By George Ralph
The biggest difficulty with data, bar it’s quality, is its quantity. It can be difficult to cut out the noise and focus on what really matters to deliver great insights to clients. Market Intelligence teams have a mountain to climb to manage the sheer volume of reports available daily and for smaller firms that do not have the luxury of a market intelligence team, other options must be considered in order for managers to stay competitive.
Natural language processing (NLP) is an increasingly popular field of artificial intelligence. The idea is that NLP applications can sort through vast amounts of unstructured (or non-numeric) data, harvesting and sorting information as they go. NLP looks for specific language or words to help to drive patterns in information that has been published for public consumption, delivering back insights to enhance decision making.
The NBER (National Bureau of Economic Research) in the US has estimated algorithmic downloads of quarterly and annual public company reports went from around 360,000 in 2003 to a whopping 165M in 2016. This provides a clear idea on how the technology has developed. NLP algorithms have advanced hugely again since 2016 so the figure now is almost certainly significantly larger than that.
The reality is that NLP is presenting a new opportunity for all investment managers to access a breadth of information that, while previously publicly available, was not practically available because of the sheer volume of data. NLP can monitor transcripts, articles, video interviews and corporate reports to collate the information on a specific subject, for example ESG performance, and deliver that information back to the manager to aid investor reporting. Traditional data analysis simply isn’t capable of giving the breadth of scope that unstructured data can if mined correctly.
Data, however it is delivered, is only as good as how it is processed, read and reported on. The value comes from harnessing robust processes and procedures as part of a firms overarching data management strategy. The second step is delivering data in a format it can be read and reported on by managers. A central dashboard which provides a view of data from multiple sources, all stored within your data warehouse, allows for clear reporting and makes your data more accessible and manageable. Having the correct tools in place to not only manage data but keep data safe are key deliverables for every firm, but also for investors and regulators too. Centralising data and decentralising cyber security is now accepted as operationally efficient and effective. Keeping a holistic grasp on data assets as they grow is vital. Should a firm suffer a data breach or attack of any kind, the first thing the regulator will want to know is where the breach occurred in the network, and what data has been affected.
As funds move more and more towards digitising operational processes, an outsourced partner can make sense. Focussing on finding the right tech partner who works with other funds with a similar strategy is a great idea. It allows the firm to benefit from the experience the tech partner has of supporting others with similar digital transformation strategies.