NLP Applications in Investment Research


The following is a guest article written by Tejas Shastry, serial entrepreneur and Chief Data Scientist of GreenKey (GK).   GK leverages NLP to convert complex, mission critical audio and text content into seamless data structures to automate and analyze real time human tasks.

This coming Wednesday, February 25th, 2021, Mike Mayhew, Chairman & Founder of Integrity Research Associates will be interviewing Tejas Shastry about how NLP is being leveraged by sell and buy-side institutions to enhance the production and consumption of investment research.  Click here to register for this free webinar.  The following are a few of the topics that will be addressed in this live event.

What are some of the latest developments in NLP that prospective users should be aware of, and why do these developments matter?

NLP as a field has been around for decades, but there are some recent drivers that have caused it to come to the forefront of business applications. 

The explosion of data we just highlighted means that firms have no choice – it’s simply too much data for a human to review and the only way to scale insights is with machines. At the same time, there have been some key advances in machine-learning and deep-learning from many of the industry leaders like Google and Microsoft that have enabled NLP models to understand language at the same level of accuracy as humans.

This means that the problem and the solution are converging.  One of the most recent advances in NLP has been a new model type called a transformer. This model type is good at what’s known as “few shot learning”. Traditionally, AI models would need millions of data points to accurately classify a piece of information. Transformer models can extrapolate from their base understanding with much less data. This means prospective users can now unlock new applications quickly with much less training data than before.

How are analysts leveraging NLP to make the consumption of research, news, earnings announcements and other unstructured text a more efficient part of their research process?

NLP solutions like GreenKey enable firms to automatically extract KPIs, topics, and sentiment from their daily voice, chat and email data.  Firms using this technology to structure their conversation data can easily summarize key points without having to manually sift through massive amounts of voice and chat files. 

GreenKey’s NLP models are trained on terabytes of data so they can understand context and extract the strongest viewpoints (negative or positive) around a particular topic or company. These insights make it possible to pinpoint missed opportunities, identify market trends and quantify the profitability of client relationships. 

How are some investment banks using NLP to extract analyst sentiment from their research reports which are being used by quant investors as part of their research process?

NLP provides an objective measure of analyst sentiment so it can be scored against a machine benchmark.  GreenKey’s platform lets analysts tag sentiment, topics, and key metrics across research documents, earnings call audio, and news using Natural Language Processing.  Investment analysts, institutional salespeople, or customers can now more easily and quickly search and query sell-side research documents that have been processed by NLP. 

How are some asset managers requesting their sell-side analysts to train and distribute NLP models for extracting key KPIs and summarizing research reports. 

Asset managers leveraging NLP are requesting their sell side analysts train models to capture their viewpoint and sentiment. Once adequately trained, they can run the models on all of their own documents going forward, providing a scalable system for reviewing research reports and extracting actionable insights.

GreenKey has over 30 pre-trained models that can be easily customized so firms don’t have to spend months or even years building models internally. To learn more about GreenKey for financial research, download our whitepaper.  

Tejas Shastry, Chief Data Scientist, GreenKey Technologies

Dr. Tejas Shastry is the Chief Data Scientist at GreenKey Technologies. Throughout his career, he has led research and development teams building machine learning algorithms across many industries, including finance, consumer electronics, materials research, and industrial IoT. At GreenKey, Shastry’s team focuses on building speech recognition and natural language processing for noisy, jargon-heavy environments in finance and public safety. GreenKey’s models focus on broad understanding of financial language and few-shot learning methods for customization.

This week’s NLP Applications in Investment Research discussion is the fifth in a series of live online events organized by Integrity Research Associates covering various sectors, themes, and topics that asset manager clients have asked about. Click here to register for this week’s live panel discussion.

Integrity Research has scheduled a new ResearchInsights panel slated for March 3rd which will address the insights and trade ideas of three highly experienced independent analysts on their views on how the U.S. Consumer Sector will perform in 2021 and beyond.  Click here to register for this free webinar.


About Author

Mike Mayhew is one of the leading experts on the investment research industry. In addition to founding Integrity Research, Mike is on the board of directors of Investorside Research Association, the non-profit trade association for the independent research industry, and a frequent speaker on research industry trends and developments. Mike has over thirty years of research industry experience. Email:

Leave A Reply