How Finance Uses Natural Language Processing

Crunching text is the new frontier

This free report offers 8 case studies of investment companies building Natural Language Processing solutions that generate direct business gains.

The applications range widely across business areas, and include:

How Deutsche Bank improves its environmental investing process
How American Century uncovers hidden market signals in equities
How State street automatically extracts research insights to keep investors informed
How 7 asset managers’ marketing content compares – to each other and to financial news

The report offers both inspiration for firms looking to deploy new solutions, and an accessible introduction to the current state-of-play.

What Is Natural Language Processing?

Machines are useful for crunching structured data – series of numbers and values. But human language is messy. We have irony, humor, jargon; the same word can take on different meanings. Language is often said to be unstructured.

But language is full of structure. From sentence syntax to paragraph structure, text is lush with hidden data, offering clues on the true meaning conveyed by a series of written words.

Natural Language Processing (NLP) is a collective name for a set of techniques for machines to uncover the structure within text data. 

Modern banks and investment managers have built their business on crunching numbers. But, with access to information no longer the competitive edge it once was, pockets of value have become much scarcer. Large volumes of text has become the new frontier for hidden market signals.

It’s also becoming harder to keep handling text data with the same processes. Ever-growing volumes of text information is overwhelming employees .

Each case study covers

The Business Goal

Featured goals include: removing grunt work, reducing errors, providing new services, and bringing down costs.

How they made it work

The solution implemented, and how it fits into the company’s existing workflows.

What's the Big Idea

Clear introductions to NLP concepts, such as: Topic Modelling, Feature Selection, and Named-Entity Recognition.

NLP in Finance 2020: Top Findings

Technical and theoretical progress over the past decade, coupled with the advent of powerful open-source software (which is often free), has made intelligent text processing accessible.

While some off-the-shelf NLP products are available, a survey by the Bank of England and the Financial Conduct Authority shows financial services firms already using machine learning prefer developing applications in-house.

Typically, a challenge arises when a company’s current processes are overwhelmed by the amount of text data the company wants or needs to process.

Pulling reliable insight from large text volume lies beyond what human workers can do. Employees working alongside machines presents a new competitive edge.

Monetary benefits are easiest to measure in terms of savings (as was the case with JP Morgan Chase) or excess returns (as was the case with Deutsche Bank).

But the greater impact is seen when companies eradicate existing moats (for example, N26), or create a whole new class of services (for example, State Street).

To get NLP right, domain knowledge is still crucial. Solutions would typically target repetitive tasks and require tailoring to the relevant use case. Invariably, this involves domain expertise.

Download the PDF

This 27-page report binds together all 8 case studies. No sign-up required.

See Text Analytics in action

Comparing the two companies’ LinkedIn strategies

Gist! instantly provides a 5-bullet summary for any article.

How can your content stand out from the crowd?