Unlocking ESG Data Transparency with Artificial Intelligence

Banks and financial institutions are in the frontline of internal and external environmental, social and governance (ESG) risk management. Artificial intelligence (AI) is an invaluable ally to help interpret the relevant data and information.

The pressure on corporations to live up to ESG principles has risen dramatically over the last decade. Employees, investors, regulators and suppliers scrutinise a business statements and actions for evidence of doing the right thing in a range of areas that include climate change, human rights and anti-corruption. When a corporation is also a bank or financial institution, that only ramps up the attention. Naturally, people will be concerned about a company’s conduct.

However, managing ESG risk is not all one-sided for a bank or financial institution. It needs to answer two questions:

1. Internal – what are the ESG risks in its business or businesses?

2. External – what ESG risks do its clients represent?

Coming up with answers to these questions is made more difficult by the vast amount of data and information that organisations such as rating agencies, multilateral development banks, research houses and others produce about the entities that are living up to best practices for ESG risk management and how they are doing so.

Extracting the nuggets

Financial institutions are taking different approaches to the problem of dealing with this task. Some are hiring for specific in-house roles; others are adding ESG risk management to an existing individual or team’s role while a third group is relying on external advisers to deal with the issues. But whatever way they choose, they are likely to find that traditional ways of evaluating a company such as examining financial records, tracking stock markets or poring over analysts’ reports and regulatory announcements to help them uncover ESG risk will not give them the understanding they require. Too much information, the majority of which may be unstructured, is coming at them from many more sources than before, such as websites, social media and electronic communication of various kinds.

This is why the use of AI, encompassing a range of tools such as natural language processing– one use of which, to use the jargon, is sentiment analysis¹ – deep learning neural networks that mimic the operations of the human brain, and sophisticated predicti ve analytics², has become critical for banks and financial institutions. Whereas humans might take days or even weeks to sift through all relevant information, AI tools can do so much quicker and with far less effort. Furthermore, AI can learn from the experience of trawling through this information and improve its efficacy over time. These approaches can help banks and financial institutions deduce where the ESG risk resides within their own business and then carry out tasks like customer onboarding that involve assessing the potential harm of doing business with an entity or individual they may not know much about.

“Technologies, often the domain of Fintech companies, hedge funds and the quant teams on bank’s trading floors, rely upon accurate and transparent bank data.”

Fintech to the rescue

Technologies, often the domain of Fintech companies, hedge funds and the quant teams on bank’s trading floors, rely upon accurate and transparent bank data. With the power of AI doubling about every three and a half months,³ its ability to extract important detail that conventional technologies, or indeed human beings alone miss, is essential for any organisation faced with making decisions based on the interpretation of data and information. 

The Fintech industry has many examples of companies that have developed technologies for the three recognised elements of ESG data and information:⁴

— Rating – firms that rate companies based on ESG criteria.

— Reporting – platforms that enable companies to measure and report their ESG performance.

— Data platforms – collecting data based on various ESG parameters for companies and funds, even if some pioneering blockchain-based Fintech companies have questions to answer about the decidedly non-ESG carbon footprint they leave behind.

AI’s ability to look through mountains of facts and figures and uncover genuine value or identify potential harm in onboarding a new customer underscores its advantages over more traditional research methods. An investor with ESG policy objectives finds it difficult to determine accurately if a company they are looking to invest in operates consistently with the investor’s values and principles. When assessing a listed company, they need to get to the underlying information about a stock to do that; after a financial institution has used the speed, depth and breadth of AI to highlight possible onboarding risks, it can then decide to avoid the issue and refuse to onboard a prospective new customer, or they can view onboarding as a long-term opportunity to influence the agenda of prospective clients on ESG issues. That applies whether the bank or financial institution is based in Hong Kong, Singapore, London or New York. 

The world’s biggest investor and ESG

Asset managers devote a considerable amount of time and resources to understanding the ESG risks involved in having individuals or organisations as clients or investing in the stock market. Norges Bank Investment Management (NBIM) is better known as Norway’s sovereign wealth fund. As the world’s biggest investor, owning shares in 9,000 companies and 1.5% of stocks globally, it takes an active role in finding out all it can about the companies it invests in.

In its responsible investment report for 2021, published in March 2022, NBIM said it had “developed better methods for identifying companies with high risk relating to the environment, social conditions, and corporate governance.”[5] It is unlikely that these better methods did not involve some form of AI. 

NBIM assesses ESG risks as part of its investment management process and manages them according to its principles for responsible investment management. It expects the companies it invests in to address global challenges in their corporate governance under eight separate headings that “largely coincide”: children’s rights, climate change, water management, human rights, tax and transparency, anti-corruption, ocean sustainability, biodiversity and ecosystems. It also publishes its criteria for excluding companies from its portfolio.

“Our motivation for responsible investment is to achieve the highest possible return with moderate risk. Companies’ activities have a considerable impact on society and the environment around them. Over time, this could affect their profitability and so the fund’s return. We therefore consider both governance and sustainability issues, and publish clear expectations of companies in the portfolio.”

AI comes into its own in situations where different agencies have rated a company’s ESG risk and the financial institution has to choose the most appropriate one for its objectives. It may decide not to go with the same agency’s rating for each of those objectives because there could be something it does not like about its methodology, or it may be inappropriate for the industry sector. AI does cross-analysis well because it can consider the external factors. It can also take into account the relative success over time of different ratings with regard to the objectives.

Robust reporting and reporters

As investors, regulators and the public take ESG more and more seriously, banks and financial institutions need more comprehensive, rigorous and independent reporting to identify and interpret ESG risk for themselves and their clients. AI is a powerful tool to help them do this given the many different sources of data and information that now exist because of its power to impose transparency on ESG ratings, rankings and indices, and to tell the story behind the underlying data, rather than the story of the rating itself.

“AI comes into its own in situations where different agencies have rated a company’s ESG risk and the financial institution has to choose the most appropriate one for its objectives”

To address any shortage in talent and skills for new ESG risk management roles, including how to use AI effectively, banks and financial institutions could partner with, or hire more from, technology companies, fintech start-ups and other organisations. For example, to understand the environmental risks of ESG, they could embrace a much broader range of participants in the market such as oil companies, environmental agencies, nongovernmental organisations, meteorologists and geographers. Generally, however, this is not something that banks are set up to do. Understanding the real drivers of ESG risk is not the natural territory of a bank or financial institution. Leaning on the expertise of specialists can change that landscape.

This article was originally published in the March/April 2022 edition of Banking Today, the official magazine of the Hong Kong Institute of Bankers.


Open chat
thank you for contacting us, for more information
please chat