You Can't Buy Your Way Out of This With a Data Lake

It is a tempting piece of logic, and it is being made in a lot of finance functions right now: put everything in one very large place, point AI at it, and the granularity problem solves itself. If the data is all there, why does anyone still need a subledger?

In this segment, host Mark Aubin puts that argument to the panel, and Mike Dolan, Partner at Ernst & Young, draws the distinction that undoes it — data platforms are excellent systems of analysis and were never designed to be systems of record. The control over business logic, the orchestration of which entries get recorded in which sequence for each individual loan or trade, has to live somewhere, and a data dump does not do it. There is also an industry-specific dimension the build-it-yourself route tends to underestimate, from LDTI and IFRS 17 through to CECL.

Adrian Mustafa, Product Director at SAP Fioneer, is candid from the vendor side: with a strong IT team and real accounting expertise, you probably can build the base functionality. His warning is about year five to year ten, and what a self-built subledger has usually become by then. See the rest of the series on the Cloud Accounting Subledger from SAP Fioneer.