The Technology Round-Up

In what ways have data management system requirements changed in recent years?
The scale of content we now have to manage is growing quite rapidly, partly due to the number of markets we operate in. And the more markets we operate in, the more feeds we take in and have to manage across multiple time zones and overnight global closes. For example, Japan closes in the morning [for Europe] while America is still asleep.

In addition, trading isn’t localized to one country. Data now has to be distributed around the globe, systematically, reliably and quickly, and when it comes to client information, privacy and regulatory rules vary from country to country.

I also think the complexity of the information we’re dealing with is increasing. We’re now seeing at the core of trading the emergence of large lumps of data that are hierarchical in their structure, and this creates a need for vendors to handle XML and relational data. This complexity of analysis and reporting has created a need to use bi-temporal databases and encouraged the growth of stand-alone data providers.

What more can firms do to optimize existing reference data systems?

We are seeing a growing push towards trading non-vanilla instruments and the data underlying these instruments is hierarchical, and nested. Firms now need the ability to store XML-based information and have bi-temporal functionality.

There is also increasing focus on instrumenting systems and reference data valuation. Obviously reference data is very valuable, but there’s other data—meta data—that’s valuable too. For example, the value of a market price degrades in seconds, so it is not only about having clean data, but also about disseminating the time reference for when the data was created. Firms need to be able to verify to trading systems that when they provide a piece of reference data, it is accurate and timely. Instrumentation data adds this key meta data to an output and can be an important differentiator.

How can firms justify investments in improving speed and capabilities of data management systems?
Reference data is as fundamental as electricity, it needs to be thought about and managed well for systems and traders to do their jobs effectively.

I think the first thing to realize is that reference data is not just important to the back office, but to most parts of an organization. You need to have a single, unified view of the data across front office, middle office and back office. In the event of a corporate action, firms need to look at what needs to change across all the data to ensure traders and trading systems can do what they need to.