Data, AI and ML – how financial services can tackle the climate change risk
We're proud to be speaking at COP26 (the 26th annual UN Climate Change Conference of the Parties). As part of this, we'll be exploring issues around sustainability in a series of blogs, articles and more, and looking at everything we're doing to help build a better future. In this article, Head of Data Analytics, Rachael Bland, explains the challenges posed to the banking industry by climate change, and how data can help them mitigate those risks – and help fight the climate crisis in the process.
According to the European Central Bank’s Financial Stability review, 80% of banks are exposed to climate related risk.
In a speech given to an ECB-EBRD joint conference this summer, Central Bank board member Frank Elderson, said "All banks have several blind spots and may already be exposed to material climate risks… All banks need to catch up, as their climate risk undertakings will eventually influence their supervisory requirements.”
On top of this, only around 40% of banks have assigned explicit responsibility for managing climate risks to the management body - and of those, three in four do not report on climate risks to management.
In 2019, the Bank of England said up to £16tn of assets could be wiped out if the climate emergency is not addressed effectively.
To give you an idea of the scale, the annual GDP of Japan, the world’s third largest economy, is worth about £4tn.
What are the risks faced by financial institutions due to climate change?
To illustrate the risks banks face, let’s look at the UK mortgage market as an example.
In England, one in six properties (around 5.2 million properties) are at risk of flooding from rivers or the sea, one million of which are also at risk of surface water flooding. A further 2.8 million properties are susceptible to surface water flooding alone and 25% of that occurs outside of areas formally recognised as being flood-prone. Add to that, 6 million homes in the UK do not have any form of home insurance.
And it’s not just floods; increased temperatures will put millions of homes at increased risk of subsidence. Data from the British Geological Survey shows that approximately a million homes were at risk in 1990 and this rises to 2.4m in 2030 and 4m in 2070.
When you consider that the outstanding value of all residential mortgage loans in the UK was £1,562 billion at the end of quarter 1 this year, just looking at mortgages alone illustrates the challenge of climate change to the banking sector.
How can data, AI and Machine Learning help?
The good news is that it’s not all bleak. By harnessing data, AI and Machine Learning, there are significant opportunities for growth and development of ‘climate smart’ finance as society and the economy move to being both low carbon and more resilient to a changing climate.
The data to give a full picture of the risk from climate change comes from multiple sources including energy performance ratings, flood and subsidence models, asset data and more. As yet, most banks have not fully harnessed the power of this data, AI and ML to help them.
By bringing this data together, banks can use AI and ML to:
- give them a full picture of their mortgage exposure
- automate the population of energy performance ratings into their systems
- factor flood and subsidence risk into their lending and underwriting models
- ensure climate risk is accurately reflected in their Internal Capital Adequacy Assessment Process (ICAAP)
How else can banks use data, AI and ML to tackle climate change?
As well as allowing banks to understand their risk, harnessing their data also provides them with great possibilities and potential, both for themselves and for the benefit of wider society.
Climate change offers financial institutions numerous business opportunities and is expected to generate £1.5t worth of opportunity.
A recent survey carried out by YouGov across 24 countries showed that 85% of adults are willing to take personal action to combat environmental and sustainability issues in 2021.
A recent study showed that 76% of UK citizens used online and mobile banking in 2020. We can see that given their important role in society and the daily interactions they have with people through their current accounts, banks are uniquely placed to influence consumer behaviour.
Encouraging carbon neutral consumer choices
Banks are blessed with phenomenally rich datasets. This presents a unique opportunity for banks to leverage the power of transaction data to help both personal and corporate customers understand their carbon footprint, and use AI and ML in real time to analyse this data to give customers tailored suggestions or product offers to help them meaningfully reduce their impact.
Indeed, firms like Klarna and Mastercard are offering this to their customers using the independently verified Aland Index blended with global merchant category codes and customers’ transaction data.
The effects of this will ripple throughout the economy, as retail spend from carbon-conscious customers shifts towards less harmful options and businesses adjust to that change in demand.
Green mortgages are typically only offered to customers when they proactively present an EPC rating of A-C. If banks use AI to incorporate EPC data as part of the application process, these could be automatically offered at the time of lending.
This saves the customer the time and effort in having to look for the EPC themselves and instead allows banks to be proactive.
Attracting more mortgages of this type lessens the increased exposure risk associated with properties rated D-G. Add in the flooding data and banks can have a real-time view of any assets they are lending against.
Other new products, such as loans that charge lower interest rates to corporate borrowers who meet or outperform sustainability targets, are a powerful incentive and can create new business, lower credit risk and support the real economy’s transition to lower-carbon activity.
What are the challenges for banks?
Many banks have struggled to move from experimentation around select use cases to scaling AI technologies across the organization.
In our experience, there are six main reasons for this:
- Data quality: Data is often siloed and not often of the best quality. More use of open and external data (EPC ratings, weather data, climate models) is needed.
- Volume of data: While banks have an extraordinarily rich dataset to draw from,working with this volume of data and drawing meaningful conclusions can be challenging without the right skills and infrastructure.
- Skills: There needs to be a greater investment in staff skills; data needs to become part of everyone’s job in the same way that IT is embedded into every area. Banks can augment with external support while upskilling their own people.
- Security: Concerns about customer data privacy and security has led to an overly risk-averse approach e.g. low use of cloud technologies and not updating marketing and profiling preference.
- Technologies: Slow adoption of cloud and continued use of older technologies means the value in the data cannot be realised.
- Lack of strategic planning: Whether it’s not engaging with business leaders early enough, not planning ahead for scaling and adopting, or investing too much up front without a clear view of expected returns, a lack of strategic planning can lead to data, AI and ML projects not delivering to their full potential.
How can Kainos help?
At Kainos, we understand the huge benefits that data, AI and machine learning can bring to financial institutions. We also understand the responsibility that we all have to do everything we can to tackle climate change, and we are seeing ever-increasing demand from our customers to help them move towards a more sustainable way of operating.
And we have experience of tackling all of the challenges data, AI and ML projects can bring, from upskilling your own in-house teams to creating a strategic plan which accounts for scalability, full adoption and ROI.
We can help financial institutions to harness data to meet climate change can help banks manage and understand the risks and opportunities as well as giving them a useful, structured and strategic project on which to begin their data, AI and ML journey.
And we believe that doing this is essential, both for the future of the banking industry as well as for the benefit of wider society. As Sarah Breeden, the Bank of England’s executive sponsor for work on climate change, says: “Integrating climate and environmental data and analytics into decision-making will allow financial institutions to identify, measure, and manage the financial risks and opportunities from climate change, and so support the objective to ensure the financial system is resilient to these risks and supportive of the transition to Net Zero.”
Want to know more?
SustAInability: how AI can help us build a better future sees four industry-leading AI experts explore how the power of AI can build a sustainable future. Sign up for early access and hear it first.