Michael Shashoua: Valuing Data and Making Data a Value

Amid all the issues affecting reference data—identifier and messaging standards compliance, corporate actions, tax collection and compliance, the quest for better data quality, getting transparency in sourcing data, and the need to track collateral in transactions—it can be easy to lose sight of the major issue on which achieving any of these things rests.

That issue is coping with the costs of data management and compliance, and managing firms’ resources to carry out those data efforts. Speakers and panelists at last month’s Toronto Financial Information Summit (FIS), hosted jointly by Inside Market Data, Inside Reference Data and WatersTechnology, reminded attendees that this applies universally in the data space, as well as in trading operations.

Bettina Wadehn, program director at the Canada Pension Plan Investment Board (CPPIB), frames the issue in terms of thinking about the value of data. “The true value lies in the analytics, and ultimately the insights and the decisions that will be made based on this data—or driven by it,” she says. CPPIB has been centralizing its data to support its investing choices and better understand its risk exposures and evaluate its performance against benchmarks. “The true asset we get from data is being able to make better investment decisions for the future of the enterprise.”

Donna Rudnicki, Toronto-based head of data management at RBC, backs a centralized model at that firm, rather than loosely federating several data sources. Managing data is necessary to derive value from it, she says. There is a temptation to try to fix every issue surrounding data, but firms must resist this. “Think big, but focus small. Pick a few key elements to focus on, and mobilize on those,” she says.

Strategizing
After choosing what data elements to focus on, as Rudnicki advises, firms have to then strategize about how they will address these elements, which Wadehn points out. For CPPIB, the choice has changed from being between one vendor and multiple vendors, to a choice of whether to outsource at all—or instead, building the necessary data management systems internally.

With data management having to support risk management, and keeping budget constraints in mind, data perfection is not always necessary.

Also, once a firm knows what its data focus will be, it must devise a strategy covering data governance, accessibility, sharing procedures, data quality and security, according to Rudnicki. “The strategy declares the organization’s conscious intent,” she says. Strategy, in effect, is the path a firm takes to get value out of its data resources.

With data management having to support risk management, and keeping budget constraints in mind, data perfection is not always necessary, says Ash Tahbazian, a senior vice president at State Street. Data users, particularly risk managers, would much rather “have something they can move forward with,” he says.

Organizing
Robert Neupauer, a director at UBS in Toronto, who focuses on hedge fund administration, says budgeting at his firm now requires preparedness for sudden, new regulatory requirements. “If the deadline for something is June 15, it’s not like you can wait for the 2014 budget,” he says. “You have to act fast. This means being flexible and able to reallocate from lesser priority projects and then put the money where it matters.”

Oliver Salvati, director of data management, group operations at Sun Life Financial, a Toronto-based global investment firm with several Asia-Pacific offices, says Sun Life looks at its “organizational construct” for data—whether it has the right organization in place to act on the framework it has established for data governance.

The takeaway, when one thinks about what budgets or costs permit for data management improvements—or at least readiness—is that a blend of centralizing data, setting data management frameworks and strategies, prioritizing the most relevant data, and choosing data sources, is needed to maximize resources for the best possible reference data.