Wall Street Widens Open Source Embrace
June 22, 2009
Whether sell-side, buy-side, exchanges or alternative trading venues, Wall Street is widening its embrace of open source technology. Today, budget-conscious capital-market firms are implementing open source order management, routing, event processing and other systems.
A key attraction these days is that open source software--whose source code is free and widely distributed over the Web--costs less than building from scratch or buying software from a third-party vendor. And with open source technology, the code is knowable, transparency needs are addressed, integration with existing applications is more modular--and there's the promise of access to the intellectual smarts of a worldwide developer community.
"More vendors are offering open source products and this will continue to evolve," says Kevin McPartland, senior analyst at Tabb Group in New York.
And as open source continues to evolve, it's finding a broader home within securities firms. Cerebellem Capital, a quantative hedge fund, for instance, is using Marketcetera's open automated trading system, which is based on EsperTech's open Java-based Esper complex event processing/event stream processing (CEP/ESP) engine. The Esper system is also embedded into the liquidity management system of Paris-based Smart Trade. At this week's Sifma Tech conference, Marketcetera is expected to announce that it's looking to work with a U.S.-based liquidity venue to roll out a software as a service (SaaS) trading platform that gives high frequency traders a high performance hosted infrastructure offering with the agility of an open source platform at a fraction of the cost of an on-premise proprietary trading system.
CME Group, one of the world's largest derivatives exchange operators, operates in a low-latency environment using Linux servers, which are the primary platform for most of its electronic infrastructure. CME is now migrating its Globex electronic trading engines and infrastructure to these Linux high-volume servers, while still running some proprietary operating systems. The move to Linux/x86 servers from proprietary Unix has resulted in a 20 percent to 50 percent initial performance gain, according to Vinod Kutty, associate director and head of distributed computing R&D at CME Group.
Bank of America and some of the world's largest hedge funds are doing mathematical modeling with R, an open source data analytics software program used in computational finance and statistical arbitrage trading. The R project works much like the proprietary Matlab package found on quantitative research desks.
And JPMorgan Chase made its proprietarily developed credit default swap (CDS) analytical engine, used to price CDS contracts, available to the International Swaps and Derivatives Association (ISDA). ISDA will give the engine to the industry to help increase standardization, offering it as open source code to promote transparency on CDS pricing.
Open and closed
Despite the advantage of using open source, IT executives still voice concerns around security and systems integration--and those issues are definitely taken seriously when managers consider open source, says John Avery, a partner at Sungard Consulting Services. Typically, these concerns are addressed in internal technical reviews, weeding out dirty code in approving open source software.
"A solution that can withstand the scrutiny of all comers--especially those intent on discovering vulnerabilities--generally tends to be more trouble-free in the long run than one only scrutinized by a small set of developers," says CME Group's Kutty. "More eyes are a good thing. Bugs will surface but there will be availability of fixes for widely used open source projects."
