The concept of holistic customer risk has been gaining traction in the financial industry for some time now – and for good reason. Financial institutions (FIs) need to be able to assess and manage risk across all aspects of a customer’s relationship with the institution. However, even with a holistic approach, different departments within an FI tend to view risk from their own siloed perspectives. This can result in a fragmented approach, leading to operational gaps and reduced efficiency and effectiveness.
Ideally, FIs need a solution that unites all aspects of risk assessment, across who and where they are, what they do, and who they are connected and transacting with. They need a way to assess the risk of customers and activities from the top down, across entities, departments, channels, segments, use cases, lifecycle stages, and so on. All these different perspectives should feed into the overall operational risk taxonomy within the institution.
Fortunately, progress is being made in financial crime risk. Traditionally siloed compliance and fraud teams are starting to work together more closely to identify overlap areas and share a more integrated view of risk. For example, preventing the onward flow of money through mule accounts is key to managing fraud losses. It is also gaining much regulatory attention as the movement of the stolen money in itself is laundering. A convergent infrastructure approach can help reduce costs and bridge the gap between compliance and risk.
Integrated risk assessment demands a holistic view of signals
When considering a holistic customer risk assessment solution, it’s important to look for a system that integrates risk signals from various sources. These sources may include onboarding, posted transactions, in-flight payments, digital or traditional interaction channels, reference information, and external data. With the right focus, it is possible to take risk signals from any source in real-time and use them for scoring elsewhere. For example, suspicious activity seen on a mobile banking app could be shared, thus affecting the decision moments later for a scheduled payment.
A good risk assessment system should be driven by advanced machine learning techniques, including supervised techniques to learn from prior suspicious activity and semi-supervised techniques to scan for evolving risks and identify outliers and unusualness. The system should be optimized for the “needle in haystack” problems that are typical in financial crime and should be tuneable via rules to match the FI’s business model. It should also provide a single book of record for investigations, alerts, cases, and documents.
An integrated strategy across compliance and fraud also meets the need for managing infrastructure costs into the future. However, it is important to ensure that the system can be easily integrated with existing processes and tailored to fit the unique aspects of the FI’s operating model.
The move toward real-time
More broadly, every assessment of where the financial technology industry is going refers to the inexorable move toward real-time. This trend is forcing new strategic approaches on where and when risk is measured, which in turn will drive integrated and harmonized risk scoring methods and systems to provide a more complete picture of customer risk. The investment needed to support this will drive the need for common infrastructure.
Risk scorecards exist in financial crime operations today. In the future, we can expect them to be broader, deeper, faster, and more integrated. FIs will need them as the backbone for decisioning on customer lifecycle stages and events as these decisions become more automated.
Contact SymphonyAI Sensa-NetReveal to learn how we can help support your customer risk assessment needs.
