Natural language processing: time is now to advance human language insights for finance industry

[Sponsored article]

With financial firms relying increasingly on data to maintain a competitive edge, the benefits of using artificial intelligence (AI) to turn this data into insights cannot be overstated. In particular, the use of natural language processing (NLP) is becoming essential for financial firms of all sizes.

“In the past few years, major advances in NLP have been achieved due to increases in processing power, data availability, open source and new techniques. These are already being used across the financial world,” says Geoff Horrell, group head of innovation at the London Stock Exchange Group (LSEG) Labs, a corporate innovation team of data engineers, data scientists and user experience experts tackling the next generation of challenges in financial markets.

To better understand the impact of NLP in finance, LSEG Labs has this year launched a research report, “NLP in Financial Services”. The report interviewed NLP subject experts and financial sector end users to outline the current market landscape and better understand how financial companies are using NLP, as well as how they could use the technology in the future.

“Having worked with NLP for many years, we wanted to take stock and see how our customers and the markets have evolved since the ‘big bang’ emergence of advanced language models a few years ago,” says Horrell, explaining the motivation behind the research.

Investment not yet matching growth

NLP is a branch of AI that helps computers understand, interpret and manipulate human language. Put simply, it helps machines understand what people write or say in a natural or conversational way. Real-world applications of NLP are all around us – from spam filters within email applications to predictive text on smartphones, and now in smart assistants such as Siri or Alexa.

Within the financial services sector, NLP is being used in a number of ways. On one level, it can be employed to create efficiencies by decreasing manual routine work and automating functions such as auditing and accounting. On another level, it is being used to review enormous amounts of data and identify risks, analyse sentiment, and spot issues and trends that may have an impact on financial markets.

Across all business sectors, the demand for NLP-related technologies is rising. According to Verified Market Research, an international market intelligence consultancy group, the global NLP market was valued at just over US$11 billion in 2020, and is projected to reach more than US$45 billion by 2028. This trend is mirrored in the financial sector. In its report, LSEG Labs found that 47 per cent of respondents said they had a larger NLP budget in 2020 compared to the previous year.

But while investment and adoption of NLP is increasing across the board, many firms are still stuck in the experimental phase. LSEG Labs found that one-quarter of respondents were not applying NLP in a meaningful way, typically using pre-packaged analytics platforms to pilot projects across a single business unit. In contrast, NLP technology was considered fundamental for less than a fifth of firms, according to LSEG Labs.

The time is now

While the use of NLP in the financial sector is still in its infancy, LSEG Labs suggests that a combination of technological and human advances are nevertheless creating ideal conditions for those companies looking to invest and scale up use of NLP and deep-learning technologies.

One of the key factors driving this is the widening availability of NLP-related vendors. These include both the growing number of fintech providers, which can provide NLP applications in areas such as market forecasting and sentiment analysis, and the more traditional big tech providers such as IBM and Microsoft.

Importantly, there are those vendors providing open-source technologies, creating a collaborative model that can help financial firms implement NLP more effectively. Examples include Google BERT and Stanford NLP.

Then there is the data itself. The rapid growth of unstructured data is providing firms with a seemingly never-ending source of potential information and insight. The amount of unstructured data is growing at an astonishing rate of 50 per cent per year, and according to the research firm ITC, the volume of unstructured data will reach 175 billion terabytes by 2025, up from 33 billion terabytes in 2018.

Data providers such as Refinitiv are also an important part of the NLP story, producing a huge volume of untapped and unstructured data such as earnings calls, news, analytics and more.

Finally, thanks to a deepening talent pool with AI and machine learning skills, there are simply more qualified people available to design, implement and run NLP programs. LinkedIn, for instance, reported a 40 per cent increase in global AI-related hires in 2020, which suggests a growing skills resource for financial firms.

“This report shows that tools are maturing, technology has evolved and data science skills are more widely available,” Horrell says. “The limit now is the vision, creativity and ability to execute in the new age of machine learning.”