1. Home
  2. Cash Flow Management & Forecasting
  3. Cash Flow Forecasting

Big data and the corporate treasurer

How can big data be meaningful for corporate treasury? This is one of the questions addressed in Journeys to Treasury, the 2017 report from BNP Paribas. PwC, the EACT and SAP that asks how treasury can give meaning to big data analytics. It points out that, with machine learning software available for as little as €20,000, and sometimes even for free, there are opportunities to understand and present data in meaningful ways to inform complex decisions. The report states that it's really the mind-set that needs to change, not the tool-set: “If the tools are available relatively cheaply on the Internet, what is really holding us back? Is it the fear that 'this is complex stuff'?”

Identifying the business issue you need to solve is the most important task, according to the report. Treasurers should be as precise as possible in defining the problem and identifying the data they need, bearing in mind that multiple sources of data might be needed and proxies may need to be created if core data is not readily available. Gathering and 'cleaning' the data is also a time-consuming step in the process, but one that can be substituted with algorithms and automation. Only after passing these three stages (identifying the problem, identifying the data sources and preparing the data) will the data be ready for analysis, which can come in degrees of complexity, as shown by the figure below, which shows that obtaining a description of what has occurred requires a much simpler level of analysis compared to predictive or prescriptive analytics.

The report also suggests that treasurers consider using software such as Tableau or QlikView, two US-based business intelligence and data visualisation software companies, which can both be used with most treasury management systems (TMSs).

The report sets out three key points for treasurers considering using big data analytics:

  • The question is not when and how you plan to use data analytics, but rather what is stopping you from doing so already?
  • The most difficult part, and therefore where most time should be spent, is defining the problem statement and the desired outcome.
  • Consult carefully with experts, not only on how to build the output but on how to manage the collected data within the regulations.

Like this item? Get our Weekly Update newsletter. Subscribe today

Also see

Add a comment

New comment submissions are moderated.