NLP in finance: Text analysis is making investment decisions easier

Image credit

NLP in finance: Text analysis is making investment decisions easier

NLP in finance: Text analysis is making investment decisions easier

Subheading text
Natural language processing gives finance analysts a powerful tool to make the right choices.
    • Author:
    • Author name
      Quantumrun Foresight
    • October 10, 2022

    Post text

    Natural language processing (NLP) can pore over thousands of words and written information to produce data-supported narratives to guide investors and companies in the financial services industry on where they should invest their capital.

    NLP in finance context

    NLP, a subset of artificial intelligence (AI), uses words, phrases, sentence structures, and other linguistic tools to identify a theme or pattern in structured and unstructured data. Structured data has specific, consistent, pre-assigned formats (e.g., portfolio and index performance). Unstructured data is a mixture of different media file types such as videos, images, and podcasts.

    Leveraging its AI capabilities, NLP uses algorithms to categorize data into a structured pattern that can then be “read” by NLG (natural language generation) systems for storytelling and reporting purposes. NLP and NLG systems can be applied together within the financial services industry to read through and assess companies’ annual reports, videos, press releases, interviews, and historical performance, among others, to develop investment suggestions, including which stocks to buy or sell.

    Disruptive impact

    Before NLP, most financial processes were manual and labor-intensive. An example is the due diligence review conducted by US-based bank J.P. Morgan & Chase for its potential clients, which used to take about 360,000 hours of manual work annually. An NLP system was able to automate much of this process dramatically, while also reducing clerical errors. Another example is the pre-trade phase, where analysts spend about two-thirds of their time collating data without even knowing whether this information would be relevant to focused projects. The introduction of NLP allows for data collection and organization to be automated, thereby allowing analysts to spend more of their time reading through higher-value information, i.e., optimizing personnel time spent within the financial services industry.

    Another area where NLP is proving to be extremely useful is sentiment analysis. Through keywords and tone, AI can go through press releases and social media to determine public sentiment toward a public event or news item (e.g., “CEO of a bank resigns”) and gauge how this sentiment may affect the bank’s stock price. NLP supports other significant services such as fraud detection, cybersecurity risks, and performance reporting. Insurance companies could potentially deploy NLP systems to assess the submissions of clients claiming a policy to search for inaccuracies and contradictions.

    Implications of NLP applied within the financial services industry

    The wider implications of NLP being leveraged by financial services companies may include:

    • NLP and NLG systems working together to collate data and write reports on annual reviews, performance and even thought leadership pieces.
    • More fintech firms using NLP to perform sentiment analysis on existing products and services, future offerings, and organizational changes.
    • Fewer analysts needed to conduct pre-trade analysis, and instead, more portfolio managers being hired for investment decision processes.
    • Fraud detection and auditing activities of various forms will become more comprehensive and effective.
    • Investments becoming victims to a “herd mentality” if too much input data uses similar data sources. 

    Questions to comment on

    • If you work in finance, is your firm using NLP to automate some processes? 
    • If you work outside of financial services, how might NLP be applied in your industry?
    • How do you think banking and finance roles will change because of NLP?

    Insight references

    The following popular and institutional links were referenced for this insight: