With the growing amount of data related to Environmental, Social and Governance (ESG) investing criteria becoming available, technology will undeniably have a huge part to play in helping the industry navigate and manage data in an efficient and insightful manner.
Advances in Artificial Intelligence (AI) have revolutionised how companies work with data. AI’s ability to create insights from massive volumes of unstructured data and automate complex tasks at incredible speeds has allowed us to achieve a level of efficiency that would otherwise be impossible for humans to achieve on their own. Here, we explore the transformative potential of AI to help socially conscious investors evaluate potential investments, and its impact on ESG investing going forward.
AI’s ability to create insights from massive volumes of unstructured data and automate complex tasks at incredible speeds has allowed us to achieve a level of efficiency that would otherwise be impossible for humans to achieve on their own.
As financial institutions increasingly pledge commitment to several climate targets and principles such as the United Nations Principles for Responsible Investment and the Equator Principles, AI systems will play a crucial role in generating data and insights required to integrate ESG factors into investment decisions. AI’s ability to process massive amounts of ESG-related data enables investors to identify information missed by traditional research methodologies. It can also provide actionable insights and equip investors with the information they need to make investment decisions that are aligned to their own ESG values and the respective principles.
The role of AI in integrating ESG data in asset management covers qualitative analysis, quantitative analysis, investment decisions and assessment1. Unlike conventional financial data, accounting for environmental, social and governance factors requires more analysis and processing, given its discretionary and sometimes unquantifiable nature. Building on frameworks like the Sustainability Accounting Standards Board, AI can be used to capture ESG data and create insights easily understood by investors. This includes correlating financial data with non-financial factors and adjusting the financial forecast for ESG-themed portfolios accordingly. Over time, these ESG-based models will improve its algorithm to provide more accurate and socially responsible investment recommendations, supporting ESG strategies of various institutions.
The role of AI in integrating ESG data in asset management covers qualitative analysis, quantitative analysis, investment decisions and assessment.
1 https://financialservicesblog.accenture.com/growth-markets/purpose-and-profit-ai-for-esg-investing
The power of AI presents a huge potential in the ESG investing space with its sentiment analysis algorithms. Sentiment analysis programmes can be trained to analyse the tone of a conversation or news article, allowing users of the data to derive deep insights without having to digest all the information2. For example, a sentiment analysis programme trained to read the transcripts of a company’s quarterly call could use natural language processing to easily identify parts of the conversation related to ESG topics, and then infer from the words used how committed a company is to advocating its ESG values or mitigating its ESG risks. While sentiment analysis is not completely new to some, the application to ESG could become an extremely useful tool for investors to navigate the expanse of ESG data and attain timely alerts of any material ESG issues for an organisation.
One such tool includes using AI as a monitoring tool for portfolios, where an indication of negative sentiments may highlight ESG issues or red flags of the underlying holding. This may reduce the risk of greenwashing as investors are able to actively monitor companies’ activities in an extremely efficient manner.
While the sentiment analysis programmes will get progressively refined over time with clearer definitions and taxonomies of ESG factors, there are some limitations in sentiment analysis programmes of today. Some of these limitations include foreign language processing capabilities which affect accuracy of the programme and scope of news and media sources covered. For example, the translation of news sources – from local language to English – may alter the connotations of certain words. This will certainly have an impact on the accuracy of the processed sentiments of the local news source, the timeliness of data and value of insights to its consumers.
2 https://www.spglobal.com/en/research-insights/articles/how-can-ai-help-esg-investing
Industry participants are of the view that the ESG investing landscape could be significantly improved by putting in place several key requirements, with the most important being the standardisation of definitions and reporting requirements3. Standardisation means creating a consistent, comparable and financially meaningful way of communicating ESG risks and opportunities to stakeholders.
Data quality and availability are of vital importance when it comes to standardisation and the increase in reporting requirements. This is where AI comes to the fore. Through machine learning and natural language processing based on the frameworks in place, we will be able to identify, sort and filter public data like annual reports and climate disclosures into financially meaningful ESG data.
3 https://www.ft.com/content/4d7accf7-5431-4ebb-a528-87db3cca1eb7
A practical use case would be identifying parts of company disclosures that relate to climate change using the Task Force on Climate-related Financial Disclosures framework and determining if it’s useful, accurate and consistent with other public data. This will vastly increase the scope of ESG data covered, considering the ability to comb through a multitude of public data across the globe. Additionally, it can improve the accuracy of data given its ability to validate, update and define relevance of such fast-moving data.
Standardisation means creating a consistent, comparable and financially meaningful way of communicating ESG risks and opportunities to stakeholders.
Going forward, AI-assisted reporting will be more relevant than ever with multiple industry initiatives afoot that aim to standardise ESG reporting criteria and to prevent ‘greenwashing’. One notable example is the EU Taxonomy – part of the wider EU sustainable finance initiative – which establishes a classification system for economic activity according to EU climate action goals. Under the EU Taxonomy Regulation financial market participants will have to complete their first set of Taxonomy disclosures by 31 December 2021 and companies will need to begin making disclosures during 2022. Besides the EU, other countries that are also developing taxonomies to encourage green investment include China, Malaysia, and Singapore. Read about how the EU Taxonomy helps define a green and sustainable future and its implications for investors and issuers globally.
Going forward, AI-assisted reporting will be more relevant than ever with multiple industry initiatives afoot that aim to standardise ESG reporting criteria and to prevent ‘greenwashing’.
The sustainability agenda is core to our business at Standard Chartered across our markets in Asia, Africa and the Middle East. We are committed to unlocking the full potential of technology and innovation to serve our clients and at the same advancing the global sustainability agenda.
Under the EU Taxonomy Regulation financial market participants will have to complete their first set of Taxonomy disclosures by 31 December 2021 and companies will need to begin making disclosures during 2022.
