Case study by Truvalue Labs
In the spirit of showcasing leadership and raising standards of responsible investment among all our signatories, we are pleased to publish case studies of all the winning and shortlisted entries for the PRI Awards 2020.
Give a brief overview of your project’s objectives and how your approach to the subject matter differed from previous research.
Truvalue Labs, a technology company that uses Artificial Intelligence (AI) and Natural Language Processing (NLP) to uncover material ESG information on companies, has created a dataset that allows investors to draw actionable insights from the increasing COVID-19 related information the company is processing through its sourcing of big-data.
To make sense of this COVID-related information, Truvalue created five new NLP-based signals that specifically related to the global humanitarian crisis at hand:
- Social Impact: Captures discussion of COVID-19 and its impact on the quality of living, access to education, social distancing, mental health, depression, access to healthy food, and related topics.
- Labour: Captures discussion involving COVID-19 and employees, layoffs, working from home, paid sick leave, unemployment, and related topics.
- Economy: Captures any discussion of COVID-19’s impact on the broader economy.
- Supply Chain: Captures discussion involving COVID-19 and supply of goods, production halts, manufacturing stoppages, and other related issues.
- Response: Captures information related to any R&D or efforts to create vaccines or to shift manufacturing to produce test kits, ventilators, sanitizer, or any other unusual, crisis-specific goods, services, or products.
All of these new signals are assessed with respect to the behaviour and actions taken by individual firms. And the dataset that was constructed from it also aggregates and organises this information by sector, industry, geography, and so on. Truvalue believes it has largely succeeded in the objective of uncovering important insights and trends that are useful for investors facing an unprecedented period of uncertainty.
Describe your methodology, including how you addressed macro trends and mechanisms for effecting systemic change.
The methodology primarily uncovers stakeholder perception on the actions of companies through early periods of this recession, as well as important macroeconomic trends observed by organising the data through various facets, such as industry, geography, SASB category, and so forth. The initial report documented evidence of five topics:
- Employee Health and Safety
- Labour Practices
- Access and Affordability
- Product Quality and Safety
- Supply Chain Management.
It was found that the relative COVID-19 data volume Truvalue captured had peaked in various regions weeks ahead of the peak in new COVID-19 cases. In addition, the firm observed that the overall relative COVID-19 volume it captured had an inverse relationship with US equity prices, which backed up its purpose as an indicator of relative dislocation.
When assessed by industry or sector, Truvalue found that relative COVID-19 data volume helped to assess what areas of the economy were more stressed, or perhaps better positioned. For example, its second research note focused on the Response Signal, which captured information related to firms that had shifted operations as a result of the global pandemic. Here it found a correlation between industries with strong volume on the Response Signal and year-to-date changes in next twelve-month profit expectations (as measured by consensus broker estimates).
External research has also validated the usefulness of Truvalue’s new signals and dataset. A report authored by a team at State Street and George Serafeim from Harvard Business School found that firms with positive sentiment on the Labour and Supply Chain signals exhibited higher institutional money flows and less negative returns.
Separately, a paper by Ola Mahmoud (University of Basel and University of California at Berkley) and Julia Meyer (University of Zurich) made use of the new signals to show evidence that through the recent market crash, US stocks with high ESG performance experience higher stock returns.
Outline how your findings have been applied in a practical context and their wider benefits for investors or the financial system.
Truvalue was able to use the new data alongside its SASB dataset for comparative analysis. For example, the data made it possible to differentiate between the perception of firm performance on Employee Health and Safety (a SASB topic issue) with respect to COVID-19 related issues, and the performance of the same company on that same topic but unrelated to COVID-19 specific issues.
Beyond the academic studies, which have used its new COVID-19 signals, investors have integrated the dataset into their internal systems to help make sense of a chaotic and uncertain social and economic landscape. Here’s how it has helped:
- With the ESG overlay on these new signals, investors have been able to assess the commitment of corporations to positive social outcomes, as well as gain insight into the quality of governance teams.
- Truvalue’s Research Brief on the Top Responders, which uses the firm’s new COVID-19 Response Signal, has helped investors identify companies that have underappreciated intangible value.
- On the fixed income side, the new signals have more specifically been used in the analysis of new issues such as Netflix’s $1bn of new debt issued in late April of 2020.
- Many institutional investors have used this data and research to help establish and frame their approaches to ESG analysis during this unprecedented time, as well as to track and monitor economic conditions. They’ve looked to relative volume trends on the Economy and Supply Chain Signals to help systematically understand the pace in which the real economy and certain activities remain disrupted or begin to return to normal.
For the financial system more broadly, these signals help consistently hold companies to account during a time when the nature of the social contract between firms and stakeholders is fundamentally shifting.
Both academia and the media have used Truvalue’s dataset to hold companies to account. While the new dataset and research has helped investors, it has also played a small role in creating better societal outcomes.