Alternative ESG data provider GaiaLens has harnessed the AI technology behind headline-grabbing app ChatGPT to revamp its news sentiment feed to investors and asset managers.
UK-based GaiaLens, which was formed three years ago, said the use of large language technology has helped the company turbo-boost its Momentum product, which scrapes trusted news sources for sentiment on companies’ ESG performance. The enhanced Momentum AI (MAI) service, which is offered predominantly to smaller asset and investment managers, can give deeper and real-time insights, identify more themes and patterns within the data and give information on how scores were reached, co-founder Sebastien Kirk said.
“It’s been a game changer in that you can rely on the sentiment analysis with an explanation in real time for more coverage, more themes and because it’s built on algorithms and machine learning AI it’s extremely scalable so you can scale it to any form of purpose,” Kirk told ESG Insight.
Deeper Insights
GaiaLens was created to build an ESG tool for a London- and Philadelphia-based investment management firm. The aim was to make ESG ratings “explainable, rather than just a credit rating-style” proposition, Kirk said.
From that, the company created Momentum, a news sentiment data scanner that differentiates itself from other similar services by providing data on specific topics within ESG themes. Each news item is tagged and grouped according to other parameters, such as whether they contain references to scandals or controversies.
Kirk said this approach offers greater flexibility to investment managers.
“My view on ESG or your view on ESG or an asset manager’s view are all going to be different,” he said. “Momentum lets investment managers examine the aim of their holdings, what are they trying to achieve – is that to achieve carbon neutrality or to have a social fund, and so on?”
Improved by AI
MAI brings improvements to the Momentum product. It has accelerated the scraping process and enables a greater number of factors to be examined. But the key benefit it brings is the ability to add explainers to the scores it produces, providing greater transparency on how those metrics were achieved.
“What that has meant is we can give explainable news sentiment in real time,” Kirk said. “AI has made it richer, easier to understand. That’s important because some of the reservations about sentiment to begin with was connected to what those scores meant, and could we prove why we arrived at them.”
GaiaLens is tapping into mushrooming demand for alternative data, and which has created a global market that’s grown from less than US$250 million in 2016 to $1.7bn in 2020, according to a Statista survey of buy-side firms. A separate survey by major US law firm Lowenstein Sandler estimated that the total market value would be in the region of $150bn in the next few years.
While alternative data has been traditionally used by hedge funds and private equity firms, the growing demand for granular corporate and non-corporate data for use in ESG functions has seen its use grow within the mainstream financial sector. Companies such as Alygne, Orenda and QuantCubehave emerged in recent years to offer sustainability and impact information obtained from a range of sources including social media and satellite images.
More Use Cases
GaiaLens scrapes data from vetted news outlets and buys in supplementary datasets from large environmental content providers. It also takes data on niche themes from NGOs and obtains governance scores using such data pools as employee reviews and board composition information.
All data is processed through proprietary validation algorithms, Kirk said.
Clients are using the service to identify where they may need to engage with investee companies or they are incorporating its scores into their quant models.
While the Momentum product is a key part of GaiaLens’ business, the company has also branched out into data vending services, controversies alerts and regulatory reporting support, which are helping investors find red flags or positive signals on individual investee companies or portfolios.
“It started with clients just wanting sentiment data because that was the new thing, but now it’s being used for much, much more,” Kirk said, adding that the AI-boosted iteration of Momentum has added even greater usability.
“Now we can basically ask MAI anything for whatever purpose the end user wants.”
Subscribe to our newsletter
