18 September 2019

Supercharging Treasury with Industry 4.0 Technology

Artificial intelligence (AI) and robotics are no longer pipe dreams for treasury. Leading treasury functions are already deploying these ‘Industry 4.0’ technologies to make their operations slick, lean and intelligent.

Contact us

HSBC France, Stockholm Branch
Birger Jarlsgatan 4, 114 34 Stockholm, Sweden

Artificial intelligence (AI) and robotics are no longer pipe dreams for treasury. As attendees at HSBC Global Liquidity and Cash Management’s recent Treasury Forum in London discovered, leading treasury functions are already deploying these ‘Industry 4.0’ technologies to make their operations slick, lean and intelligent – without significant budget or disruption.

“For a corporate to maximise its return on Industry 4.0, it needs to invest in Treasury 4.0 to support it.” This was the clear message from Lance Kawaguchi, Managing Director, Global Head – Corporates, Global Liquidity and Cash Management, HSBC, during his opening address to over 90 corporates at the bank’s recent event entitled ‘How the Fourth Industrial Revolution is Driving Treasury Evolution’.

Kawaguchi also emphasised the growing need for treasurers to think creatively around new technologies in order to add value to their organisations. This message was echoed by the keynote speaker, Diane S. Reyes, Group General Manager, Global Head of Global Liquidity and Cash Management, HSBC, who highlighted the bank’s own Industry 4.0 to Treasury 4.0 innovations.

Revolutionising correspondent banking

Next to take to the stage were Ryan McAuliffe, SWIFT Payments Innovation Specialist, SWIFT and Andrea Feay, Group Treasury Manager, NSG, to explain how SWIFT’s global payments innovation (gpi) is leveraging technology to change the face of cross-border payments.

Feay began by outlining the challenge: “Cross-border payments are a bugbear, since there is a total lack of clarity. You send the payment and simply hope for the best, and if it doesn’t arrive on time, you have to try to track it down, via the originating bank,” she explained. “We recently had a payment that was untraceable for seven weeks so to be ready for situations like this, we build in cash buffers in case a payment doesn’t arrive. This is inefficient from a liquidity standpoint but SWIFT gpi for Corporates is changing that and helping treasury departments to be as slick as possible,” she enthused.

McAuliffe went on to explain the benefits of gpi, from faster use of funds to transparency and improved remittance data. He explained how, rather than being just a DHL-style tracker for the banking industry, gpi goes further. “Our unique end-to-end transaction reference (UETR) number, which is created upon origination and maintained throughout the entire payment chain, not only provides complete visibility on where a payment is and when it has arrived, but also on deductions and FX. To help corporates with reconciliation, we also ensure remittance data is not modified throughout the chain,” he noted.

According to McAuliffe, SWIFT gpi represents a “rejuvenation of correspondent banking for the digital era – and works for the benefit of banks and corporates alike.” He shared some interesting statistics with the audience, stating that:

  • 3,500 institutions have committed to adopt gpi
  • 500+ of those are live and sending gpi transactions on a daily basis
  • Live traffic is happening in 130 currencies across 1200 corridors, with over $300bn moving on a daily basis
  • 40 % of those transactions occur within five minutes and over 90 % within 24 hours

Bringing these numbers to life, Feay explained that, although NSG is early on in its SWIFT gpi journey, the company now has a portal where it can track payments, rather than ringing up its bank(s). NSG chose to work with AccessPay on this solution. “We generate the UETRs in our TMS and then Access Pay displays that information. All MT199s are also displayed there,” she commented. “The ideal solution would be to have this straight into the TMS, and I know SWIFT is working on that with other gpi for Corporates customers, however we are pleased with the progress we have made thus far.”

Feay also explained that she is excited to see developments around gpi for incoming payments. This, said McAuliffe, will be piloted in Q3 2019 when beneficiary banks will be extended additional visibility on what’s incoming and which specific account the payment is destined for, with the ultimate aim of filtering that information down to the underlying client – providing the treasurer with even greater visibility and certainty.

Testing the waters

To further explore the potential that emerging technologies hold for treasury, a panel of industry experts then shared their Treasury 4.0 experiences and insights with the audience. Kicking off the discussion, Séverine Le Blévennec, Director, EMEA Treasury, Honeywell spoke about the use of Robotic Process Automation (RPA) in treasury.

“The great thing about robotics is that it’s not invasive, and it is honestly very quick to implement. If you compare it to a TMS implementation, it’s a dream! At Honeywell, we have several RPA projects underway, including one to optimise the intra-day cash forecast of the in-house bank. By using robotics there, we are saving two hours’ work each day for our senior treasury analyst,” she explained.

While RPA requires extra controls, Le Blévennec sees great potential for the technology. “Across treasury, I have 48 processes in mind to apply RPA to. A robot is always happy to do a boring job. It doesn’t take holidays. It’s a tool that raises the productivity and profile of the treasury team, allowing them to focus on where they can add value, such as optimising our investments and our relationship with different business partners.”

Le Blévennec also shared some statistics around the intra-day cash forecast RPA project, stating that the “ROI was 2,000% - within the first three months. And the upfront investment really is minimal. The caveat is that to reap the rewards of RPA, you have to do your homework first,” she commented.

Tackling another buzzword, AI, James Kelly, Group Treasurer, Pearson, spoke about using the technology to assist with cash flow forecasting. “If you operate a decentralised treasury operation, it can be tough to get all of the business units to contribute to a forecast – on time and accurately. It simply isn’t their priority. We were facing this challenge at Pearson and I either had the option of training people to perform better forecasting, or looking for a different solution. I chose an AI-enabled approach,” he explained.

“We have rolled out a vendor solution which interfaces with the company’s key finance and ordering systems to provide a detailed cash flow forecast in a single dashboard. We now use this as the backbone for our iterative forecasting process, but we are constantly tweaking it to improve it even further,” said Kelly. “It really has been a revolution – and I certainly haven’t had anyone from the business units say to me that they’re upset that they’re now less involved in cash forecasting! In all seriousness, this new approach to forecasting dovetails much more closely with Pearson’s digitisation journey and provides treasury with an opportunity to be a strategic business partner.”

Philip Fellowes, Regional Head of Global Liquidity and Cash Management, Europe, HSBC, then explained how APIs are driving treasury evolution. “APIs are not new, but, as Diane alluded to, the hype around them has increased with the rise of open banking. APIs will enable better communication between different players in the banking ecosystem, whether that be the banks and fintechs, banks and corporates, or corporates and fintechs. So, connectivity is evolving because of APIs and mobile cash management apps are becoming richer in terms of functionality. This is good news for treasurers, who should experience more seamless interfaces, regardless of their chosen device or partner, he noted.

Fellowes also spoke about distributed ledger technology (DLT), dispelling the myth that it is not strictly relevant for treasurers. “At HSBC, we have participated in numerous successful DLT projects – some of which are now live – in the supply chain space, and in securities services. From a liquidity and cash management point of view, we are working on solutions to change the way payments are made, whether that be through cryptocurrencies or fiat currencies. We are also considering extending our own internal blockchain solution for FX settlement into the corporate space.”

The panel then spoke about some of the risks associated with moving to a Treasury 4.0 environment, with Karen Van den Driessche, Assistant Treasurer, Avnet, commenting on the rise of cybercrime. “By digitising we are inevitably opening ourselves up more to cyber criminals. But that’s not the whole story. By automating, deploying robots, and getting rid of paper, we are also reducing the potential for human error, and that needs to be taken into account in the risk equation.”

In terms of preparing for Treasury 4.0, the panel also debated the need for different partner ecosystems going forward. “We would like to see more co-creation between corporates and banks, as well as vendors,” said Van den Driessche. “SWIFT is doing a good job of bringing different players together with initiatives such as gpi and its KYC Registry, but there is room for more collaboration in an increasingly digital and ‘open’ world. And it would make life much easier for treasurers to have everyone working together, on standardised solutions and innovations.”

Fellowes concluded the panel session by re-iterating the great potential that Treasury 4.0 holds. “As we have seen, robotics can be an extremely powerful tool in treasury – not replacing people but enabling them to perform better. Likewise, AI can improve treasury’s performance significantly, and turn data into actionable insight. Yes, there are risks such as cybercrime to consider, but the opportunities are arguably bigger, especially if – as the panellists have said – banks, vendors, fintechs and corporates all work together to build an optimal Treasury 4.0 environment.”

Forget science fiction

The final session of the day saw H.P. Bunaes, General Manager, Financial Services, DataRobot, explain several use cases for AI and machine learning (ML) in corporate treasury, starting with receivables management. “Every treasurer wants to know how quickly they will be paid. By taking all of your historical data from every buyer, it’s possible to build a machine learning model and run it against your current receivables. The model will then plot out exactly what your payments curve is going to look like,” he explained.

Receivables data can also be fed into a wider cash flow forecasting model, said Bunaes. “We find it works best to run separate models for inflows and outflows, and then aggregate those to produce the complete forecast.” While this makes the process sound very simple –there is inevitably a great deal of work involved and huge amounts of data required. “The more granular the data and the more data you have on each transaction, the higher the fidelity of the signal, and the more accurate the forecast will be,” he explained. “It is also important to take time to teach the model and tell it anything that it can’t intuitively know, such as known non-recurring transactions.”

Fraud detection and prevention is another interesting use case for ML. “As fraudsters become more sophisticated, corporates and banks must be more creative in detecting and preventing fraud in real-time. One of the best ways to do that is to use both supervised and unsupervised learning to create a multilevel detection system. This involves building models to filter out false positives, continuously learn to detect new fraud schemes, and to look for anomalies which may require closer scrutiny.”

In addition, Bunaes mentioned hedging as a potential area for using machine learning. “When you hedge in order to mitigate enterprise risk, you actually create counterparty risk. Using machine learning, banks can model when they have a deteriorating counterparty, to estimate the default risk for all of their open derivative positions, for example.” For many of the day-to-day challenges treasury is looking to solve, it is possible that machine learning could provide a solution. But wherever machine learning is applied, it is important to remember that “the model is only going to be as good as the data that you’ve used to train it,” warned Bunaes.

Levelling up

Having clean data is indeed one of the cornerstones of Treasury 4.0, explained Kawaguchi in his closing remarks. “The ultimate goal is to spend less time wrangling data, and more time acting on it – to the benefit of the business.” Other potential benefits of Treasury 4.0, said Kawaguchi, will include “a more secure environment for treasury practitioners; one where technology works for them, not against them.”

Greater transparency and efficiency should also result from the shift to Treasury 4.0. “And at HSBC, we are embracing emerging technologies in order to garner those efficiencies on behalf of our clients. We are also proactively helping treasurers to transition towards the new digital environment, not only through innovative solutions and proofs of concept, but also through knowledge-sharing at important forums such as this,” he concluded.

For more information on how HSBC can help meet your needs please contact your local HSBC representative or visit gbm.hsbc.com

You are leaving the HSBC Commercial Banking website.

Please be aware that the external site policies will differ from our website terms and conditions and privacy policy. The next site will open in a new browser window or tab.

You are leaving the HSBC Commercial Banking website.

Please be aware that the external site policies will differ from our website terms and conditions and privacy policy. The next site will open in a new browser window or tab.

You are leaving the HSBC Commercial Banking website.

Please be aware that the external site policies will differ from our website terms and conditions and privacy policy. The next site will open in a new browser window or tab.