The valuation of companies is still mostly based on financial valuation. Environmental, social, and governance (ESG) criteria are less taken into account because they are too disparate. The start-up ValueCo, incubated at IMT Starter and presented at VivaTech 2023, offers a database for asset managers to better integrate these ESG criteria into their investment decisions.
Asset managers, who invest in a company through shares or bonds, base their decisions primarily on financial analysis. However, taking into account environmental, social and governance criteria — called ESG criteria — for extra-financial analyses is a complementary approach that makes sense in the context of responsible and sustainable investments.
However, the analysis of ESG criteria involves the capture of a large quantity of data that is not very standardized, for which investors generally have neither the time, nor the expertise, nor the internal resources. The latter will therefore contact data aggregators and rating agencies, such as MSCI or S&P Sustainable1, to act as an intermediary between them and listed companies. These agencies aggregate the thousands of data and indicators available on companies to generate aggregated scores, which they sell to asset managers in particular to help them in their investment decisions. These scores also make it possible to assess regulatory compliance or the shareholder commitment of a company, to monitor risks or to communicate.
The problem: these ratings are very different from one agency to another. While the ESG score is a rational indicator, the methodologies for building it are opaque and extremely variable. For example, one agency will produce a score out of 100, while another will use a letter rating. As investors generally only have this information to do their extra-financial analysis, they will buy several ESG scores, and based on these various signals, create an internal decision support score themselves. ” In the end, no one really knows what is behind these ESG scores, and it is very complicated to justify a responsible investment approach. ”, notes Mathieu Joubrel, CTO of the start-up ValueCo.
” For their part, investors have a fiduciary responsibility to their clients: they are obliged to do what is best for the return on the investments they have been given. As a result, ESG is generally only an ancillary factor in their investment decisions. ”. It is therefore at the end of the value chain, in this relatively immature ecosystem, that ValueCo intervenes: the start-up, incubated at IMT Starter (which involves schools Télécom SudParis, Institut Mines-Télécom Business School, and the ENSIIE), captures market sentiment on ESG and sustainability factors of listed companies.
ValueCo collects ESG scores developed internally by asset managers on all companies in their investment universe — between 500 and 30,000 companies for listed investors — to make a database. These scores generally include an aggregated ESG indicator, indicators by environmental, social, and governance sub-pillar, and sometimes more specific indicators, relating to biodiversity for example. With this Benchmark, the startup provides an unprecedented global vision of ESG scoring: it is able to give meaning to investor scores, but also to capture market signals concerning the sustainability of a company.
This data is made available by ValueCo on a SaaS platform (Software as a Service), available by subscription to investors who contribute to the base. On the user side, the platform is a “press button”: all you have to do is select the name of a company to get a dashboard with analyses. Investors can position themselves in relation to the average on a particular company, position it in relation to other companies in an identical market, or even filter by a specific score to exclude the lowest rated companies from their portfolio — or conversely select the best rated companies.
The ValueCo team is also working to set up APIs (Application Programming Interface or application programming interface) to connect to its customers' systems and exchange information live: the scores are immediately injected into the database and the customers benefit from internal access to the analyses.
The global vision of ESG ratings offered by ValueCo is beneficial to all stakeholders in a market which, until now, has evolved in silos, with each party keeping its own evaluation method to itself. The start-up encourages contributions from investors, either by offering paid access to its benchmark solution, or in exchange for a number of free services, including participation in research initiatives and privileged access to results. Working with collective intelligence is also an argument in its favor: ValueCo provides market information to the listed companies themselves, so the investors who contribute to the database participate in providing feedback on their perception. In this way, ValueCo's approach is appealing because shareholder engagement is low, with no financial implications for investors (all it costs is to export the scores), but the responsible impact is significant.
Indeed, ValueCo addresses all the stakeholders in this ecosystem. Although 90% of its customers are asset managers, the start-up also targets institutional investors and "issuers", i.e. listed companies themselves. Institutional investors include insurance companies, commercial banks, pension funds, etc., which pool savings and entrust them to asset managers. Thanks to the analyses provided by the ValueCo platform, these investors have a better understanding of the approaches taken by the managers they finance.
Issuers, for their part, have an interest in subscribing to the platform, not only to gain investor insight into their own sustainability performance, but also to benchmark themselves against their peers. ValueCo provides them with tools to identify their strengths and weaknesses, and adapt their CSR strategy in line with best market practice.
"In a financial analysis, the input data are clear, and at the output it's the return that's sought. It's a deterministic and unambiguous process," explains Mathieu Joubrel. "When it comes to ESG, there's a huge amount of data that's much softer and harder to capture, so it's legitimate to have different opinions. The problem is not understanding what these scores mean." The divergence of ESG scores is well theorized, and three main sources are identified.
The first and most important source of variability is measurement divergence. At the start of the value chain, there is a strong need to collect the right data at company level: the same signal name, for example "Safety at work", sometimes covers different indicators from one company to another. So, even with an identical methodology, the heterogeneity of input signals creates discrepancies in output ESG scores. However, this source of divergence is set to diminish: many players are working on the deployment of technological solutions to retrieve and standardize indicators within companies, so that they are comparable. Europe, meanwhile, is working on the development of norms and standards to complement these technological initiatives.
The other two sources of divergence are scope (the criteria selected to create the score) and weighting (the relative importance of these criteria in the calculations). These parameters constitute the "secret recipe" of the rating agencies: they depend on the definitions of CSR performance and the models developed by each agency. These biases are not destined to disappear, but a shift is taking place, as investors are less and less inclined to use the aggregated and opaque score models of rating agencies. With the standardization of input data on companies, investors will tend to collect more granular data themselves, and then decide on the scope and weighting to create their own ESG scores. "In any case, the less divergence there is in measurement, the more comprehensible the scores collected by ValueCo will be, as the differences between scores will then be explained mainly by investors' internal approaches. We'll then be in a better position to help them position themselves and take control of their indicators", says the CTO.
Once the ESG score database has been compiled, ValueCo's data processing relies essentially on statistical methods. "Our aim is to bring understanding to our customers in a market that is already confusing. So, for the time being, we don't want to complicate this system with deep learning algorithms that will themselves be difficult to explain", points out Mathieu Joubrel. "Having said that, there are subtle, weak signals to be picked up, for which we are considering using artificial intelligence. For the moment, we're at the research stage, but there's clearly an interest in integrating slightly more advanced technologies into the platform in the future."
Research is indeed a fundamental pillar for the ValueCo team, which collaborates with several universities and institutions, including MIT and Grenoble École de Management, sharing data and analysis with them. The aim of this approach is not only to participate in publications and advance the ecosystem, but also to recruit talent in data science and fundamental research. After two years of development, ValueCo is now in a period of transition and is taking advantage of the strategic support provided by the IMT Starter incubator to anticipate its move to scale. The start-up plans to expand its team in the near future, following a fund-raising round scheduled for the end of 2023: the VivaTech trade show should provide an ideal platform for this.