Continuing our journey to becoming a climate positive bank

Context

Like other financial institutions, the emissions produced by our lending activities far exceed those produced by our own operations. This is inevitable for a bank like Atom. Our business model relies on supplying finance to tens of thousands of households and small businesses. Until the wider economy is decarbonised, these firms and households will continue to use fossil fuels to heat their premises and power their livelihoods.

Even though parts of the economy and some customers may be slow to adopt low-carbon solutions, lenders like Atom can still do a lot. We can raise awareness, offer more options to customers and change our policies, pricing, and products to promote the shift to a zero-carbon economy.

Since 2019, Atom has had a "Climate Change Sector" list that prohibits lending to high-risk environmental sectors. This list, which includes industries like oil, gas, mining, and primary plastics, is updated twice a year. The review focuses on energy efficiency, greenhouse gas emissions and waste management within each sector.

Data limitations

So where do we get started in tackling this problem? Step 1 of our strategy is always to ‘Measure’ — to identify and quantify the emissions we are responsible for. But when it comes to financed emissions, data quality and availability continues to pose a significant challenge.

Across many sectors, it is broadly accepted that there are significant limitations impacting financed emissions calculations. One of the most prominent is the lack of reported carbon emissions and availability of actual energy usage data. Recognising these limitations, the PCAF uses a data quality scoring system to show that carbon reporting still relies on estimates. This helps everyone understand the journey ahead to get better data.

These data quality scores range from 1 (most accurate) to 5 (least accurate). For us, the most important measures of financed emissions stem from the emissions associated with power, heating and cooling associated with the buildings that are secured against our residential and commercial mortgage books, and a fuller understanding of the emissions from all activities undertaken by the SMEs that are funded through our partnership with Funding Circle.

Metered data on real energy use is required to achieve data quality scores of 1 or 2. To date, these measurements have not been routinely available in the UK and, generally, Energy Performance Certificates (EPCs) are used as a proxy for estimating energy use. This has been a significant limitation to the overall data accuracy that we, and other banks, can operate at.

We've done a significant amount of work over the last year to improve our options and the quality of our data. This will help us better measure the size of the challenge we face and share accurate information about our lending emissions with customers, regulators and other stakeholders.

We have recently published the results of a collaborative trial with Experian . It revealed that we — and other UK banks — are likely over-reporting emissions from residential mortgage lending by as much as 50%. This study, undertaken with our partners at Durham University’s School of Mathematical Sciences and supporting findings from University College London’s (UCL) Energy Institute, found minimal variations in actual carbon emissions between properties of very different EPC ratings, suggesting a significant issue with the current EPC assessment process.

EPC ratings were not designed to be the basis for calculating carbon emissions but have become common currency for estimating financed emissions and for designing and selling what are sometimes called “green mortgages”. Our analysis suggests that there needs to be an industry-wide reappraisal of the approach to both reporting and to claims that a “good” EPC necessarily equates to lower carbon emissions.

By using actual meter readings instead of EPC estimates, we found a huge 50% reduction in emissions from 1,100 properties. While rethinking how we use EPCs might be inconvenient, this difference is too big to ignore. It suggests that the path to net zero for UK residential properties could be significantly less costly if we focus on the energy sources used to heat, cool and power homes.

Consequently Atom, alongside Experian and industry group Bankers for Net Zero (B4NZ), is urging the energy industry, other banks, the government and the Bank of England to share and start using actual emissions data to gain a clearer understanding of the true cost of the transition to net zero. We are calling for accelerated EPC reforms, including a shift from estimated to actual energy performance data, and a greater focus on providing low-carbon power and heat to all.

FY25 reporting

For the second year running, we are proud to disclose the financed emissions from all assets on our balance sheet. This includes our unsecured loans, which come from our partner Funding Circle, and our buy-to-let mortgages, which we offer in partnership with Landbay. We hope our transparency will encourage others to do the same and help them benchmark their own financed emissions.

To ensure our calculations are as accurate as possible, we’ve worked with multiple partners who specialise in calculating financed emissions.

For FY25, we have still mainly used EPCs to gauge energy use. However, we are actively exploring options for more accurate measurements for next year. The market doesn't currently offer our preferred option of routine reporting using either permissioned or anonymised meter data.

We are committed to working with other innovators and new companies that can solve this challenge for us and other lenders. This problem won't be solved by any one firm working alone. We are dedicated to collaborating with both the banking and property sectors to increase understanding, transparency and action on data quality and emissions reduction.

Data quality for residential and commercial mortgages

We do not offer 100% finance for any mortgage. We limit our residential lending to a maximum of 95% of the property value and our commercial mortgages are typically below 75% of the property value, often much less.

We only calculate and report the emissions for the portion of the property we finance. As homeowners pay off their mortgage, they become responsible for a greater share of the property's emissions, proportional to their equity.

Additionally, for these calculations, the property value is fixed at the time the loan was originated. This is a widely accepted practice in reporting protocols and it prevents the bank from benefiting from a reduction in carbon emissions that results from rising property values.

Based on this method, Scope 1 and Scope 2 emissions from a property's energy use (those associated with power, heat, coolant etc) are attributed to Atom and other lenders. The amount of emissions attributed to a lender is determined by the ratio of the outstanding loan to the property's original value when the mortgage was issued.

Given that we are currently unable to access metered energy use information, our calculations for our residential mortgage portfolios correspond very closely to a data quality (PCAF) score of 3. This is consistent with the disclosures of other mortgage lenders and provides a foundation for improvement in future years.

For our commercial mortgages, the score is closer to 3.1. There are fewer data points available for these properties — we know less about what is happening in individual buildings because one small business is very different from another. Consequently, the data quality score is lower. There is a lot of valuable work to be done to increase the quality of information and disclosure of emissions from SME loan books and we are actively engaged in discussions on this topic, with partners such as B4NZ and the British Business Bank, and a peer group of other engaged lenders.

Data quality for unsecured business loans

For unsecured loans, emissions are attributed based on the ratio of the outstanding loan to the SME's financing needs. Due to a lack of available data for small and medium-sized businesses (SMEs), we chose an economic activity-based approach to measure the emissions from this part of our balance sheet.

This method uses sector-specific emissions factors that calculate the amount of carbon emissions per unit of revenue. While we used an approved methodology from the PCAF, this approach has known inaccuracies. Therefore, the numbers disclosed in this section should be treated with caution.

Our unsecured loan portfolio received an average PCAF data quality score of just 4.0. While this is a starting point, this level of estimation, and the uncertainty it creates for making decisions about pricing or policy, is not acceptable long-term. We will continue to investigate how to improve data quality and then take steps to reduce the carbon intensity of this type of lending.

It’s important to note that carbon emissions from unsecured loans are significantly higher per pound than those from secured loans. This is because unsecured loans finance an entire business, and the calculations include Scope 1, 2, and 3 emissions. In contrast, secured loans only account for the Scope 1 and 2 emissions directly related to a specific property.