By Joseph Neu
The real fail point for treasury management systems is access to underlying data.
One of the clear messages from the recent NeuGroup meeting cycle is the growing realization that data will be driving treasury’s future as it is disrupted—like everything else—by digitalization. Not to be overlooked in all this is how data and digitalization will disrupt the traditional treasury management system (TMS).
I have always thought the chief fail point for the traditional TMS is its inability to deliver the out-of-the-box reports that most treasuries need. The bigger problem being exposed now by the proliferation of data analysis and visualization applications (many self-built) is the lack of access, or limited access, to required and increasingly nice-to-have data.
In simple terms, treasuries are coming around to the same realization reached by other corporate functions: To become data-driven, they need to create separate “data lakes” and data warehouses to capture raw data in native formats without truncation and then structure it for defined purposes, such as feeding cash forecasts and risk analysis.
Access to all the data is critical because you don’t always know the importance of missing data; for example, the actual currency of payment vs. the translated amount recorded in the general ledger system. Plus, the connection of everything via the internet of things is now creating data that no one imagined existing before.
“The chief fail point for the traditional TMS is its inability to deliver out-of-the-box reports that most treasuries need.”
The sources of data relevant to treasury and finance functions can be substantial. This was abundantly clear in a map of one member’s data ecosystem showing his finance function’s data sources. Unfortunately, a lot of data provided by sources in native formats is lost when it’s pulled into widely used finance systems.
For example, the data that your corporate purchasing card vendor generates is much more detailed than what you will be able to find in commonly used expense management systems, which are not designed to pull it all in. Similarly, all the data a customer submits with a payment via the banking system does not show up on the payee end (thanks to field limits and the lack of standardization in the MT940 or XML format messages, though the latter is better).
So as treasuries create data pools (lakes and warehouses) structured for their needs and build applications to feed data into and then analyze, they will increasingly be bypassing traditional TMS systems. The traditional TMS may still be a source of data for the pool—as the book of record for trading and intercompany transactions, for example—but its value could end there and still be threatened by made-for-purpose applications that better sync with treasury’s cloud-based data pool. This should provide a clear mandate for traditional TMS providers to radically rethink their approach to data and revisit the business case for a single, comprehensive system.