s data strategy should cover: 

1. How does the organisation handle a lack of data? 
2. How do they address data quality considerations? 
3. How is data being converted into useful and reliable information? 

Scope 3 emissions are expected to make up the majority of financial institutions’ emissions (given the large emissions contributors that exist across the supply chain and downstream). Currently, a number of financial institutions find it challenging to accurately estimate these emissions given their reliance on the Scope 1 and 2 emissions of customers. Because customer emissions data is often not available, several assumptions must be made to achieve reasonable estimations. These assumptions will likely need to be appropriately supported to satisfy future regulatory requirements, and so companies must create a holistic data strategy that adequately addresses existing limitations. 

From the outset, organisations should identify what data is available internally and externally, so they understand the options available to meet data requirements. For internal data, consideration should be given to existing data that can be repurposed for calculating and estimating emissions. When leveraging external data, financial institutions should assess the reliability of customers’ reported numbers and, where accuracy is not deemed sufficient to be reliable, assumptions need to be made using other sources of data.

Financial institutions should develop a data strategy that’s relevant to their level of maturity and their requirements. For example:

• For banks: It is expected that carbon emissions quantification will be assessed through industry-sector and asset-class dimensions. Different sectors will likely require different data sources and strategies for dealing with missing data. Similarly, different asset classes will rely on different attribution factors for emissions. Any assumptions made can have a material impact on the emissions calculated and so appropriate review processes and adjustments should be considered within the strategy.

• For insurers: Insurers will need to consider the carbon and equivalent emissions from a portfolio and policy-type lens for accurate quantification. The data strategy would determine which sources of emissions would be in scope when measuring a customer's emissions and detail any assumptions.

• For investment firms: Investment firms will calculate the facilitated emissions from managing deals between two companies. A unique consideration for investment firms is understanding how synergies between two companies may impact absolute emissions.

While the list of data requirements is extensive, there are numerous objectives that can be achieved simultaneously, and a well-thought-out data strategy will help to achieve this. Building a solid foundation of emissions-related data can help financial institutions prepare for quantification, projections, and scenario analysis, as well as establishing and meeting net zero targets. A clear assessment of needs will also ensure cost-efficiency, support an effective strategy for carbon footprint reduction, and bolster an organisation’s ability to make the most of new opportunities.

Use case 3: Credit risk management and modelling

It’s increasingly clear that climate-related risks will affect the level of losses that banks can expect on their lending portfolios and, as such, banks will have to incorporate climate-risk considerations in their credit risk processes and underlying models. These considerations include credit risk origination, portfolio monitoring and credit risk modelling (both provisioning and capital).

It's important that banks assess the impact of physical and transition risks on their clients’ credit risk at the onset of any new relationship, as well as during ongoing portfolio monitoring. The data points required will depend on the client’s segments, as well as product and collateral types. 

When defining a data strategy for this use case, consider:

• An accurate assessment of physical risk is essential for both retail and non-retail portfolios, with a key focus on portfolios with assets secured by commercial or residential properties, as well as agricultural lands. Detailed location information is critical for this assessment, and banks need to assess whether this information is readily available and at the right level of granularity. A range of approaches could be used, including optical character recognition technology to read existing PDF files containing location data, using information collected as part of valuation exercises, or relying on external data providers. 

• Transition risk assessment is likely to require a granular view of how specific industries will be impacted by the decarbonisation of the economy, with transition risk impacting the probability of default as well as the loss given default (through the increased risk of stranded assets). A mapping of exposures to these industries is critical to ensure an accurate assessment, though the underlying data can be incredibly complex with a large number of corporate clients operating across a range of industries and countries. Remediating these data points will not only benefit transition risk assessment but will also support other credit risk processes. 

While it might be challenging for banks to update their credit risk models to incorporate climate-risk considerations, banks benefit in the long run. Having an initial assessment of how credit risk models could be updated in the future means that the data strategy captures the data fields required for future model rebuilds. Key considerations in this assessment include whether the bank will rely on environmental, social and governance (ESG) ratings to feed into their credit risk rating models for corporate clients, or how the value of collateral will be updated to reflect increased risks of natural disasters based on location. 

Banks could consider conducting basic exploratory analyses to assess the relationship between climate-related risk factors and historical credit losses. This will require collecting data such as granular hazard scores (which may not be straightforward) and organisations without an in-house natural periods model may consider external sources of data.

Given the complexity and granularity of data required to assess the impact of climate-related risks on their credit portfolios, banks might want to consider a tiered approach. Put simply, a tiered approach requires different depths of information for different industries or regions depending on the impact of climate change on those segments, as well as their materiality for the bank. Linking focus segments with the risk appetite of the bank and their net-zero strategy will also ensure efficient allocation of effort during this data exercise.

_{¹The Australian Prudential Regulation Authority conducted a Climate Vulnerability Assessment of Australia’s five largest banks (2022)
²Insurance Council Australia. (2023, October). Historical Normalised Catastrophe}