It was my first day in Settlements and I was getting an earful about a trade that failed to settle. It turned out to be a missing account number in the SSI. I remember thinking, if there was a way to notify me and my team in advance, we could have fixed the issue earlier and the trade would have gone through
Nick Smith, Principal Business Process Transformation
Intro
… a conversation between Brett Adams, CTFIS’s Head of Product, and Nick Smith, who leads CTFSI’s Business Process Transformation team as they discuss the origins of CTFSI’s Prediction Fails Management solution for SSI fails.
Challenges in Traditional Trade Settlement Management
Brett: Nick. Tell us about some of the challenges you’ve seen in traditional trade settlement processes?
Nick: Absolutely. In the past, I’ve been part of teams that were overwhelmed with manually handling trade errors. The process is manual—it was highly error-prone as a result. We often had to reject hundreds of trades daily due to manual errors, leading to delays, increased costs, and higher operational risks. It was clear that a more robust solution was needed.
Identifying SSIs as a Key Area of Improvement
Brett: That sounds incredibly frustrating. How did you come to realize that addressing bad SSIs could be an opportunity for improvement?
Nick: It was through firsthand experience. I remember my first day in Settlements, someone was giving me an earful about a trade failure. It turned out to be a missing account number in the SSI. This small issue caused a significant delay and client frustration. Over time, I saw that many trade failures could be traced back to SSI errors. These weren’t isolated incidents; they were pervasive problems affecting overall efficiency daily and over thousands of trades. The biggest challenge is banks are adding more complexity to their SSI standards to manage their own operational models. There is no SSI “rule book” to look up how each bank wants their trades to settle.
Brett: What was the turning point for you in deciding to focus on this issue?
Nick: The catalyst came when we had to manually reject hundreds of trades due to one missing SSI. One day stands out—I had just gotten off the phone with an irate client whose trade had failed yet again. I thought, ‘there has to be a better way to handle this.’ When I came to CTFSI, I was asked on my first day if there were any processes where leveraging AI could be added into a Settlements workflow. My first thought when brainstorming with CTFSI Labs was to use machine learning to learn all the banks’ rules of SSI formats and to catch SSI issues while there was still time to fix them.
Historical Context: Why Wasn’t This Solved Before?
Brett: Why do you think organizations have struggled to solve these issues traditionally?
Nick: Well, it is not because the issue isn’t widespread! I’ve seen fails in the hundreds of millions of dollars stemming from bad SSIs impacting trade settlement. Personally, I think operations teams have their hands full running their day-to-day core operations, focusing on urgent issues. At the end of the day, the client is the one who often is providing the bad SSIs to the bank and there hasn’t been a way previously to validate they are correct until the trade fails. And attempts to utilize new technologies like AI often start at an enterprise level with little to no concept or appreciation of how to connect these capabilities to the last mile of operations workflows.
Leveraging AI for Trade Settlement
Brett: How have you been able to leverage your experience managing trade settlement teams to develop a preventative fails solution at CTFSI?
Nick: At CTFSI, we recognized the need for a transformative approach. We integrated AI into trade settlement workflows, automating routine tasks and employing predictive analytics to foresee and mitigate potential trade failures. Our AI models analyze historical trading data to identify patterns and anomalies, enabling proactive management within frontline teams. The biggest distinction compared to other AI Trade Fail predictors is that ours can tell you the reason for the fail prediction for SSI issues.
Innovative Solutions and Real-World Impact
Brett: Can you give us an example of how these AI solutions can impact trade settlement?
Nick: Our AI-driven solutions can detect discrepancies in SSIs and automatically flag them for review, reducing the dependency on manual checks. This advancement has the potential to improve the speed and accuracy of trade settlements. Implementing these AI solutions is a game-changer for last mile operational challenges.
The changes can seem small, but I remember one day when a typically complex and error-prone batch of trades went through without a hitch—it was a revelation! No seriously, I was so happy to have managed to cure this seemingly small problem, but it had such a positive impact on operations. And while the tightening T+1 window is adding additional challenges and making it that much more important to get it right, with the types of solutions we are creating, I am increasingly optimistic that the notion of one-day eliminating SSI-driven trade fails completely is no-longer fiction and we will be able to experience that feeling with increasing frequency in the days ahead…