StreamBase Systems is rolling out a major new release of its real-time processing platform to help financial services firms analyze data on the fly for automated trading or real-time compliance applications. At the SIA Conference, the company is introducing a capability dubbed "stream record/playback" that lets a user capture streaming data from the past and play it back.
"With the benefit of hindsight, someone can try out a trading strategy on yesterday's data or last week's data," says Bill Hobbib, vice president at Lexington, Mass.-based StreamBase Systems (booth 2019). Similar to using a VCR to record a television program and fast forwarding through the first hour, the StreamBase platform can capture the real-time data and then play it back and test it, according to Hobbib. Someone can play back the last six months of data or focus on what happened at noontime, he says. The capability is aimed at developers of automated or algorithmic trading applications as well as analysts in the front office who are building or implementing trading strategies.
Additionally, StreamBase is unveiling pre-built analytical capabilities, including moving average and volume weighted average price (VWAP), and methods of finding the best bid and best offer, as well as higher-level analytics, such as the Black Scholes model for options pricing.
For traders or developers writing their own black-box algorithms "with their own secret sauce," they can make their own custom C++ or Java code part of the StreamBase application, says Hobbib. The advantage is that people who have spent months developing analytics to run on batch data - but like the concept of running on real-time data at 50,000 to 100,000 messages per second - do not have to rebuild all their Java code. "Now, you can run it on 10,000 MPS without having to redo your work," explains Hobbib.
StreamBase also is releasing what it calls "real-time output/display and visualization." Within its graphical development environment, someone can create a real-time spreadsheet to track a portfolio of stocks. It's also making it easier for customers to generate an interface automatically to track which stocks are the most active, or send an alert to their desktop when a condition is reached on the 50 stocks in their portfolio. This is possible through Excel or a custom Java interface, which shortens development and deployment time, notes Hobbib.
Next, StreamBase is introducing the concept of "super-boxes extendability," in which users can aggregate a library of black-box algorithms or analytics and capture them as a macro and reuse them. "You can write this once and deploy it to other asset classes, locations and geographies," says Hobbib, who adds, "It's a way to roll out algorithms and a way to reuse what you built."
Finally, Streambase is offering starter applications or templates that will take people more than half of the way into building a working application, such as processing real-time data from a firehose, but it also can be used for automated trading and real-time compliance. "The concept of processing real-time data on the fly is still getting broadly adopted. Even some of the bigger firms ask, 'What can you get me out of the box?'" says Hobbib.
Ivy is Editor-at-Large for Advanced Trading and Wall Street & Technology. Ivy is responsible for writing in-depth feature articles, daily blogs and news articles with a focus on automated trading in the capital markets. As an industry expert, Ivy has reported on a myriad ... View Full Bio