Like it or not, the global treasury ecosystem is currently in the midst of a sweeping technological transformation.
Treasury teams are being forced to take on a greater strategic role within their respective organisations while simultaneously being pushed to whittle down overheads and produce greater value using less resources – and for most organisations, the solution to these mounting pressures revolves around innovation. Corporates know they’ve got to become more efficient, and increased automation is the most obvious pathway forward.
That’s where robots come in.
It can be surprisingly difficult to try and disentangle the concept of ‘robotics’ from Hollywood sci-fi fantasy – but within the context of treasury, robotics is actually just a simple blanket term used to describe any number of software programmes that facilitate the automated execution of a process using a series of rule-based decision-making algorithms. Many of these computer-assisted workflow tools have been in circulation for years now, and corporates are finally working to capitalise upon their actual potential.
According to researchers at Capgemini, most corporates have already gone a long way toward fully automating a variety of individual processes – while seven in 10 finance professionals say they recognise the potential for full automation within the next three years.
Yet for those corporates keen on getting a head start with their respective digital transformations, there’s no need to wait three years in order to automate treasury processes. Enterprising fintechs have already brought a number of increasingly advanced treasury bots to market that can be easily integrated into existing TMS and ERP systems – and while plenty of teams are still unprepared for a fully automated mapping of end-to-end processes – existing treasury bots are already making a huge impact in the back office, C-level and everywhere in between.
Tomorrow’s bots today
Robotic process automation (RPA) has only entered the mainstream over the course of the last five years or so, but it’s totally streamlined the back office. RPA bots are designed to utilise pre-set guidelines in order to automate labour-intensive and highly repetitive tasks like manual cash pooling, auditing, financial closings, data consolidation, prepping FX exposures and invoicing.
This automation is absolutely critical for treasury teams working to demonstrate strategic value at the C-level.
According to Nordea’s Future Treasury Report 2018, treasurers are planning to spend less time than ever on back-office tasks over the next six years – and by calling upon automated RPA tools they can trust, finance professionals can subsequently expect to free up time and bolster efficiencies by allowing bots to take on these laborious data entry tasks.
Not only does this speed up internal efficiencies and enable treasurers to focus on higher value processes, but it also saves corporates a pretty penny on labour and otherwise costly human errors. Companies that use RPA bots to handle accounts payable functions process invoices 43% faster and for a cost of 40% less than those companies that invoice manually – which is why it’s hardly surprising the market is already littered with tried-and-tested RPA bots.
Entry-level RPA solutions like IBM’s Business Automation Workflow Express can help teams to set up a fully automated end-to-end process flow and integrate those workflows into existing ERP systems without any friction. Meanwhile, providers such as Automation Anywhere are able to take RPA functionality to the next level by enabling corporates to create their own, bespoke RPA finance bots that are built to address unique corporate challenges specific to each team.
The next level of robotics in treasury is all about intelligent cognitive automation, and it’s a bit more exciting. Using artificial intelligence (AI) and machine learning, second generation treasury bots are now able to take on far more complex tasks and draw from unstructured big data sources in order to make decisions or more sophisticated forecasts totally unassisted.
AI-powered treasury bots are already proving particularly effective in areas such as automated liquidity planning and FX exposure planning, as they’re able to draw from a huge range of untraditional data sources such as news reports, incoming market data or even corporate tweets in order to inform and improve predicted movements. Treasury bots are then able to spot and learn from any potential patterns – subsequently helping to detect fraudulent transactions and facilitate better-informed decision making at the C-level.
Treasury robots that utilise machine learning and AI functionality are already hitting the market in the form of advanced virtual assistants.
Last year, JP Morgan rolled out a new pilot AI bot that enables corporate treasury clients to bypass its online cash management portal and simply ask a voice-activated assistant such as Amazon’s Alexa to send wires, export data or get balance information in real-time. With the help of machine learning, the bot is built to bank commands and then learn from client patterns in order to start prompting treasurers with suggested reports or activities.
Bank of America Merrill Lynch has developed its own chatbot assistant, while NICE’s own finance bot promises to reduce manual and spreadsheet-driven processes like fraud analysis down to a two-minute conversation. That reclaimed time is absolutely invaluable to already overstretched treasury professionals – although it’s worth pointing out that the industry as a whole has yet to embrace this brave new world in treasury robotics.
Barriers to adoption of robots
As technology continues to advance at breakneck speed, there’s certainly a degree of inevitability when it comes to the increasing involvement of robots in treasury.
According to researchers at EY, RPA software has the power to help teams regain up to 70% of the time they would have spent on particularly manual data processes – which is probably why Capgemini estimates this new generation of intelligent automation in treasury should add an extra $512bn to the global revenues of financial services firms by 2020.
Around 64% of companies that have deployed treasury bots have reported improvements in customer satisfaction, and 35% of financial services firms have directly traced up to 5% of recent topline growth back to these treasury automation tools. Yet for all that success, only 15% of corporates are currently reporting mature use of RPA in treasury, and just 5% of treasury professionals say their organisation has mastered the use of AI bots.
So, why have treasury bots still not cracked the market? While a certain degree of sluggishness can probably be pegged down to a general hesitancy to embrace full automation, a high proportion of corporates simply do not have adequate enough infrastructure to facilitate the effective use of robotics.
In order to successfully carry out automated processes, spot patterns and make decisions, treasury bots need a constant flow of inconceivably large data volumes – and the only way programmes can gain access to that much data is by plugging into existing cloud-based infrastructure. According to research conducted by IDG, one in four organisations still haven’t integrated a cloud-computing solution, which means a quarter of firms couldn’t yet integrate treasury bots even if they wanted to.
Yet as the industry crawls ever-closer to full cloud adoption and treasurers begin to untap the true potential of big data, mass uptake in the use of treasury robots and bots will almost certainly follow. Robots aren’t around to make treasurers obsolete or hijack corporate decision-making processes – they’re designed to make life easier for teams and deliver much-needed added value at the strategic level. If nothing else, the incredible success of early adopters certainly seems to suggest that treasury bots deliver on those promises.