The Annual EuroFinance Award for Treasury Excellence: Stripe
The payments company has reinvented the treasury function with the help of data science and software engineering. Nicholas Dunbar reports.
It was an offer that Chris Van Woeart couldn’t resist. “I wasn’t thinking about a treasury role when I got the call” he says. “But the opportunity to build treasury-related solutions from scratch was too exciting to pass up”.
The call was from San Francisco-based payment processing firm Stripe, which was founded in 2010 and is already worth $20 billion. Van Woeart was previously in charge of the liquidity pool at Wall Street giant Goldman Sachs but soon after the move to Silicon Valley he was questioning everything he had done before. “Being in a really fast-paced, innovative environment with a lot of technology was a big difference” he recalls.
The company’s core product is an application program interface (API) that allows customers to embed secure payment facilities in the code of their websites. These customers, who are based in more than 25 countries and number above one million, are effectively outsourcing components of their treasury operations to the company.
They include other fast growing start ups such as Lyft and Spotify, that depend heavily on Stripe’s API to gather real time payments from customers and disburse them to suppliers
Being a privately held company, Stripe doesn’t disclose financial results. As a proxy, one can use Netherlands-listed payment processing firm Adyen, which trades at a 288-times multiple of earnings and processes €70 billion of transactions per year. If Stripe had the same multiple, its current earnings would be $70 million.
Where engineers and data scientists outnumber finance professionals
As a technology startup, Stripe taught Van Woeart to view treasury as a problem in software engineering and data science. “Before I got there, our engineers started to solve some traditional treasury problems through automation. Even now, we have far fewer finance professionals on the team than engineers and data scientists”.
“At Goldman Sachs why did I have so many analysts?” he muses. “Upon reflection, they were going into databases each day, downloading data into spreadsheets, performing analyses and then sending out reports to management. At Stripe, we took a different approach.”
“We could start with a blank sheet of paper and leverage a lot of our technology and payments architecture. That was a contrast to inheriting a lot of legacy systems and processes that many larger institutions tend to be saddled with. At Stripe data is near real-time and accessible, so our Treasury engineering and the data scientists can be incredibly high-leverage in partnering with us to automate our processes and analytics”.
“As a result we’ve seen efficiencies. For example, no one is in spreadsheets trying to do cash or currency forecasting on a daily basis”, according to Van Woeart. “Instead we’ve worked with our data scientists to build very predictive models with good error bars that are very dependable and able to learn as our business changes. You can pick up things like seasonality and the model recognises and learns from it. We’ve been able to automate a lot of the work that you otherwise would’ve had to staff with finance professionals”.
Nor does Van Woeart need to follow the well-trodden path of buying a treasury management system. “When I joined, we asked whether we needed a third party Treasury Management System (TMS), but in the organisation there was already the semblance of what a TMS would do – it was just a bit decentralized. So we just had to work to stitch it all together”, he says. “In the end, people who’ve joined Stripe’s Treasury team with TMS experience actually say that what we have is better than what they’re used to”.
Traditional treasury problems meet new solutions
However, Van Woeart cautions that even allowing for the benefits of technology, the fundamentals of running a treasury at Stripe aren’t that different from his previous work at Goldman Sachs.
“Many of the problems are very similar from a treasury perspective, with the pillars of liquidity management and risk management being predominant”, he says. “What are the risk models and stress scenarios that we need to be prepared for in order to support our customers for the long run? There are so many businesses that are now dependent on Stripe to run their money flow, so we take risk management very seriously”.
While the fundamentals may be the same, when it comes to banks or regulators, the approach is changing and evolving, according to Van Woeart.
“For example, we are now looking to work with our bank partners to leverage APIs and computers to handle managing our currency needs, rather than hopping on the phone as in the old days”, he says. “Or how when it comes to fraud we are machine learning focus. For example, you can find correlations of fraud transactions based on the behaviors of someone making a charge or the end-point from which the charge is originating. We’re working with our partners to think outside the box in these areas as new technologies allow us to do so”.
As a parting shot, Van Woeart urges his peers in treasury not to fear new technology. “At the end of the day we’re not moving wheel barrows of gold, we’re actually moving numbers or digits”, he says. “That can be solved through technology. And, if it’s done through APIs, it can be real time, efficient, and transparent.”