Two of the biggest issues facing trading firms today are the total cost of trading and the ability to generate a return.

An integral part of this is the sourcing of market data required to power various trading applications and execute trading strategies.

“The cost issue is intensifying as firms look to diversify their strategies away from traditional cash equities into the OTC markets or other regions, or both,” said Henk D’Hoore, head of product, data feed services, at financial data provider Interactive Data’s trading solutions arm.

However, in order to understand whether a particular strategy will generate the required return, a firm must have access to a breadth of data to fully assess that asset class or new venue.

“The two major issues are the cost and size of the data,” said Philip Enness, director of markets infrastructure at IBM, a technology firm. “The challenge will be offering a flexible infrastructure that supports a broad set of analytics while minimizing the duplication of data, which requires maximizing access to that data.”

Interactive Data’s Feed Handler creates an output format that emulates the data format used by Thomson Reuters’ Reuters Market Data System, or RMDS, which allows companies to consume market data and distribute it to applications that would otherwise only consume RMDS data.

RMDS deployments tend to be large, enterprise-wide installations and well embedded into a firm’s architecture.

“We estimate there are more than 2,500 installations globally and each of those is powering multiple downstream applications,” said D’Hoore at Interactive Data. “The Feed Handler presents data to applications using the same data models they were originally designed to utilize. This enables applications to consume global cross-asset data, including tick-by-tick level one and full order book data, in exactly the same way as their legacy feed sources.”

With the current budgetary constraints that many firms face, it makes little sense for them to remove RMDS, which is an open market data platform, as it would be a major undertaking.

“Recognizing that issue, we have circumvented the problem by representing our own data from the consolidated feed in a format that is known to these applications,” D’Hoore said. “As a result, a firm gains the rich global content available in our consolidated feed through RMDS but without the need to reprogram the business critical trading applications that feed off that RMDS installation.”

The Feed Handler is optimized both in terms of performance and business logic.

A single instance of the Feed Handler run on commodity hardware can, for example, process a 400,000-instrument watch list universe and up to 900,000 updates per second with virtually no added latency.

“As such, customers can take advantage of the less than 10 millisecond ticker plant latency that the Consolidated Feed is capable of outputting in addition to its full-tick level 1 and full order book features,” D’Hoore said.