The financial services industry is fundamentally shifting. Fee compression, compliance and regulatory changes, heightened volatility, and burdensome technology costs have added layers of complexity in recent years. At the same time, equity trading volume is expanding at a record pace, growing from an average of 9 billion shares a day in 2019 to 14.7 billion in 2021, often presenting unwelcome transactional errors and reconciliations.¹ Despite these challenges, asset managers have continued to embrace transformation as they pivot their strategies and grow their businesses. It can be argued that there’s one key driver to that transformation: operational efficiency.

For asset servicers, the pursuit for greater efficiency is no different. Like many in our industry, BBH has spent a considerable amount of time exploring how automation, artificial intelligence (AI), and machine learning can enhance our work as servicers of the back and middle office. The promise of automation, of course, is that it can substantially augment the work of human employees in a fraction of the time and cost involved, while also reducing risk. What’s equally critical is that these solutions can scale and grow to meet manager’s needs as business expands. In our experience, we have achieved the timing and scalability benefits through the application of automation. Even so, the advent of automation and AI tools does pose major questions for asset servicers and managers alike. How do you ensure you invest in the most impactful solutions? What organization structure can boost end-to-end optimization for both the asset servicer and their clients? Do operating teams now need to hire a cohort of employees with advanced engineering degrees?

As we reflect on our own AI and automation journey, we’d like to share our perspective on what has helped us eliminate millions of exceptions and thousands of processing hours while generating meaningful insights for our global servicing teams and clients.

Start Small, Think Big

In 2017, BBH created a dedicated team of Business and Systems resources tasked with testing the use of emerging technologies in solving operating challenges. While many firms — including BBH — were actively utilizing algorithms in the front office, we believed there was opportunity to evolve our middle and back office functions, tackling the operating and control challenges that existed outside of our core platforms.

Our hypothesis at the time was that these newer tools and technologies could help us drive efficiency, control, and mitigate risk in lighter and faster ways than traditional methods. We also believed that by creating internal competencies in this space, we would achieve broader benefits across the organization through the reuse of solutions that could address operating problems for us and our clients.

We soon realized that deep business subject matter expertise (SME) was equally important as technical know-how, both in terms of opportunity assessment and in solution design. As a result, we paired our technical “lab” with a business SME team to work side-by-side. The result is a highly integrated and connected business and technology team that is centered around both transforming process and products, as well as transforming people to enable our workforce of the future. Three years later, the benefits of this structure are profound. We’re able to generate more ideas, align our capacity to execute on the most impactful concepts, and help to more quickly justify the efficacy of our investments. We also created and implemented strong governance and controls that ensure our solutions are reusable, configurable, and quick to market.

A recent example is how we achieved 90% end-to-end automation in the bank loan cash settlement process for both BBH and our client. The solution was developed and deployed in under six weeks by combining our data transformation and process digitization capabilities with our Infomediary platform.

Another example of this in practice is our NAV transformation journey toward what we refer to as “STP NAV.” While we had a high degree automation in our NAV process, there were still thousands of exceptions at the instrument and fund level that required investigation. Given the compressed time cycle, this exception process can be resource intensive, so to execute against our vision of a STP NAV, we focused our AI/ML applications on streamlining this processing.  

This transformation began with the automation of cash reconciliations using our LINC platform which applies natural language processing and machine learning to automatically classify, escalate and, where appropriate, resolve breaks. We then progressed to reducing security level exceptions by leveraging our ANTS platform which uses predictive analytics and anomaly detection in our security pricing process to eliminate the noise of reviewing false exceptions. This helped eliminate 80% of our pricing edit valuation volume.

 Most recently, we applied these methodologies and tools to fund level reviews with our Guardrail application. This application monitors NAV movements against a benchmark and historical performance and flags variances, while also identifying possible reasons for abnormal deviations, and brings those automated insights for BBH analysts to review.

These three applications create an AI ecosystem that can leverage data between each other and review the NAV as whole providing users with a fully integrated NAV oversight process.

While transformative, this was a multiyear journey, in which we challenged ourselves with the question of how we could more quickly scale our capabilities across the enterprise. The answer was to put the solutions in the hands of those closest to the business challenge.

Make Automation Accessible Through Democratization

At its core, democratized automation means making automation accessible to everyone — no engineering degree or extensive technology background required. The idea is that the employees or business teams who perform a given process are the ones best-positioned to identify opportunities for automation in their work — while also mitigating demand on our specialized Systems teams — and that these incremental enhancements enable a culture of firmwide innovation.

To that end, we deployed an array of self-service automation tools for business users that addressed a common set of challenges. We estimate that 80% of the issues that business users face boil down to consuming unstructured data, reconciling that data, identifying exceptions, and addressing disconnected data. This self-service suite enables those closest to the business to configure solutions to address those challenges. Since launching these applications, we’ve seen exponential adoption across the organization which is now being utilized by over 60 different business units with more than 750 workflows automated and a robust pipeline of opportunity for continued scale.    

In addition, this platform can be leveraged to access and explore applications of our AI/ML models.  When a use case is identified, our teams can test it with real data in real time, evaluate the results, and iterate on their approach. We’ve created an ecosystem between business users and IT that allows us to learn fast thereby saving time and cost prior to validation, and to scale quickly. We can then apply effective trials to other similar business processes. This approach helps to mitigate the “leap of faith” that many organizations struggle with when seeking to justify a business case and move beyond proof of concepts to production grade AI applications.

What Are the Benefits to the Client?

Above all, the power of automation allows our service teams to surface and focus on the most impactful issues to our clients. These tools allow us to react more quickly, anticipate issues, and magnify weak signals. But what separates our automation solutions from many others is that we’re marrying advanced technology with true subject matter expertise. Finally, by using these types of enterprise applications, we can now document and systemically capture how and when certain data points are used. We can then call on this repository of past use-cases to help asset managers understand what data is most important so that it can be accessed quickly and repeated when necessary.  As we consider the pace of change in how our clients access, manage, and ultimately consume data, these capabilities are imperative for any service provider to remain agile in meeting these evolving needs while also creating opportunities to co-create new products and service offerings.