Our services in the selected location:
  • No services available for your region.
Select Location:
Remember my selection
Your browser is out of date. It has known security flaws and may not display all features of this and other websites

GSAM Perspectives

Text, Tone and Topic


Note: Separate multiple email address with a comma or semicolon.

Send me a copy


Note: Separate multiple email address with a comma or semicolon.

Your Name:

Your Email Address:

Send me a copy

Exploring Potential Benefits of Natural Language Processing

While computers today may not yet be as adept with language as humans, natural language processing will become a crucial part of the investing process in the coming years.

Language is a complex and difficult problem for computers. The fastest super-computer in the world can perform up to 54 trillion calculations per second,1 yet it struggles to produce and understand our language at even a toddler’s level. Vast amounts of information in the world are stored in language form, from textbooks and encyclopedias, to websites and social media, so researchers are hard at work to enable computers to bridge the divide from the digital bits and bytes to the world of language. This technology is called “Natural Language Processing” (NLP), and it represents one of most promising areas of machine learning within the field of modern computer science.

In the past decade, enormous advancements have been made in NLP technology. Today, online translation software can convert most documents into a hundred languages with astonishing accuracy. Numerous online businesses now offer customer support with a virtual agent, a computer that is able to interpret a number of common questions and refer users to the proper answers.

Over the past several years, research has been conducted on how natural language processing can be used to discover trends and sentiment in the market, and in the future, this technology will be a crucial part of what it means to be an investor.

Analyst Research Reports & Earnings Call Transcripts

Investors often utilize analyst research reports to help select stocks. For example, investors may take the current average earnings per share (EPS) estimate for a company (the “consensus” of the analyst community), and compare it to the average EPS estimate from the previous quarter. The companies with the highest average EPS revision, meaning those companies that analysts view as having the most promising and improving earnings prospects, are then favored according to this signal.

Investors have focused on headline EPS forecasts—even quantitative managers—to process a single number from analysts’ reports across different companies over time. A consequence of this approach was that the body of the analyst report (the actual prose) was often ignored. However, much of the information in a research report is conveyed through its words, tone and phrasing. Rather than relying on headline EPS figures, investors can look at subtle shifts in tone and diction in any analyst’s research reports on a given company. Analysts are reticent to change their buy, sell or hold recommendations frequently, so will often subtly shift the tone of their report to indicate a potential change in view while keeping their headline rating the same. By utilizing natural language processing to parse through an analyst’s writing, investors can help predict changes in their headline forecasts before those shifts take place.

A similar analysis can be performed on the transcripts of companies’ earnings calls. Management will frequently hint at shifts in their viewpoint through their phrasing and tone. By utilizing natural language processing to parse through earnings calls, investors can isolate subtle shifts in management’s sentiment around their company, which can be helpful in more accurately forecasting future performance.

In both of these cases, natural language processing is a powerful tool that enables investors to gauge subtle shifts in how analysts and company management are thinking about the future of a company using non-numerical, language-based data.

Topic Analysis

Natural language processing can also help identify what trends are driving the market and what companies are best positioned to take advantage of those trends. For example, a positively trending topic such as wind energy in Europe may point to greater investor interest in renewable energy, potentially impacting the producers, suppliers and related companies in the space in ways that were not so obvious at first glance. Identifying these trends objectively and dispassionately is not a task at which humans tend to excel. Humans are often affected by unconscious biases, favoring trends they prefer and neglecting those that interest them less. Computers are experts at weighing evidence without prejudice. However, much of the information investors use to identify trends is conveyed through human speech and writing—e.g. news reports, online blogs and company filings. Natural language processing allows us to parse through millions of news articles and other text-based data sources per year, extracting those trends that are most compelling with efficiency and objectivity.


With over 13,000 public corporations around the world, global investors would have to digest hundreds of thousands of pages in annual reports and over 30,000 hours of earnings calls every year. For decades, people have utilized computers to help us grasp an increasingly complex world around us, limited to dealing with numerical and easily quantifiable data. Natural language processing is helping to remove that limitation. Computers are increasingly able to answer questions like “What topics are trending in the market place?” and “How are research analysts thinking about the oil industry?” Successful investing has always been about maintaining an informational and analytical advantage. With the enormous amount of information embedded in the language around us, it is likely that natural language processing will be a critical tool for tomorrow’s investors.

Related Perspectives

GSAM Perspectives
The Role of Big Data in Investing

GSAM’s Quantitative Investment Solutions team shares their synthesized views on the role of big data in potential investments and key techniques for approaching big data.

GSAM Perspectives
Big Data is Fundamental

GSAM’s Fundamental Equity and Fixed Income teams discuss how big data can be a strategic advantage or disruptive force in companies and sectors.

GSAM Perspectives
The Political Gets Analytical

Big data has become a powerful force in the election process and is likely to serve an increasingly central role in future political camaigns.