HSBC, Citi Contend With Standards Variations

Data managers that are attempting to achieve consistency when aggregating data from different sources are having to contend with multiple standards in their efforts.

Data standards improve the quality of data, says Cornelius Crowley, associate director of information standards in the Office of Financial Research (OFR) at the US Treasury. However, being able to reliably compare data requires analysis and aggregation of data, he adds. "The decisions you make really depend on how well you can compare that data."

Semantic data has emerged as a way to better manage databases that goes beyond just aligning relational databases with the use of the XML format standard established in 1998. Layering meanings onto sets of data can make it easier to build databases, according to Michael Lang, chief executive of Revelytix, a provider of data integration software based in Sparks, Md.

"The big challenge is getting a core model that everyone buys into-getting multiple stakeholders to buy into whatever contract or settlement they need," he says. "If you get a core model onto which you can hang everything, then any problems after that will be pretty small."

Groups with different functions within a firm, such as traders and financial managers, tend to want to put data to different uses, Lang observes. "So it becomes very hard for them all to agree on the right way to describe that data," he says. "Internally, the need to do analysis for management is one reason to have standards. Another reason is the need to do transactions with hundreds of other firms."