Kyriba strategist: tech trends that will dramatically impact treasury.
Corporate treasury departments are likely to look much different within even a few years. That’s because a variety of technologies are unfolding that will make treasury functions more efficient, cost effective, and—if possible—leaner still in terms of staff.
That was the message delivered by Bob Stark, vice president, strategy, at treasury management solution (TMS) provider Kyriba. Speaking at a recent NeuGroup T30 meeting, Mr. Stark, whose job entails no sales responsibilities, presented his research on the latest technologies. Unsurprisingly, one of those was bitcoin, which at the time had just dropped in one day from a record high of just over $11,000 per coin to below $9,400, after starting out the year at $997. It’s since climbed above $17,000.
“Given that volatility, in my opinion it’s a commodity because it has very limited supply and that supply is not increasing in any meaningful way,” Mr. Stark said. “There’s no scenario where it’s necessary to pay in bitcoin, unless in response to ransomware, and that’s why the demand seems unrealistic.”
Nevertheless, he said, the concept of cryptocurrencies will likely flourish, mainly because they are a critical component of distributed ledger systems for which a multitude of use cases are now being tested and developed. For example, many nonbank payment channels are using distributed ledgers to offer cross-border payment services free of the fees often levied in the traditional banking system, because they can make money solely on the FX spread to generate revenue.
“Many are doing this by using their own cryptocurrency. The actual conversion is in their cryptocurrency, and they will net settle into fiat currencies at the end of a day or other period.,” Mr. Stark said. “It’s early days still, but this is the kind of use case we can expect to see more of.”
Mr. Stark said that the migration of cryptocurrencies from peer-to-peer payments for individuals to the corporate community will elevate overall interest in distributed ledger technology, and some companies are exploring the possibilities. There are large online retailers, for example, that are rumored to be exploring how to pay suppliers in digital currencies.
Eventually, a standard cryptocurrency may emerge, pushed by the market or regulation. So far, however, there’s been little demand from Kyriba’s clients for its TMS to support cryptocurrencies, Mr. Stark said, and while some corporates such as Starbucks are accepting bitcoin, the digital money is typically converted immediately into a fiat currency.
Their volatility makes it unlikely that bitcoin or ethereum, another relatively longstanding cryptocurrency that offers business friendly functionality such as self-enforcing “smart contracts,” will become that standard.
Allison Berke, executive director of the Stanford Cyber Initiative, noted that distributed ledgers often use the cryptocurrency to levy a cost for getting transactions approved, and using a volatile cryptocurrency would make forecasting those costs over a period of time difficult. “If I were building a blockchain for my business, I might not want to use bitcoin or ethereum, given their volatility,” she said.
Likely impacting corporate treasuries sooner will be the increasingly widespread use of application programming interfaces (APIs), which provide a framework to integrate applications more easily. Mr. Stark noted that Europe’s unfolding PSD2 regulations aim to give nonbank service providers better access to bank payment systems, providing corporates with more payment options. Several large US banks, he said, are already piloting API technology to comply with the new regs, and they and other banks are expected to spread the technology outside of Europe. So far testing has been limited, generally with a single bank customer, he said.
“APIs will offer another alternative for corporates to connect to banks for reporting and payments, potentially rivaling existing solutions such as file transfer protocol (FTP), SWIFT, or regional protocols like EBICS,” Mr. Stark said. He added that in terms of corporates connectivity with their banks, a move to APIs could enable a shift away from batch processing to real-time updates for information such as bank reporting.
APIs also facilitate cloud-to-cloud integration, in which multiple applications are configured to share data in the cloud and communicate with one another. APIs permit pre-built integration rather than creating a unique file interface for each project, reducing implementation time and cost and standardizing onboarding and security.
Mr. Stark said banks are now examining whether communicating via APIs presents an opportunity to move away from FTP and batch systems. When they’re on APIs, he said, banks will have to decide how to offer and price the more efficient reporting services, but that may still be a couple of years away.
“Banks are starting to ask their customers questions, such as how important it is for them to get real-time reporting vs. the batch reporting they get now, so they can understand what the technology’s value is,” he said.
The biggest impact on treasury, Mr. Stark suggested, is likely to come from robotics and especially when artificial intelligence and machine learning are applied so that software begins learning tasks and adjusting to changes on their own. Today, companies can buy robots—essentially algorithms designed to perform specific tasks—to automate processes that currently require treasury staff participation.
“Not a lot of differentiation from what TMSs now do,” Mr. Stark said. “But AI will distinguish between what we see today in treasury automation and what we’ll see in the future.”
Machine learning, a component of AI, takes robotics beyond predictable programming and enables software applications to evolve without human reprogramming, replicating the experience factor. Mr. Stark said that limited applications of machine learning already exist today in other areas, but not yet corporate treasury. He added that in the short term it likely will be used to detect certain types of transactions, such as fraudulent ones.
A T30 member said payment batch files could first go through a robotic system that looks at each one for potential fraud, because it has already been trained on billions of transactions, and it has been told which ones have been fraudulent.
“That’s a perfect example,” Mr. Stark said. Based on the data, here’s a scenario that could hypothetically happen and we want to stop it. Or maybe payroll is on average X dollars and it suddenly doubles in value; that may not be wrong, but it’s important to take another look.”
In the longer term, machine-learning robotics can be applied to anything that can be optimized through a person’s treasury experience, such as cash forecasting, FX hedging and tax optimization.
“Increasing productivity is the tip of the iceberg, and once the productivity gains are achieved it can be applied to more value-added functions,” he said, adding, “once that happens, that’s where treasury jobs will start to change.”
Mr. Stark said certain tasks will evaporate or be replaced, but automation should allow treasury staff to focus on more strategic decisions. “Of course, one can argue that a robot could end up making those decisions instead of you, and then more of treasury’s day is taken away by robots, and there will be fewer treasury jobs,” he said.
He added that treasury may have more time than other departments before facing that reckoning, given treasury teams are already small and members are highly specialized, whereas larger departments with more predictable tasks will be software developers’ early targets.
Another T30 member said that once the software arrives and the cost analysis is done, if implementing robots becomes far less expensive than hiring people to do the task, the decision will be a slam dunk. Fortunately, treasury-related robotic software now is not yet driven by AI, and companies are wary about purchasing it because it may have to be replaced in a relatively short time.
“But the evolution will happen, and it won’t take long,” Mr. Stark said.