by: Simon Axon
Climate risk is racing back to the top of the agenda at many banks. As they learn to adjust to the changes enforced by the COVID pandemic, the attention of customers, regulators and shareholders is returning to the other global crisis facing all of us – climate change. With the COP26 Glasgow Summit just around the corner we can all expect renewed scrutiny on this critical area. But how does data play a role and what should banks be doing now to prepare for this increased scrutiny?
Pressure to respond to climate change comes from many different angles – but two in particular stand out. Customers want to see evidence that their Bank is acting to mitigate environmental damage, and bank regulators are demanding proof that banks are resilient to the wide range of new risks stemming from climate change.
Reputational Risk
The first is a reputation and marketing challenge. It is no longer adequate to just include pro-environmental policies in the annual report, or even to develop and publish detailed sustainability reports highlighting the actions taken by the organisation to mitigate its own impact. Both are increasingly essential in the current environment, but on their own they do not go far enough. Customers – including personal banking as well as corporate and investment clients – demand evidence that their Bank does not support environmentally damaging businesses with its lending or other services. Adequate action on ‘Finance Emissions’ will be critical to the reputation of any financial institution. If customers aren’t convinced, then they are increasingly likely to take their business elsewhere.
Financial Resilience Risk
From a bank regulator’s point of view, they need evidence that banks are taking account of increasing climate risks in their risk and governance processes. Demonstrating understanding of direct risks (for example business disruption from extreme weather) is just the tip of the iceberg. Banks must be able to ‘show their working’ and highlight a detailed understanding of the impacts of climate change risks to loans, investments, value-chains and, ultimately their capital base. They must also demonstrate mitigation of ‘transition’ risks as economies shift away from fossil-fuels. It is critical that banks not only understand the financial threat to the ongoing solvency of the organisation created by these complex interactions and interconnections but can demonstrate that they have effectively modelled this risk across a wide enough range of scenarios.
Increasingly complex data sets
Both challenges require data to be at the heart of the solution. Increasingly that data must be sourced, integrated and understood not only from across the entire organisation, but also from many external sources. Long-range predictions and modelling from academics and meteorological experts will need to be combined, for example, with investment data by sector and geography to model the impact of climate events on assets and revenues. Forensic data on investments must include environmental performance of investees as well as their exposure to climatic events. This too will require ingesting data from a wide range of external sources.
Layering in these new data sources can quickly create a messy, complex data environment with silos of specific information held by separate teams, duplication of effort and data, and inconsistencies as different data are combined in different ways to generate different results. Governance and auditability – the ability to show exactly how a specific decision on climate risk exposure was made – become virtually impossible.
Top-down data strategy
A planful, top-down, approach to creating an enterprise-wide data platform is essential to avoid this outcome. Understanding what data is where is the essential first step. Understanding its lineage, sensitivity and quality begins to create the building blocks of a single, trusted source of data that can underpin risk modelling and decision-making. Huge reductions in the time and effort need to find, prepare and ‘wrangle’ data prior to building models will make data science processes more efficient. Data features with proven utility can be stored, catalogued and re-used not only further improving efficiency, but providing audit trails to support governance and show why and how specific risk decisions were made.
This data infrastructure needs to be agile and flexible. We are not advocating the heavy integration of all data – not only is that cost and time intensive – but in the fast-moving area of climate risk, new data and new sources of data are constantly appearing. Banks must have the ability to ingest all types of data and rapidly update models to take account of it. Any enterprise data strategy must incorporate not only the robustly engineered features that can be used time and again in production, but the ability to quickly experiment with lightly integrated, or unintegrated data to explore and predict new outcomes with new data.
Over the past eighteen months or so focus has drifted from the climate crisis – but it will return rapidly over the coming months. Stakeholders from across the spectrum will be asking increasingly searching questions of banks about their climate risk capabilities and strategies. Answering them will require effective use of granular data from multiple sources, at scale. Now is the time to ensure data architectures are well deployed to meet these challenges.
