Data Analytics for Treasury

Don't drown in the data lake!

Data Analytics seems to be a big buzz word today and rightly so, it should be. However, is it the panacea that everyone is expecting it to be?

Corporates and FIs have been investing in data related technology over the last few decades and more significantly over the last few years with the primary objective and a hope that there will come a time when all information will be accessible at their fingertips on a real time basis to evaluate and make decisions. Great progress has been made in the right direction where we are now able to store Terabytes or Petabytes of both structured and unstructured data, can easily perform real-time processing, allow for decentralizing computing as the need arise while maintaining the important cost-value proposition. Giant leaps have also been made in the technology to perform advanced analytics on the data and to make the user interface friendlier for the business users to independently work on. Things look really promising to achieve the primary objective until we look further in the detail.

Emerging technology is not without pitfalls

Despite having access to the best technology available as of now, most organizations are plagued with fundamental problems linked with the usage and quality of data. Organizations tend to have multiple systems interacting with each other, hence the so called golden ‘Source of Truth' data might originate from different places. One set of processed data is often sent to another application which has to analyze it again for its use. Invariably this analysis or re-processing of data lead to errors, resulting in data quality being compromised. Further processing of this data leads to the introduction of even more errors at various downstream data processing nodes. Understanding the potential issues surrounding the usage and quality of data are critically important for the corporate treasury function, because most of the day to day decisions we must take are based on information gathered from different source systems.

Statistics: data enhancement or simplification of truth?

Another challenge organizations face is the way the analytics is designed. Decision makers tend to overly depend on the output of statistical techniques to infer results. There is always a possibility that the analytics can be poorly designed or can be misused by people by providing biased data aiming to get a certain outcome. Even the strongest of statistical techniques can only predict results with a certain probability and are susceptible to providing false positive or false negative outcomes. This is in addition to the challenge of making sense of the outcomes as the design can churn out relationships that would be statistically explainable due to the way the data is, but not make any business or common sense. For e.g. some data can present a strong correlation between the consumption of ice-cream and the sales of oil tankers, but does it make any business sense? Perhaps not..... From a risk management point of view, we have seen this dependency on statistical techniques in the calculation of ‘at Risk' numbers. By using statistics, we try to simplify portfolios of very complex financials risks into one number!

Data can exist and be analyzed using any technology or statistical methods, but an organization needs people to adopt and evolve their decision making around data centric activities to make full use of analytics. This is important especially considering that the technology evolves at a blistering pace and it can be hard for an organization to focus on continuous upskilling of their resources. There is a risk of organization being stuck with an obsolete technology just because their resources cannot evolve as quickly and lose out on the competitive advantage which the newer technologies offer.

Added value for treasury management

Fortunately, all these challenges can be effectively managed by adopting better processes and controls, establishing a single golden source of truth data, liaising with partners who specialize in specific areas like data management, analytics and establishing a comprehensive employee training regime.

In the current state, the use of data analytics in Treasury Management has a long way before it can be blindly trusted in decision making. Nevertheless, organizations should adopt and embed it into the very fabric of the organizational culture as it has limitless potential. However, there should be a constant effort to evolve it to get the best results and make it into the panacea which everyone expects it to be.


If you have an yquestions, please get in touch with Mridul Sharma via +44 20 7730 2510.

Share this article: