Artificial intelligence bridges the gap between big data and sustainable investing

BlogBig data and artificial intelligence, SustainabilityAccounting & FinanceDigital Transformation
Kristof Stouthuysen

By Kristof Stouthuysen

Professor of Management Accounting & Digital Finance

22 March 2021
iStock-929942360

Sustainable investing is booming. In his 2021 letter to CEOs, Larry Fink, the CEO of Blackrock – the world’s largest investment firm – announced that his firm will immediately stop investing in companies that ‘present a high sustainability risk’ and that it will make sustainability an integral part of its investment strategy. Fink is not alone in his view that sustainability matters and should become the lens through which investors evaluate their investments. Despite the Covid-19 pandemic, inflows in mutual funds and ETFs that invest in sustainable assets reached 326 billion dollars in the 12 months to November 2020, an increase of 97% compared to the same period in the prior year.

However, sustainable investing does not come without challenges. An important one is how to measure sustainability. Whereas, traditionally, investors evaluate firms based on well-defined and understood measures of financial health and stability – such as ROA, leverage ratio, Tobin’s Q, etc. – for which high-quality data is widely available, assessing the environmental, social, and governance (ESG) performance of firms can be a daunting task. As opposed to financial reporting, sustainability reporting is only in its infancy and, despite many reporting guidelines and standards already available, it is often not regulated.

Therefore, many investors rely on ESG data providers for their sustainability data and analytics. However, these ‘rating agencies’ are not without flaws either. Evidence exists that the ratings of firms tend to diverge across different raters and that there might be a rater bias. Moreover, ESG providers often rely on data obtained from surveys and that are reported by the companies themselves, which makes these data highly subjective and susceptible to greenwashing.

The lack of high-quality sustainability data might keep investors from moving into the sustainable investing field. A recent Blackrock study shows that 53% of their respondents cited the ‘poor quality or availability of ESG data and analytics’ as their biggest barrier to adopting sustainable investing.

Nevertheless, in a world where more than 2.5 quintillion bytes of data are generated every day, large amounts of ESG information are already there (to clarify: a quintillion has 18 zeros!). The problem is that much of this data is unstructured, which makes eliciting meaningful information challenging. Luckily, artificial intelligence (AI) can help. AI has the ability to analyse (large amounts of) unstructured data, allowing investors to evaluate and analyse ESG data that are more granular and objective.

AI and alternative data

Think about (social) media. Every day, a tremendous amount of data is generated through social networks, news websites, blogs, and many other sources. Investors can use natural language processing (NLP) – a field of AI specifically aimed at understanding and deriving meaning from human language – in combination with big data analytics to scan large amounts of web sources in almost real-time. In doing so, investors can discover ESG risks and opportunities at an early stage. For example, the data might show that a retailer is investing in autonomous delivery to reduce its carbon footprint, but they might also show that one of its suppliers has consistently been violating human rights. TruValue Labs, a San Francisco based company, already uses a combination of NLP and big data analytics to provide ESG insights into over 16,000 securities.

Dutch investment firm NN Investment Partners has also started using natural language processing, big data analytics, and machine learning (ML). But instead of analysing written text, the firm analyses speech in company conference calls. This allows them to better capture a manager’s attitude towards ESG.

This process of using AI to analyse content to capture the beliefs and sentiments towards a firm’s ESG performance is called ‘sentiment analysis’, and it has major potential for applications in sustainable investing.

But the use of AI in sustainable investing goes much further. The variety of data sources investors can leverage through AI is huge.

UK start up Cervest, for example, combines satellite imagery and other environmental observational data with AI and ML to model climate risk. Investors can use this to effectively incorporate climate risks in making their investment decision. Satellite imagery can also be used to model an organisation’s direct environmental impact by evaluating deforestation and river pollution.

A firm’s societal impact, on the other hand, can be determined using telecommunications and demographic data. Using data aggregated from telecommunication providers, Distilled Analytics, a Boston based company, was able to show the increase in local GDP associated with firm investment.

A complementary benefit of AI is that it not only enables the analysis of this rich amount of data, but it can also do it fast. After conducting one year of analysis on the automotive sector, TruValue Labs found that the same analysis would have taken a human analyst approximately 6 years.

AI for sustainable private equity investment

The data quality problem is even worse for private equity investors, as businesses are smaller and have fewer resources to conduct sustainability performance evaluations.

Since AI models allow the use of alternative data sources, they are also suited for private corporate sustainability assessment. After all, private firms get media coverage too, and satellites do not discriminate between listed and non-listed firms. The use of AI and big data analytics will thus create more opportunities for private equity investors to invest with impact.

Overall, AI and big data analytics are powerful tools that aid investors, both public and private, in making their investment decisions. They are fast, provide an almost real-time analysis, and help elicit more objective information by combining multiple data sources. Using AI thus decreases the uncertainty inherent to sustainable investing.

By buying a minority stake in Clarity AI, a tech platform that uses big data and machine learning to provide sustainability insights, Blackrock recognizes the importance of AI in helping them deliver on their promise of becoming more sustainable. As the leader in their industry, they set an important example and drive the transition towards AI for sustainability. Both the technology and the data are already there, so it is our belief that many other investors will quickly follow.

Get in touch!

Kristof Stouthuysen

Kristof Stouthuysen

Professor of Management Accounting & Digital Finance