background image

How can sustainable investors benefit from artificial intelligence?

author image

By Dirk Söhnholz

· 5 min read

Artificial intelligence (AI) can help to generate better, more timely, and cheaper information for sustainable investments. AI can also help to identify more sustainable investment opportunities and analyze them to a greater extent. Additionally, it can help investors to create portfolios which have a better fit with their specific sustainability requirements than most products offered today.

The tricky question is: how should AI-based sustainable investments be implemented? Here I try to provide some answers.

AI is not clearly defined. In this article I don’t differentiate between machine learning, deep learning, and AI.  For ease of understanding, I am not differentiating between environmental, social, and governance (ESG) and impact investing, or other sustainable investing approaches either.

Same return, lower risks with sustainable AI investing?

The main question is whether AI can help to improve returns for investors. In the past, a huge amount of brain and computer power and money were invested to try to generate higher returns than the markets. Many quantitative traditional investors and hedge funds with deep pockets tried in vain to consistently outperform passive benchmarks. I do not expect that AI will change that.

However, AI may help to reduce investment risks, especially sustainability risks. These risks can be measured with ESG ratings. These ratings are often based on a lot of data and on unstructured information from all kinds of formats including, for example, from corporate video conferences with stock analysts. With AI, it is easier to cover more investment product issuers and update ratings more frequently. ESG Book and are early suppliers of such AI-based ESG ratings. If AI helps to reduce investment risks, risk-adjusted investor returns can become better. Therefore, AI may not help to identify future sustainable outperformers, but it may be useful to avoid unsustainable underperformers.

Portfolio construction: is AI a threat for active or passive sustainable investments?

Discretionary active investing has typically underperformed systematic passive investments in the past. I do not expect that AI will change that. But AI can help to create sustainable starting portfolios. With Search4Stocks, for example, offers a free test service to create such starting portfolio universes.

AI can also be used to change investment rules for passive investments. If the current rule is to review the ESG ratings annually, with AI the review frequency may be increased, because the rating-relevant data may be available faster. Also, the investment universe may be expanded because AI can help to assemble reliable ESG data for companies, which due to limited human resources or high costs could not be used previously, e.g., small cap companies.

I doubt, nevertheless, that (overlay) risk management of portfolios can be significantly improved by AI. In the past, more frequent, or more complex risk signals to change portfolios have typically not led to higher portfolio performances.

AI can help to attract, educate, update, and retain clients

It seems obvious that AI can help to target marketing better to individual requirements including those of sustainable investors. Custom marketing may become cheaper and more convincing.

AI can probably also help to improve investor education. With AI support, it should become easier for investors to better understand the many different facets of sustainable investments. For example, this could be achieved by preparing good FAQs, which can be answered even with standard AI large language models (LLMs) such as Bing, ChatGPT, or Google Bard. These can be used for questions such as: is sustainable investing typically good or bad for investor performances? Product providers can train their LLMs with specific sustainable investment information to enhance their usefulness. Thus, potential investors can get better systematic comparisons of different offers such ESG-, SRI- or SDG-ETFs. A question for a proprietary AI can be: how does this specific offer help to improve the sustainability profile of my current portfolio?

Clients are known to stay longer invested in customized compared to standard (sustainable) investments. Overall, therefore, upfront AI-based sustainable portfolio customization is attractive for investors and providers alike.

Also, AI can help to generate more frequent and more detailed sustainable investment reporting for clients. That could help to increase sales and retain clients. But frequent information can be a sales risk too. Typically, there is negative information about every investment. If Investors receive additional (AI based) negative information about several of the companies they have selected for investment, they may abstain from trying to invest sustainably altogether. My recommendation for such cases is: try to invest as sustainably as you can. Even if this is not perfect, it is more sustainable than traditional investing.

AI for direct ESG indexing and self-customized sustainable investments

In my opinion, there is an even more attractive proposition than provider-based portfolio customization. I advocate self-customization for sustainable investments. Direct or custom ESG indexing (Custom ESG Indexing Can Challenge Popularity Of ETFs ( typically allows investors to customize rules-based sustainable start portfolios such as an index.

AI-based sustainability information can be used for self-customization, e.g., by providing up-to-date controversy information for the components of the start portfolio. Based on this information, even without detailed financial education it should be easy to deselect some stocks. Self-customized sustainable portfolios can be more ‘sticky’ than third-party customized offers and therefore even more attractive for providers.

AI may also be used to support individual bond or shareholder engagement with target companies. Adding custom AI-based sustainability services to a technology such as the one provided by Say Technologies may be the way to go in the future.

Conclusion:  Few downsides but many upsides of AI-based sustainable investments

Since there are not enough well-educated experts on sustainable investing, AI can help by filling some gaps. AI can probably also support internal functions such as compliance. However, traditional finance employees may be at risk from AI.

Another drawback is that AI applications, especially if they use image and video creation, can use up a lot of energy.

Even though there are some risks associated with AI-based sustainable investing, these are clearly outweighed by the benefits such as portfolio risk reduction, marketing, educational and reporting improvements, and better (self-)customization.  

This article is also published on Gitex Impact. illuminem Voices is a democratic space presenting the thoughts and opinions of leading Sustainability & Energy writers, their opinions do not necessarily represent those of illuminem.

Did you enjoy this illuminem voice? Support us by sharing this article!
author photo

About the author

Prof. Dr. Dirk Söhnholz is the CEO of Söhnholz ESG GmbH, previously Diversifikator GmbH, and a respected professor of Asset Management at the University of Leipzig. He has been a member of the DVFA Sustainable Investing Commission since 2018. Dirk’s prior roles include Managing Director at various divisions of the Feri Group and executive positions at Gerresheimer Glas AG and TelePassport GmbH. His academic work includes significant contributions to asset management, especially in ESG and SDG frameworks, with numerous publications since 1992.

Other illuminem Voices

Related Posts

You cannot miss it!

Weekly. Free. Your Top 10 Sustainability & Energy Posts.

You can unsubscribe at any time (read our privacy policy)