Like Tiger Woods changing his swing after winning the Masters, we set out to fix what wasn’t broken when we decided to migrate to a .Net platform from Java. The decision was made because we saw an opportunity to advance beyond our capabilities with our present system, much like Tiger saw a way to advance his game beyond the limitations imposed by his present mechanics. And we’re happy to report that the results have been about as good as Tiger’s.
Liquidnet is an electronic marketplace that we believe is the most efficient venue for institutional investors (such as pension and mutual fund managers) to trade equity securities worldwide. These institutional investors are using our secure interface to enhance the quality and speed of trade execution, gain price improvement for their trades, and, ultimately, lower overall trading costs. With a responsibility to continuously be a technology leader in the buy side trading and block liquidity, questions arise without a competitive prompt. We asked ourselves what our limits were while using Java to build and manage our user interface, simply because we realised that in order to stay ahead of our competition in the future, java would not be able to support our needs.
In 2004, we began the process of investigating alternatives to Java. We started with the question: “What would we do differently if we were starting Liquidnet today?” While our original interface for our US platform served us well as the company grew to a global trading platform, Java was not the strategic platform to enable us to take our Trading System to the next level. The global trading platform provides our customers access to our global natural liquidity pool of more than 7.5 billion shares per day on average (May 31, 2008) listed in 29 countries throughout North America, Europe, Africa, Asia and Australia.
Therefore, we identified our overarching goal – create a user experience that would be unmatched in the industry. That led us to Microsoft’s Windows Presentation Foundation (WPF). Microsoft was betting that WPF, it’s future user interface framework based on .NET, would ultimately change software development to enable creating rich, multimedia applications. Since this goal was aligned with Liquidnet’s business goal, our decision to migrate to .NET was sealed as a necessary stepping stone to get to that visionary system.
This transition from java to .net was achieved seamlessly because of our automated full-scale regression tests, utilised in past technology upgrades which accelerated our time to market. It was because of this that the migration was possible to run concurrent with normal operations from August 2006 to August 2007. Despite the rising trade volumes during the period and normal trader work-flow, our users’ experience remained undisturbed.
