Taking a diligent approach to ESG risk
RepRisk has harnessed both AI and human analysis to collect ESG data on more than 120,000 companies and infrastructure projects. Its Executive Vice President Alexandra Mihailescu Cichon explains the RepRisk approach
Environmental Finance: What’s the RepRisk approach to assessing ESG risks, and how does it differ from other companies in the ESG data business?
Alexandra Mihailescu Cichon: A lot of providers in this space come from the SRI industry. We do not. We come from credit risk management in investment banking – and this has fundamentally informed how our methodology and services have developed.
RepRisk started in 1998 as a risk consultancy focused on the financial sector. Our main product came out of a client request in 2006 for a list of 100 companies it should be careful about doing business with. We devised a systematic methodology to identify companies and projects, and also sectors and countries, linked to ESG risks and violations of international standards. We realised that this was a very interesting set of data – which gave us the idea to launch our flagship product, the ESG Risk Platform, an online searchable database of, to date, over 120,000 companies and 30,000 infrastructure projects.
We use public information only – what we call an ‘outside in’ perspective. So we don’t look at a company’s policies and processes, we look at its on-the-ground performance. Every day our research starts by screening more than 90,000 sources and stakeholders in 20 languages. These aren’t just the ‘Google-able’ sources – we specialise in going down to the regional and local level, which is where those 20 languages play a key role in capturing risk as early as possible.
EF: What’s the role of machine learning and AI in your research process?
AMC: There are traditional providers that tend to be strong at human analysis, and there are new providers who are fully tech-driven. RepRisk has leveraged both those elements from the beginning. This allows us to deliver not only timely but also relevant, actionable data – what we call ‘due diligence quality’ data.
We screen and pre-process over half-a-million pieces of news through our inhouse AI tool, every day. It does issues identification, sentiment analysis, entity extraction and natural language processing. While anyone can do AI, for powerful machine learning the machine has to be trained with tagged or labelled data – and RepRisk has a unique dataset: human-tagged and analysed data that spans more than 12 years.
EF: The company collects ESG risk data on infrastructure projects as well as companies. How – if at all – does the process differ?
AMC: We are looking for adverse information that links a company or a project to an ESG issue or hot topic, everything from palm oil, or human trafficking or corruption. We cover projects the same way we cover companies, with the same issues- and event-driven methodology. This approach allows us to cover any company in the world that has ESG risk exposure – public or private, large or small, in any sector or country. Essentially, we can cover any client portfolio, any investment universe, any asset class. We’re the only provider to cover projects and we have the largest dataset of private companies and of companies in emerging and frontier markets.
EF: How do your clients typically use the risk data you generate?
AMC: While we work with non-financial multinationals, UN agencies and some NGOs, our two big client segments are banking and insurance, and investment management. For the former, we’re used as a due diligence tool, for client onboarding, KYC [know-your-client] processes, client and transaction reviews, and reputational risk management. They use RepRisk to check for material ESG risks and violations of international standards linked to a particular client or project.
On the investment management side, we’re used by everyone from large asset managers to pension funds to private equity firms, to screen and monitor portfolios for ESG risks in a systematic way. Our coverage is very broad and our data is updated daily, allowing clients to flag pockets of ESG risk in their portfolios.
EF: You have recently announced that RepRisk is integrating the Sustainable Accounting Standards Board (SASB) materiality framework into your products. How will that work?
AMC: There are 28 ESG issues that make up our research scope, as well as the 57 hot topics we track. We’re mapping the SASB issues to our issues so clients can see our risk research through the SASB materiality lens. Clients will be able to see what the material issues are for SASB for each company, and then also get the data to see where the ESG risks are. As a next step, we’ll map our data to the 17 Sustainable Development Goals.
EF: How do you see customer demand evolving for ESG data products?
AMC: Clients are getting more sophisticated and are looking for alternative data sets and to complement or re-evaluate what they are currently doing. In general, ESG is more accepted and more mainstream. Before, it was about explaining why they need ESG integration, or helping them make the case internally. Now, they are figuring out how to do it, and our job is to explain how we can support them on that journey.
For more information, please see: www.reprisk.com