Michael Shashoua: The Culture of Data Management

At the Asia-Pacific Financial Information Conference in Hong Kong last month, I found that certain data management considerations are universal. That was evident in a panel discussion I moderated at the conference, concerning enterprise data management, in which service providers joined practitioners from Credit Suisse and JPMorgan Chase to share their issues with data integration and centralization. They spoke of their own experiences and issues as practitioners in the region, but also about problems that occur in every region.

Every firm has a data management culture—like it or not, observed Hugh Stewart, sales director at SmartStream. How centralized that culture is can depend on the distribution of ownership and stewardship, more so than simply whether a firm’s policies are centralized. Yet, centralization should not necessarily mean having a central group governing data management, but rather bringing together those handling the same data task throughout a firm, said Srini Venkataramani, global head of reference data strategy at Credit Suisse.

“It’s about making sure there’s a standard and consistent way of handling data, and as far as possible, we do one task once across the firm, as opposed to repeating it,” Venkataramani said. “I completely get that. But if firms are able to centralize data management, it’s a great step toward the next level of data maturity, which is the data management culture and governance. You have a single group of people going through the pain and therefore have no option but to come up with a strategy to fix the problem.”

So centralizing data tasks through management seems preferable to centralizing a repository for data itself, as Darren Measures, head of risk at JPMorgan, observed. “As much as you try to go to a centralized data strategy, it’s never worked because there’s always something that sits off to the side,” he said. “You have to agree to what is called new business, and make sure every product you roll out is consistent with that. On data definitions, templates and standards, everyone is accountable. You make it part of your governance and project management.”

Empowering a Czar
When data is not well managed or organized throughout a firm, what can happen is what Venkataramani experienced at Credit Suisse: A vendor’s analysis told the firm that it was re-leasing or re-buying the same pieces of information an average of 24 times because of problems with its infrastructure.

Lack of an integrated infrastructure was forcing the firm to re-acquire the same pieces of data instead of re-purposing it to different departments, projects and tasks, all of which needed it. Venkataramani said that having a data czar is the best means for a firm to ensure it is properly re-purposing data it acquires from service providers. Having tried steering committees, consumer groups, centers of excellence—a gamut of management techniques—he asserts that a czar is the best way to avoid getting mired in discussions about pain points. Designating a leader for data adds accountability, he argued. That position has to be at a senior level—assigning this role to the CFO of the investment banking division, as Credit Suisse has done, puts the czar title where it should reside, he said.

However, that feeling is not universal. JPMorgan’s Measures said that a data czar designation can become inflexible, and no one knows what can happen in the event of mergers, the introduction of new products, or other issues that affect data management.

These considerations and recommendations suggest that the path to take for managing data is subjective. Placing power and responsibility in the hands of one data czar can be effective if it fits with a firm’s corporate culture. Choosing whether to organize data based on function or ownership and stewardship requires assessment of what can be most effective—again, dependent on the corporate culture.