Treasury in CEE: Exploring the Digital Frontier
By Eleanor Hill, Editor
The European Association of Corporate Treasurers (EACT) recently teamed up with TMI for the 2018 CEE Conference in Budapest – covering a whole spectrum of treasury topics for the region, from FX to geopolitical risk. One of the key plenary sessions explored innovation and digitisation in corporate treasury with a focus on artificial intelligence and cybercrime. On the panel for this discussion were Daniel Chapman, AVP Strategic Accounts EMEA at HighRadius; Matt McQuillan, C2FO’s Managing Director, UK; and László Tóth, Head of Corporate Banking and Global Subsidiaries Group, Hungary, Western Balkan and Baltics at Citi. The moderator for this lively session was TMI’s CEO, Robin Page.
AI and robotics
Advances in artificial intelligence (AI) and robotics inevitably come up in any conversation about digitisation, but how can they be applied specifically to treasury operations? This was one of the first questions addressed by the panel, with Chapman commenting that AI has significant potential in areas such as cash application and invoicing.
“This is about taking huge data sets that exist across your customers – swathes of information about how you’re paying them, how you’re invoicing them, how they’re paying you – and using AI to be able to intelligently predict when things are coming,” Chapman said. “I see AI playing a much more significant role in cash forecasting going forward, helping treasurers to more accurately see what their customers are doing and to better project future cash positions.”
McQuillan then explained that C2FO uses AI to “learn what cash is worth to suppliers”. As a dynamic discounting marketplace between the supplier and the customer, the company receives information every second of the day from thousands of suppliers who make offers to their customers on invoice discounts for early payment. The marketplace uses machine learning to take all of that big data and essentially “pick the invoices with the greatest potential to get the best return and to pay as many suppliers as possible with the cash available,” he noted.
Another interesting point raised was that, in many large companies, people who work next to each other might not really talk to each other – and AI could help bridge that gap. Said Chapman: “It’s always fascinating when I sit by someone who’s in cash application and they’re trying to apply cash as quickly as they can. Meanwhile the deductions person who sits across from them is drinking from a fire hose of unstructured data and information – getting overwhelmed. “When you start doing the mathematics of where the information is, you see that if the cash application person would code the deductions a little better, the deductions person would be freed up to bring greater value to the organisation. AI can fit into that to help reduce manual burden, but also to help determine where people should focus their efforts in order to reap efficiencies.”
One of the unfortunate by-products of digitisation is the growth of cyber-fraud. Tóth noted that Citi has three large centres around the world – in Budapest, New York, and Singapore – that operate as a cyber-security function for the bank’s entire global network. Here, the different functions are dealing with various aspects of cyber-security, he said, from market intelligence to network performance monitoring, detecting intrusions and much more.
Nobody can have a completely watertight IT system, he said, but it is imperative to have a strong defence mechanism to deal with an event if and when it occurs, especially for treasury, since payments will still need to be made and received. “Communication between the different parties – internally and externally – is therefore one of the most important elements of a successful cyber-security system. It’s not necessarily about just preventing the intrusion, but much more about managing the damage that is caused if an intrusion happens and identifying any cyber-actor as soon as possible,” he cautioned.
Tóth also commented that, as with all systems issues, the weakest link when dealing with cyber remains the human element. He explained how Citi uses a group of ethical hackers to target their system internally, probing staff with different phishing exercises, for example. This enables the firm to identify how strong the weakest link is, and also to offer education to staff on a regular basis as a way to pre-empt any behaviour that could be exploited by cybercriminals.
Chapman then explained how machine learning can also help corporates to protect themselves against potential cybercrime, through solutions such as sanctions screening. He spoke about how Citi Smart Match – an intelligent reconciliation solution developed in conjunction with HighRadius – can use a company’s sanction list to automatically scan customers. “If you receive a payment from a company or individual on the list, the system will flag them so the payment is not applied immediately. A member of the team can then investigate further before taking action.”
The same goes for suspicious payments, Chapman explained. “Perhaps you were expecting a payment from a large company in a particular country but were actually paid by a third party from elsewhere. The system is able to create rules so that if you’re not paid by exactly who you think you’re going to be paid by, even if the invoice number and customer number are correct, the transaction will still be flagged so that someone can assess the situation and make sure that there are no issues before the payment is applied.”