Can the tech behind financial services be more sustainable?

24 Feb 2022

James Chenery. Image: Verne Global

Verne Global’s James Chenery explores how the level of tech needed in the financial industry may be causing challenges when it comes to meeting ESG goals.

When it comes to environmental, social and governance goals, particularly where sustainability is concerned, technology can be a double-edged sword. While new and emerging tech can help create solutions to sustainability challenges, innovations can also be quite power hungry in terms of energy usage.

While there has been much focus on the likes of energy providers and data centres, other services such as fintechs and banks can also play a central role in decarbonising the economy. Investing in sustainable projects is a step in the right direction, but James Chenery, sales director at Verne Global, said financial service companies also need to look at their own back-office operations.

“What makes things difficult is that the financial industry is it is a significant user of energy. Financial markets are all electronic, while financial services increasingly use AI for all sorts of purposes, ranging from risk management and fraud detection to asset portfolio optimisation and customer communication,” he told SiliconRepublic.com.

“For example, quantitative trading leverages AI to identify patterns in vast data sets and generate insights and predictions to inform strategic trades. For global banking, McKinsey estimates that AI technologies could potentially deliver up to $1trn of additional value each year.”

To support the modelling required to generate these insights, high-density computing power and vast amounts of data are required. According to Chenery, even simple data lakes are energy intensive because when servers are idling, they can still draw up to half their power capacity.

“In order to be compatible with sustainability efforts, financial services have to find ways to minimise the impact of their operations, particularly when it comes to rapidly advancing technologies such as AI.”

Minimising environmental impact

While it may be common knowledge that complex AI models and data lakes are energy intensive, being able to measure and reduce the impact of this technology is easier said than done.

However, Chenery pointed to the Greenhouse Gas Protocol (GHGP), a set of standards for accounting and reporting carbon emissions across the entire value chain. This would allow financial services to gain an accurate measure of the impact of the technology they are using.

“AI models in financial services, as with other industries, are only going to get larger as they expand and draw on more data to become more accurate,” he said.

“The high energy costs of leveraging advanced technologies such as AI means that some of the highest carbon costs will be occurring outside of the company’s four walls, ie in the data centre. To measure data centre efficiency, financial institutions can also use power usage effectiveness (PUE), a globally recognised indicator of how efficient a data centre is.”

Using measures such as PUE comes with challenges as it can be an oversimplification due to the type of power being used. For example, even though two data centres have the same PUE number, one might be fed by power from a grid reliant on fossil fuels while the other might be powered by 100pc renewable energy sources.

“Additionally, adopting new technology means retiring legacy equipment and applications. The financial services sector has been around a long time so, for example, mainframes still form a key part of many banks’ infrastructure,” said Chenery.

“While there are generally plans to retire these legacy systems, primarily to meet ESG objectives, many data centres cannot handle the higher density of new equipment. Alongside efficiency and renewable energy, financial institutions will want to ensure there is an abundance of stable power available to keep their high-density compute running optimally.”

What leaders need to think about

Finding the balance between leaning into the technology that can make processes more efficient and ensuring that same technology doesn’t use up too much energy can be a challenge for leaders.

Chenery’s advice is to look beyond the jargon and think about renewable versus green when selecting a data centre.

“Green can mean carbon offsetting to meet green objectives. While this is a step in the right direction, the data centre will likely still be using power from a grid reliant on fossil fuels, meaning the financial institution or data centre itself is then having to offset this use of fossil fuels in other areas,” he said.

“What’s more, as the GHGP includes all emissions generated across the entire value chain, financial institutions should think about leveraging their significant influence on companies delivering ancillary services to the data centre such as suppliers of mechanical and engineering infrastructure, servers, switches, cabling and telecommunications.”

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.

Jenny Darmody is the editor of Silicon Republic

editorial@siliconrepublic.com