Front-Office Demand for Tick Databases Grow

Demand for tick databases as part of front-office strategies is rising within the industry, according to recent surveys conducted by database and complex event processing (CEP) platform provider Sybase.

The surveys, conducted during April's TradeTech Europe 2010 in London and June's SIFMA Financial Services Technology Expo in New York, report that 60 percent of the more than 100 respondents indicate that they saw demands for tick databases, according to Sybase officials. The survey results also find that increased competitiveness and speed are changing trading operations, putting pressure on algorithmic models and blurring distinctions between the front and middle offices.

To address demand, "there are a lot of new network companies entering the space and a lot of data is going to in-memory databases and data-caching products from a number of companies—including IBM, Oracle, GigaSpaces and some smaller ones," says Sinan Baskan, senior director of business development for financial services at Sybase. "The space is getting crowded with quite a bit of components—that's where a lot of investment is going to go."

According to Baskan, data fragmentation will figure into the integration of front and middle offices. "Data backbones are merging," he says. "Everyone wants to put them on the same data flow and management layer, but the applications they will use continue to be different and the tools they will use will also be different. There is a lot of focus on exposing front-office data to the middle office as quickly as possible. The shortcut to that is connecting middle-office applications to the data sources in the front office. But that still doesn't solve the problem of data being fragmented across different units."

Firms have "a lot of intent" but have not yet put a lot of investment into front- and middle-office integration or tick databases, says Baskan. "Building a shared data management platform for the front and middle office is the most expensive step people can take, in acquiring new technology and developer time," he says. "The intent is there and sincere. It's one of those things that will evolve a little slower."

Survey respondents also say keeping trading models current is a major challenge, with 89 percent calling the feeding of models with multi-frequency data either important (65 percent) or very important (24 percent).