Data Governance for Asset Management
Every company needs some kind of data governance strategy. They might not call it that. Depending on their size and business model, they might not even need a Chief Data Officer. (Gasp!) But to keep a business running, you have to keep your ducks in a row. Naming ducks and keeping them organized is the essence of data governance.
For asset management companies, good data governance and engineering are especially critical. The unique mix of data use cases in asset management magnifies the risks and opportunities of applied data science. Other companies might be able to pick and choose their data governance battles. Asset management has it all: performance analytics and forecasting; home-grown and acquired data sets; upline and downline channel partners — not to mention multi-agency regulatory oversight and a demand for ever-faster and transparent decisions. Asset management companies risk falling seriously behind their peers and rivals if they don’t have at least a long-term strategy that checks all the boxes.
Your Requirements = My Requirements
For mid-level asset management companies, in particular, one principle is absolutely critical: You must accept your customers’ and partners’ requirements as your own. A lot of service delivery companies tell their clients, “Your success defines our success.” In asset management, this is simply objectively true. Investment firms and RIAs don’t demand insight into performance details just to test you. They do it because their ownclients and regulators are demanding it of them. To succeed and grow, you have to enable and support their success. Investment is a close-knit industry, and the best way to expand your client base is to inspire trust and loyalty. This means prioritizing your clients’ user experience alongside your own operational needs, keeping in mind that transparency is currency.
The good news is that getting your clients’ uses and views right will drive out other requirements and opportunities, making your whole operation more client‑focused and more effective. In asset management, for instance, information about the buying and selling patterns of fund wholesalers is publicly available. Many large firms are extremely predictable, making the same kinds of sales and purchases at the same time each year; others respond in measurable ways to leading market indicators. Of course, aggregating, analyzing, and visualizing take effort. But the intelligence to be able to anticipate your clients’ trading patterns—and redirect your sales teams accordingly—can yield dramatic savings in the first year, as well as ongoing efficiencies as those predictive algorithms get refined.
A similar “bucket-sharing” approach is important for internal data requirements, too. McKinsey touches on this in their article, “Achieving Digital Alpha in Asset Management,” when they note that successful firms have “erased the traditional boundaries between their operations and technology groups, combining their budgets and development strategies,” and “made their operations and technology capabilities central to their competitive strategies.”
A Single Data Cosmos
Uniting operational and analytic data requirements within a single data cosmos requires leadership that thoroughly understands the company’s revenue mechanisms and can assess and foresee the profit potential of new data wells and analytic tools. Ideally, this is a single person. (See “Chief Data Officer: Rule Maker or Rain Maker?”) Practically speaking, though, it can also be achieved by making sure your data stakeholders include people who are directly responsible for revenue—and that these people are tightly integrated into the data governance mission, not just spectators. When all of this works well, data stewardship will emerge simultaneously, and nearly organically, from both the technical and revenue‑making sides of the business.
It’s also important to seek out, bring onboard, and keep onboard “minor” stakeholders who will personally reap the benefits of the “unified data cosmos” in their work. Engaging these people early ensures that their processes and requirements are accurately reflected in the analytical models. Keeping them engaged, through rapid iteration and consistent messaging, helps assure the long-term success of data changes by building their skills, allaying their uncertainty, and giving them a view of their success in the new data cosmos. Practically speaking, it also helps the company recognize and allocate budget for data changes across the full breadth of the organization.
It boils down to this. In asset management — as in other industries — implementing good data governance isn’t about making and enforcing rules. It’s about creating leverage and unleashing organizational lift.
A final thought. Where in this picture is compliance? In the highly regulated financial sector, it is tempting to make compliance the chief driver for data governance, especially if the organization has been sanctioned in the past. However, it’s extremely difficult to track measurable ROI against a sanction that has been avoided—and the more successful the changes are, the harder it gets. But in a “data cosmos” focused on organizational lift, good governance begets good data—and, as a result, regulatory compliance falls out almost free of charge.