Tuesday, 8th October 2019 By Pascal Vilhelmsson
Curing the data overload – using NLP to augment the Investment Process

Investors generally have access to the same types of information. In recent years however, there has been a drastic increase in the amount of data flooding the financial markets. Quantitative data has long been used to drive investments, but the recent increase in unstructured data (data that is not organized in a pre-defined manner) has presented another challenge for asset managers, and investors in general. Gartner predicts that the data volume will grow 800% in the coming years, of which 80% will be unstructured. How will asset managers deal with this data growth, if they already struggle to analyze today’s data?

Up until recently, the growth of data provided an immense advantage to investors, but as this amount of data continues to grow it is becoming less of an advantage and more of a curse. There will be a shift from acquiring as much data as possible to acquiring capabilities to analyze all that data. Today, one of the most time-consuming tasks of asset managers is to sift through all that data. However, recent developments in artificial intelligence (AI), and more precisely, the development of Natural Language Processing (NLP - a subfield of AI) provides a tremendous opportunity to optimize and streamline the analysis of unstructured data and to capture actionable insights.

NLP facilitates more efficient data analysis

NLP is the process by which computers interpret and understand textual data to extract and make sense of the human language, providing valuable insights. NLP can, for example, capture noteworthy developments from the array of millions of texts published each day, providing specific alerts or machine-readable inputs for trading, risk-assessment, or portfolio management machine learning models. Sources of relevant data that NLP can analyze ranges from news, earnings call transcripts, bond prospectuses and many more. This allows asset managers to include these new data points in models that previously only included quantitative factors. Although qualitative factors have been used mostly intuitively in the traditional investment process, this new technology allows models to be built from the ground up with quantitative and qualitative data to account for data interdependencies previously uncaptured. This increases the accuracy of these models and greatly reduces the time required to analyze the mountains of textual/qualitative data available.

AI: an opportunity or threat for asset managers?

Today, numerous mainstream news articles talk about how AI will take over jobs, which can be a frightening fate to face. However, in the near-term, AI technologies like NLP will not replace asset managers, but rather augment the investment process. Adding another tool to the asset manager’s arsenal. Technologies like machine learning and NLP have the potential to re-define the way that asset managers work. Transitioning from being an interface along the data crunching and analysis process to becoming facilitators of data streams and AI-based analytics tools. Ultimately, freeing up asset managers to deal with the more nuanced aspects of the job – coming up with investment theses and taking investment decisions and coming up with investment theses.

Developing this kind of technology in-house takes time and requires the buildup of extensive and top-skilled data science teams. With the cost pressure in the current environment, caused by regulatory requirements and flow of funds from active to cheaper passive investments, developing internal technology is not feasible for many asset managers. The gap between the large and small in the asset management industry is already widening, with large asset managers being able to use their scale as an advantage to offer cheap passive investments and attract top tech talent to combat the data analysis challenges. Like a snowball rolling down the hill, it will continue to pick up momentum and become harder to stop. The largest asset managers will continue to grow and increase the disparity to small- and medium-sized asset managers. Performance will inextricably be linked to a firm’s ability to analyze the large amounts of data, presenting a real challenge to small- and medium-sized asset managers.

 Leveraging technology for a competitive advantage

All is not lost. Rather than incurring the exorbitant costs associated with hiring data scientists and developing internal solutions, smaller asset managers will have to start looking towards fintech startups to acquire capabilities otherwise reserved to the large asset managers. There are a plethora of technology solutions appearing on the market for all stages of the investment process and all types of investment products. These can range from sentiment analysis for twitter data, machine translation for financial documents, all the way to credit risk assessment for corporate bonds. Likewise, large asset managers can also benefit from the collaboration with fintechs. Lean and agile startups working to improve the investment process can add the needed speed and flexibility in testing and evaluating new technologies that large asset managers may lack.

The Asset Management industry is at a pivotal point. The industry is facing pressure from all sides and waiting to implement technology will only add another challenge to the long list. There will be winners and losers in the race to adopt advanced AI solutions. It is up to each asset manager to decide whether to wait or to take action today.