AI-driven ESG strategies for traders and investors – how to spot value in a sea of vanity metrics

ESG compliance is simultaneously becoming more important and more demanding. Data processing is chief among the demands, with organisations required to handle and process ever-growing volumes of data to publish for public scrutiny.

For investors interested in that data, there’s a three-pronged challenge to overcome. Firstly, not all companies collect and publish all of the relevant data. In fact, Professor Shawn Cole, who teaches Sustainable Investing at Harvard Business School^[1], believes that as few as 30 companies worldwide publish comprehensive ratings. Secondly, companies are incentivised to manipulate the data to give a more positive impression of their performance. Thirdly, and most importantly, no two companies collect and share their ESG data in the same way^[1].

AI is relieving some of that burden. Natural language processing (NLP) and machine learning models make it possible to extract richer and more meaningful insights and data from other unstructured sources, like emissions reporting, weather station readings and satellite imagery.

As well as lighten the load on ESG teams, AI is surfacing useful strategies for traders and investors to take advantage of.

Climate risk modelling for companies - telling new stories with old data

AI has the capacity to more clearly identify climate risks compared to traditional methods, using real-time climate data, historical climate data and climate forecasts. It’s already happening and it’s a $31bn market^[2].

Climate AI technology is still in its infancy. Smart investors might be attracted to companies partnering with the early-stage disrupters.

Among the privately held startups in this space, Jupiter^[2] stands out for its capacity to serve sectors that will always need and rely on climate modelling. Those include real estate, agriculture, defence and logistics, and clients include AstraZeneca, CLS Holdings, Entergy and Hawaiian Electric^[3]. Their technology can support compliance teams with existing regulations, disclosures and a general adoption of stronger climate resilience practices. The company raised $54 million in a 2021 Series C round, taking its total, as of late 2024, to $84 million.

Another notable startup, ClimateAi^[2], focuses more squarely on climate risk mitigation. Its platform delivers high-resolution risk assessments and forecasts for things like crop yield and supply chain disruption, supporting clients like Dole and Nuveen^[4].

Carbon credit trading and greenhouse gas reductions are all well and good for large corporations. But what about actually reducing or preventing climate damage? This is the less talked-about but more important aspect of Net Zero because it’s extremely hard to measure the cost-effectiveness of the investment.

ClimateAi and tech company NEC^[5] developed a climate adaptation model for African farmers by determining the return on investment (ROI) for different projects, like rice and cocoa production.

Established companies have been quick off the mark too, bringing to market new ways to manage their own climate risks. IBM’s Watson can already provide advanced climate risk analysis and forecasts.

Credit ratings company Moody’s bought climate data firm Four Twenty Seven in 2019 and has spent the time since incorporating climate risk tools into its offering to large financial services businesses. Their ESG Solutions now offer climate risk scoring for a variety of asset classes^[3].

Pricing signals in carbon markets - adding new meaning to Net Zero ambitions

Carbon markets allocate a price to carbon emissions. These systems underpin the concept of carbon credits and allow countries, institutions and companies to trade carbon credits to meet their climate targets. They also create financial incentives for reducing carbon emissions.

The cost of emissions is ultimately paid by the consumer, either directly through increased prices at the point of sale, or indirectly through things like increased health care costs, and growing insurance premiums owing to the cost of droughts and floods.

The global economy has what you might call a ‘carbon budget’, with the aim ultimately being for the world economy to reach Net Zero, which means removing the same amount of carbon from the atmosphere as we produce.

To achieve an efficient transition to a Net Zero economy, markets need reliable and easy-to-understand energy benchmarks. Carbon markets can provide powerful signals to investors in service of this broader goal. But whenever there’s a collective economic incentive to reach a goal, there’s an incentive to under- or over-report.

This might be where AI comes in. To illustrate, let’s take an example from the past. Imagine back in 2014, a risk management team had created an AI platform for standardising and converting emissions data reported by major car manufacturers.

Now imagine they’d applied this data across all of the major manufacturers for the previous ten years. They’d have eventually spotted an outlier in Volkswagen, which was found to have been underreporting emissions from diesel vehicles.

Now imagine that ahead of the inevitable regulatory sanctions, that same risk management team had advised clients to short VW stock based on the expectation that the price would fall when the scandal broke and sanctions hit. This is how AI use in ESG can generate powerful tradeable insights.

Is ESG a reliable predictor of future performance?

Traders and investors can and do use AI to analyse ESG data, financial performance and economic signals to predict future ESG risks and potential opportunities. Potential tradeable insights might include companies at increased risk of supply chain disruption, companies with poor adaptation and possible mitigation strategies.

Climate risk modelling is another area where traders and investors may identify opportunities; extreme weather events, disruption to resource availability and even drops or spikes in climate-related consumer demand for certain products and services.

By using AI to analyse the ESG posture of a company, group of companies or even entire sectors, savvy traders and investors can identify undervalued stocks as well as stocks that may be about to experience volatility.

If you’d get value from an AI tool that can deliver institutional-level insights and curated analysis, check out Tradu’s Analyst AI. Users benefit from sentiment alerts and insights tailored to their own trading activity.

The bottom line

ESG has evolved. What was once seen chiefly as a set of metrics for marketing departments to lean on when making claims of being ‘ethical’, it’s now a popular but potentially unreliable predictor of a company’s future financial performance. AI is making it more reliable, helping traders and investors make sense of vast swathes of complex, unstructured data to arrive at smart decisions, build automations and identify risks.