Getting Under the Surface of Standards

“If you speak to five people about data quality, you will get 10 different definitions of what it means,” suggests Llew Nagle, head of consumer service management for reference data at Deutsche Bank. This multiplicity of answers frustrated the data quality aficionado so much that, 18 months ago, he set up a working group to pin down what is meant by the term. Under the sponsorship of the EDM Council, the group agreed on a common definition of data quality that organizations can now use to make sure they are singing from the same hymn sheet.

In his current role, Nagle is confronted with the problems of organizations when they use corporate actions data to trade. One of the big talking points in the corporate actions market at the moment is standardization, with the relative merits of ISO 15022, ISO 20022 and XBRL being frequently debated. While Nagle agrees standardization helps when people are communicating with each other, he is skeptical about the idea that standards provide a “silver bullet” for all the problems in the space. He uses an unlikely zoological analogy to explain: “I always tend to imagine that if you look at a swan, it looks beautiful on the surface, but underneath it does not look so pretty. I tend to see messaging standards in a similar way. Just because we are presenting a very standardized view going outside the organization, it doesn’t necessarily mean that, internally, the data quality and standards are up to scratch.”

Nagle goes on to suggest standardization can sometimes adversely affect data quality: “A lot of these standards, if implemented poorly, or implemented under a lot of time pressure, can degrade your data quality within an organization, because you are just adding another layer of transformation.”

Having joined DataStream (now Thomson Reuters) after graduating from the University of Durham, Nagle has been working in the industry for 16 years, including executive positions at Barclays Capital from 2001 to 2010. He would like to see more descriptive standards, such as ISO country codes and currency codes, which define what is in a message, as these help with the symbology and terminology people use internally.

Although their models can “vary wildly” between systems, Nagle thinks most companies have a reasonable understanding of how to model a corporate action. He believes timeliness is now a more serious problem where corporate actions are concerned, saying: “The thing that tends to kill people at the moment is late notification of where a corporate action creates a new line [on a trading system], what that line should be, and co-ordinating the times of when these things actually happen.” He continues: “The most common issue I see at the moment is problems around associated static changes of a corporate action, so either creating a new line or name change or symbology change. They tend to be the major things I hear. So it’s more the outcome of a corporate action, rather than the corporate action itself.”

The obvious solution to problems with timeliness is straight-through processing (STP). Nagle explains: “STP is the ambition of everyone. It involves the minimum amount of additions or touching of data as it moves through an organization. So from a corporate actions perspective, it is the full automation of either creating a new line based on a corporate action or triggering a dividend payment based on the information we have coming through.”

The biggest change Nagle sees is that businesses are starting to understand the importance of data. “If you look across the market, the people with a deep understanding of reference data and how it should be used are very much in demand, because a lot of organizations have very deep and strategic programs that are actually in play at the moment to try and resolve some of the deficiencies they perceive they have. I see that very much continuing.”

He predicts the organizations investing in reference data now will soon see the benefits: “I think you are going to start seeing a lot of organizations getting a lot leaner and smarter about not just the way they deal with data but also the way they operate.”