Investing in companies with strong Environmental, Social, and Governance (ESG) practices not only appeal to individuals’ moral compasses, but 90 percent of more than 2,200 individual studies reveal that the business cases for ESG investing are well-founded. (An informative read on this blog, What is ESG Investing.)

As demand grows for sustainable investing or investing in companies with high ESG scores, the use of Artificial Intelligence (AI), and, in particular, Natural Language Processing (NLP), to gather insightful data on companies with strong sustainability, social impact, and governance practices is paramount to recognizing companies that are committed to making a positive impact from those those that are looking for quick gains.

In this article, we describe a use case to reveal corporate labor practices in S&P500 companies with the Accern NoCodeNLP Platform.

Why Corporate Labor Practices Matter

Employees suffer from uncertainty, irregular and lower wages, longer working hours, absence of collective bargaining and representation rights, and often have ambiguous or disguised employment statuses.

Their physical and financial vulnerability are increased due to organizations excluding benefits such as social security schemes, safety and health, maternity leave, and other labor rights that protect employees.

Additionally, 152 million children are still in child labor, 40 million people are in different forms of forced labor and forced marriage, and countless people are excluded from learning opportunities due to discrimination by sex, age, disability, ethnicity, and political and/or religious affiliations.

Identifying Companies with Fair Labor Practices

With the wealth of text data from news sources, blogs or other public data, early warning signals can be identified from companies that are involved in a series of anomalous labor practices.

These insights can assist asset managers in assessing risk and in making informed decisions for portfolio performance. Click the image to enlarge it.

Manually tracking financial results, annual reports, news and social media posts for a company’s ESG performance can be extremely labor intensive and error prone.

Today, portfolio managers are beginning to leverage AI / ML / NLP for the fast and accurate analysis of text data and gain the ability to track more companies, find crucial information that may be hidden, and analyze sentiment of data.

AI Use Case for Corporate Labor Practices

In this use case, we use the Accern NoCodeNLP Platform to track and identify companies in the S&P500 with positive and negative labor practices.

  • Observation Period: January 1st, 2022 to June 30th, 2022
  • Data Sources: Public News and Blogs
  • Company: All S&P 500 companies
  • ESG Issues: Social – Labor Practices
  • AI Model: Accern ESG NLP Model

Deploying the Accern ESG NLP Model takes minutes, and the results can be viewed via an API, visual dashboard, or in a Jupiter Notebook.

The main dashboard presents the results such that asset managers can quickly understand the insights to make informed decisions. Click the image to enlarge it.

The above screenshot of the dashboard shows how data is consolidated and analyzed:

  • Signals refers to the number of hits or mentions about labor practices.
  • Articles is equal to how many unique articles there are on a company’s labor practices.
  • Entities include all of the companies in the S&P 500 that fit the labor practice description.
  • Events is one because we are only looking into company labor practices.

Understanding the data:

  • 141 entities (unique companiesin the S&P 500 that mentioned Labor Practice activities during the observations period.
  • 7,696 stories/documents contained mentions of employment discrimination, child labor, fair labor standards, labor law violations, employee harassment directly related to the company.
  • 9,582 signals were identified (a story may have multiple signals) in the 7,696 stories. A signal is a text snippet that mentions a Labor Practices event and the respective company (or variation of the company name, product, leadership or subsidiary.
  • 131 of the articles were unique. Accern uses an algorithm which identifies and clusters articles which have been republished or are contextually about the same topic and company.

The graph below shows the data sources: users can choose sources that are credible to the event, such a local newspaper. Click the image below to enlarge it.

Labor Practice Signals

The following graph shows the volume of Labor Practice signals across the observation period of January 1st, 2022 to June 30th, 2022. Click the image to enlarge it.

The spikes in labor practice signals in early February, April, and May indicate increased activities around labor practice topics than the average norm.

ESG Labor Practice Signal Key Terms / Themes

The word cloud below shows some of the topics mentioned around labor practices during the observation period. Click the image to enlarge it.

The immediate intuition based on the following key terms from the generated signals is that there are several issues that could indicate a hidden trend among certain industries or companies:

  • labor relations
  • minimum wage
  • work stoppage
  • minimum wage
  • labor unions

Digging into the insights further could reveal a number of warning signals that can provide insights into the labor practices of specific industries or companies.

We can further drill down into the supporting arguments and documents, as shown in the image below. Click the image to enlarge it.

The Accern NoCodeNLP Platform extracts key passages from full documents and understand the underlying details of the warning signals generated.

Additionally, users can analyze the sentiment of each documents with the Accern Adaptive NLP models to identify whether there is a positive, neutral, or negative connotation with the news or company.

Customizing Sentiment Analysis with the Accern Platform

Users can quickly extend and customize the analysis on Accern’s platform by

  • Adding additional content sources from the integrated datastore.
  • Building additional ESG themes to track.
  • Adjusting the NLP model and the parameters of the analysis.
  • Comparing the impact of ESG activities across companies, industries and sectors.
  • Identifying which companies have related or co-mentions and the market perception.
  • Understanding ESG warning signals and critical insights

Watch this 3-min video to see how quick and easy it is to run an ESG use case in the Accern NoCodeNLP Platform.

Building an ESG Use Case on the Accern NoCodeNLP Platform

More AI / NLP ESG Use Cases

Here are more ESG use cases that can be built with Accern:

  • Idea Generation and Surveillance: To identify and monitor investment opportunities within the ESG space in real-time.
  • Risk and Compliance: To uncover risks and early warning signals within the ESG space.
  • Company Research: To improve initial searches by eliminating noise and focus on the core aspects of analysis within ESG.
  • Building Alpha Trading Strategies: To build factor models and investment signals using ESG data.

Schedule a demo to learn more about the Accern NoCodeNLP Platform and how it can drive ESG-focused insights for your investment portfolio.

Share this Post!