As an analyst at Aite Group, I frequently speak to business and technology leaders at capital markets firms to understand business strategy, the role that technology plays in that strategy, and the common challenges when striving for success.
One of the talking points often brought up during my interviews is how far along the firm is on its “digital transformation journey,” a nebulous term to describe IT system modernization and deployment of technology to drive innovative business growth. This usually entails the replacement of legacy systems, automation of manual tasks, moving onto cloud environments, the addition of business intelligence and analytics tools, or even integration of artificial intelligence and advanced machine learning techniques in core business processes.
Delving deeper into the conversation almost always leads to the acknowledgment that strong data management is foundational—not only to a firm’s competitive position but also for keeping up with the growing regulatory burden and reacting to external market shocks.
Many senior executives across the banking and asset management community would still admit that in the recent past, they haven’t placed enough investment and attention on data management. This has finally started to change, but at varying levels across the industry. Tier-1 investment banks often lead the way with strong data-focused leadership and significant investment in IT architecture, which usually lead to legacy systems and modern technology coexisting, along with custom development work and off-the-shelf applications.
As more firms take the initial steps toward digital transformation, executives who plan to complete larger-scale data projects will require a rational and cohesive data strategy. This is tough to do at an enterprise level, especially for global firms with multiple siloed business units, further complicated by the autonomy they’ve granted to their regional offices. Some of the most important lessons learned from our recent analysis of capital markets firms include the following:
Silos are a problem, but absolute centralization is not always practical: Firms often suffer from cultural, not technical, challenges of centralization. It becomes difficult to place rigid processes on teams across different functions, enforce a narrow view of data, and standardize taxonomies across business lines. Instead, federated architectures in which each business unit has defined responsibilities for its own data and can speak its own language might be the better option for a firm-wide data strategy.
A strong data governance framework can take the firm forward: Introducing governance frameworks will mean appointing data stewards for each business function. This can help establish a strong federated program at the grassroots level. Each of these stewards would then form part of a steering committee, led by the top data executives to help provide centralization and guidance. This is especially useful when trying to set common data definitions and guidance for various teams that have started utilizing new alternative data sets to derive investment signals or further insights.
Data transformation is a marathon, not a sprint: Newer technology implementations should be deployed alongside existing architecture and done so in manageable sizes to ensure operational gains can be delivered quickly. This is superior to delivering change through a big-bang approach that will likely lead to delays and lost momentum as well as demand internal project support over time.
Always link business outcomes and benefits to data management technology investments: It’s common to hear about a major project, once completed, delivering very little and becoming a “white elephant” or ornamental in nature. For example, certain firms with recent data lake projects have realized they didn’t consider the downstream applications enough and put too much focus on building data lakes, partly driven by the fear of being left behind. Data storage is but one element to consider. Firms need to consider how best to cleanse, transform, and normalize data to make it fit for purposes across business units with different requirements from trading, investments, risk, operations, and compliance.
These suggestions are but a few lessons learned. The list will continue to grow as firms adopt new technologies and uncover unexpected challenges. What is clear is that firms that do not take a step forward will inevitability lag behind, especially in capital markets with ever-increasing data velocity—and the need to process and analyze data in real time. This does not mean that firms shouldn’t take a step back every now and then to reevaluate their internal strategy; they must ensure a loud feedback loop exists between the teams that the data management function supports and the hierarchy controlling the enterprise data strategy.
