Only a small portion of corporate finance departments has adopted the most advanced technologies, and just over half say those technologies, including robotics and natural language processing (NLP), will be “must-haves” over the next three to five years. However, this is a timeline they may want to reconsider because the new technology requires complicated training; getting started now would put corporate finance teams ahead of the game.
KPMG teamed up with Forbes Insights to survey CFOs, controllers and other financial executives early this year and recently publish an analysis titled “Digital Transformation: How advanced technologies are impacting financial reporting and auditing,” looking at how companies are dealing with the digital revolution. Marc T. Macaulay, US cognitive technology audit leader at KPMG, noted that the report focuses more on the financial reporting and audit side, although the findings are relevant to treasury as well.
A full 69% of respondents said their organizations use advanced technologies for the financial reporting function, with 30% responded that its use is limited but increasing. However, while three quarters of respondents’ companies are using predictive analytics and workflow automation, and 89% are using the cloud for financial reporting, only 38% and 35% are respectively using advanced robotics and natural language processing (NLP), in which a machine will read unstructured data.
“So the journey is still in just the formative stages with some of these more advanced technologies,” Mr. Macaulay said.
Mr. Macaulay noted that the key question for treasury is how to harness data to make decisions across treasury functions, ranging from cash forecasting to hedging risk to liquidity management, that are aligned with the company’s strategy and risk appetite.
“The treasury must determine the best data management solution to address the identified business problem and not vice versa,” he said.
The study revealed that 77% of finance teams use predictive analytics. Mr. Macaulay said it’s unlikely they use the technology to forecast key factors such as FX and interest rates, relying instead on forward curves and more conventional tools, but those forecasts could be “sensitized” to determine the impact on the specific company.
Unsurprisingly, 75% of respondents employ workflow automation, the development of which has been underway for years. Less than 40% of respondents are using robotics and natural language processing, which Mr. Macaulay described as more advanced but nevertheless generally available to corporates over the past few years.
In the case of NLP, said Mr. Macaulay, the machine—typically an application in the cloud—essentially has to be trained. For example, a treasurer who wants to analyze certain analysts’ reports regularly would have to teach the NLP application to monitor for when the reports are issued, download them to the system, read them, look for the specific attributes the treasurer cares about, and then, probably in conjunction with other technologies, provide that information to the treasurer.
“Now, the treasurer or his/her staff won’t have to read the reports, and they can instead have the machine do that,” Mr. Macaulay said. “The difficulty comes in having to train it. The first time the application may do relatively well but not perfectly, so it has to be retrained, and then retrained again and again, each time getting better. The technology is there, but does the treasurer have the data as well as the specialists to train the machine?”
Employing robotics and the other more advanced technologies is similarly challenging. In the meantime, only 26% of respondents saw these advanced technologies as becoming a “must-have” in the next year or two, at least as far as financial reporting, while 53% see them as must-haves in the next three to five years.
“Given the speed of technological advancements, the latter may need to reconsider their sense of urgency on this issue,” the report noted.