Is treasury missing the artificial intelligence opportunity?
Artificial intelligence has the potential to help treasurers deal with regulation changes, increasingly demanding customers and continuing globalisation, yet uptake of the technology remains low. Is this a missed opportunity or is it just a matter of time before AI is commonplace across treasury functions?
Corporate treasurers are under constant pressure to deal with issues such as regulatory changes, increasingly demanding customers and continuing globalisation, making the job more pressured than ever. As a result, treasury departments are progressively turning to technology such as robotic process automation powered by artificial Intelligence (AI).
It’s easy to see why AI uptake is beginning to gather momentum (albeit slowly, as we’ll discuss later). Bob Stark, VP of Strategy at Kyriba suggests that there are promising developments in AI that can help under pressure treasury departments.
“Today, Robotic Process Automation (RPA) are complex algorithms that automate treasury tasks,” he says. “In many cases they work with treasury systems to minimise manual processes. With the addition of machine learning, treasury robots will not only automate today’s cash and treasury tasks, but will also recommend actions to take and identify exceptions to policy – potential fraud is a good example – that require more human involvement.
AI and the battle against fraud
Bill North, Head of Global Sales at Pelican adds that AI, and specifically machine learning, can bring great value in stopping one of the Treasurer’s primary risks: company funds being paid to fraudulent recipients.
“As treasurers face ever increasing pressure to protect company assets, machine learning is proving to be a strong solution in the battle against payment fraud. Global criminals are increasing their attempts to get money out of large corporations. They are finding success and it is relatively easy as most attempts do not involve getting inside company networks. Most attempts have one thing in common: they rely on human error.
“Machine learning (ML) takes the human error element out of the equation. ML creates a normal ‘baseline’ using historical payment data and then analyses each and every payment to see if anything is an exception. It spots anomalies that humans miss. In addition to being part of a more secure strategy, ML also keeps internal costs down as there is no need for manually creating rules to cover all potential scenarios.”
Enhanced decision making
Another example of how AI can help treasurers is in decision making. Richard Hayes, CEO of Mojo Mortgages – a fintech that uses AI as part of their mortgage experience, comments: “One of the huge benefits of AI (that may help corporate treasurers), is that it uses algorithmic decisions to justify why certain recommendations are pursued. For example, AI will use statistical evidence to validate certain decisions, rather than someone manually trawling through extensive pieces of data, or going with their ‘hunch’.
“With evermore demanding customers, corporate treasurers can use their AI to improve the overall efficiency of their role. At our company we use AI to compare the mortgage deals of over 90 lenders in a matter of minutes – by using AI to improve the efficiency of our service, our team are able to set more time aside for the customers that need to speak directly to an adviser.
“Whether it’s a platform that provides greater insight for your customers, or a piece of technology that can conduct international financial analysis at a quicker rate, corporate treasurers can use AI to save both time and money.”
Data is the key
As has already been touched upon, underpinning AI is data. “In a volatile and uncertain global environment, having a holistic real-time view of the situation and being able to react quickly to changing market conditions is key,” explains Michael Walker, Head of Sales Enablement – Corporate Banking & Payments, at Finastra. “The increasing availability of open APIs means data sources from both inside and outside of a company’s systems can be integrated at speed to provide a contextualised view and support better decision making.
“Artificial intelligence, or augmented intelligence to use another phrase, allows corporate treasurers to make sense of vast amounts of data where time-intensive human analysis would previously have been needed. These tools release staff to focus on value-add tasks, whilst the likes of robotic process automation (RPA) can then be used to automate subsequent activities and decisions.
“As an example of how AI and RPA can be used to deal with the pressures of business, think of a corporate that manufactures goods from raw materials and has a complex, global supply chain. Using AI, they can monitor and forecast cashflow and demand for products, predicting shortfalls and gaps between invoicing and payment to better assess the financing required from their banking partners. They can consume real-time data from IoT sensors to dynamically manage their inventory, while leveraging sentiment analysis to monitor external market forces and predict the best time and location to purchase raw materials. RPA can then be used to automatically raise repeat purchase orders or finance requests to the relevant suppliers or banks respectively.”
The missed opportunity
AI clearly has potential to help treasurers. When it comes to low-value, high-volume areas that include data entry, automated decision-making for reconciliations and identifying trading breaks it can definitely help. Likewise, there’s potential to automate regulatory reporting and predictive analysis. Yet, statistics suggest the technology is far from becoming mainstream.
A 2018 report from the Association for Financial Professionals (AFP) warned of only minuscule current adoption of tech-based tools, including AI, a trend that analysts described as “troubling”. The 2018 AFP Technology Survey, underwritten by BELLIN, the AFP surveyed 708 corporate finance executives about their thoughts on adoption of technologies like AI and blockchain. Amazingly, it found that just 1% had implemented AI extensively and 10% had used it to some extent. Only 20% of those surveyed said they will implement it in the next couple of years. That’s despite corporate treasurers clearly saying they have an interest in disruptive, cutting-edge FinTech solutions.
“The gap between enthusiasm for emerging technologies and implementation is very troubling,” says AFP President and CEO Jim Kaitz. “These technologies are disrupting every organisation, especially the finance function. If treasury and finance do not embrace these emerging technologies and implement them to help make their organisations more successful, they risk being left behind by innovators inside and outside their organisations.”
Reasons for delays
According to the AFP, there may be a disconnect between what corporate treasurers expect in terms of positive impact from these technologies and what they actually get. When assessing how these professionals anticipated positive effects like greater accuracy, faster decision-making, cost reduction and streamlined operations, the survey found professionals expected these positive impacts prior to implementation more than they experienced post-implementation.
For instance, 45% said they expected greater accuracy when adopting these tools, yet only 32% percent experienced this positive effect when they adopted emerging technologies.
Meanwhile, other treasurers may be put off by the relative newness of AI technology – which also means it’s unproven in many circumstances, making return on investment hard to measure and complex. At a time when it’s often hard to convince CFOs of treasury needs explaining something that’s part cost-growth minimisation, part indirect revenue and part lower fraud losses can be awkward, to say the least.
Furthermore, significant percentages of companies in the midst of adopting these tools are experiencing challenges, including a lack of adequate skills and technological resources. Nearly one-third cited the high cost of tech implementation as a top roadblock, as well as the steep learning curve and struggle to accurately assess return on investment (ROI).
While it could be argued that treasury is missing out on the AI opportunity, there are signs of progress. “The key to the artificial intelligence opportunity is harnessing data, and through business intelligence projects, treasury is doing a better job at gathering the information to make artificial intelligence a reality,” says Kyriba’s Stark. “I think this gives us a glimpse into what will be commonplace within five to ten years.
And while it’s true that barriers to adoption do exist, BELLIN CEO and Founder Martin Bellin, suggests companies should not let the challenges of technology adoption overwhelm them.
“Embracing new technology can be daunting, but it is the only way forward,” he comments. “The world of treasury and finance has a unique opportunity to not just keep pace with new trends like AI, but to be a driving force behind them.”
That’s surely reason enough for treasury to look harder at AI and how it can revolutionise the function – and to ensure it doesn’t miss the AI opportunity.
Do you think you’ll be using AI in your role in the next year or two? What are the main benefits? If not, what are the main barriers holding back adoption? We would love to hear from you – either email [email protected] or head to our Twitter page.