1. Home
  2. Refinitiv Labs
  3. Projects - what we're working on
  4. COVID-19 Company News Tracker
Colourful lines of data displayed against a black background

COVID-19 Company News Tracker

Identifying COVID-19 risks and opportunities for companies, industries and supply chains

Leveraging Natural Language Processing to help firms more quickly and effectively detect signals of risk and opportunity in news data.

Problem and opportunity

Keeping up with the most relevant and meaningful COVID-19 news and data presents huge challenges globally. 

Economists, investment analysts and traders need to be able to track the positive and negative impact of events on companies, industries and supply chains. But the speed and volume of developments reported in the news makes this extremely difficult. 

From November 2019 till August 2020, Reuters reported over 150,000 news articles mentioning COVID-19.

Screen shot from COVID-19 company news tracker showing search results featuring relevant headlines

Company search results displaying relevant news headlines with identified risks and opportunities

Solution        

Generating insights when unforeseen, high-impact events occur

Refinitiv Labs has built the COVID-19 Company News Tracker, which provides time series signals related to the coronavirus that enable customers to quickly identify signals of risk and opportunity in Refinitiv news data.

The machine learning (ML) and natural language processing (NLP) model provides insights on companies and industries of interest, when unforeseen, high-impact events create uncertainty in financial markets.

The COVID-19 Company News Tracker is now available to use in the Macro Vitals app in Eikon and Workspace.

Screen shot of COVID-19 company news tracker with time series charts

Time series charts comparing company and peer risks and opportunities, as well as peer fundamentals

The COVID-19 News Tracker in action

Combining ML, NLP and high-quality data

The prototype includes:

  • An ML and NLP model built on basis of an existing Refinitiv Labs risk tracker and Google’s open-source NLP model, BERT
  • Pre-trained with 2.7 million relevant words from the Refinitiv News Archive and 7,500 annotated articles
  • Automatically determines whether a news article contains a COVID-19 risk or opportunity for the companies or industries mentioned
  • Tracks risk and opportunity trends over time, and compares company results against those of their peers
Screen shot of COVID-19 company news tracker featuring value chain diagram

Interactive value chain diagram illustrating risk and opportunities mentioned for related companies

A collaborative approach

Refinitiv Labs takes a collaborative, customer-focused approach to building solutions to real problems in financial markets by combining customer feedback, extensive data capability and exceptional partner technologies.

Collaborating with our customers

  • Strategists, economists, traders, wealth managers, company analysts and data specialists were consulted throughout the development process to create a relevant solution to their needs and pain points
  • Customer feedback informed the functionality and user interface of the application
  • Additional use cases suggested by customers to improve the ML model

Refinitiv partner technologies:

  • AWS CloudWatch
  • AWS Batch
  • MySQL
  • Node.js
  • Angular
  • BERT
  • BERT as a service
  • Python
  • NLTK
  • spaCy

Looking ahead

Graphic illustrating a black arrow moving inside a blue box

What Refinitiv Labs is thinking next...

The team is working on additional improvements and functionality, including:

  • The ability to determine the temporal nature of news stories, and to show customers when the data relates to past events, ongoing events or developments expected in future
  • The potential to deploy the model to detect and better respond to unforeseen, high-impact macro events in the future