Picture four old college buddies, deciding to ditch their regular jobs and start their own business in a rambling old house. By the fourth year, the business begins raking in a million pounds a year. Surely this happens only in movies, right? Wrong - the Hollywood scenario actually happened to four Imperial College, London, graduates who quit their jobs in 1995 to form Xenomorph, a front office data management and analysis system based in Wimbledon.

Before setting up Xenomorph, all four worked in the financial services industry: one was a quant at Bankers Trust, another was a techie at Logica, the third was a quant in the equity derivatives group at JP Morgan and the fourth traded equity derivatives at ING Hong Kong. They agreed that one of the most intractable difficulties facing derivatives trading desks was the long turnaround time for integrating new pricing model analytics with historic and real time data in risk management and data analysis systems. Their new system was designed to address this problem.

Marketing director Brian Sentance describes Xenomorph as a suite of products that allows "storage, integration, distribution and analysis of trade data". However, it is not a pricing tool, and Xenomorph doesn't supply data. Instead, the product suite is designed to make it easy for risk managers and derivatives traders to link together a variety of supported market data feeds with in-house and third-party option analytics. This allows rapid derivation and storage of key trading data such as historic implied volatilities and historic zero curves.

One of the key components in the product suite is the Xenomorph Data Explorer, which is designed to allow easy viewing and manipulation of data sources and databases, without sacrifices to database power or flexibility. A risk manager or trader can use Data Explorer to visually design complex data structures to support new derivative instruments without the need for programming or SQL experience.

Data Explorer makes use of the proprietary native Xenomorph Database (XDB) format. XDBs can store a comprehensive amount of historical and nonhistorical data, such as prices, yields, correlation matrices, basket compositions, volatility surfaces, yield curve definitions and regions of Excel spreadsheets. XDBs are designed for the highest possible performance in storing and retrieving time series data - data that Xenomorph classifies as historic storage of any shape of financial data, not simply numbers for prices and yields.

The maximum size of a single native XDB is 95 gigabytes, and there is no limit to the number of instruments that can be stored in one database. This disk storage is used extremely efficiently, needing only a 1GB XDB to store 300,000 years of daily price data.

Xenomorph argues that although the bigger shops are more likely to use large relational databases such as Oracle or Sybase, for time series data (date/value pairs), databases such as these are not always the best storage medium. Access times to time series data can increase tremendously as new instruments and their histories are added to the database. Xenomorph tries to get around this problem by offering high-speed access at all times, a task it was specifically designed to do.

During the first 18 months of its existence, the company was based in an old house in South London and funded by the four directors themselves. Now it employs 20 people and anticipates that 1999 turnover will be in the region of 1 million pounds.

Xenomorph's first client was Bankers Trust, and it now has 11 clients and 20-plus sites. Newer clients include General Re Financial Products, ING and Rabobank. Sentance admits, however, that potential clients are more likely to be smaller hedge funds and boutiques such as Tokai Asia and Wharton Investment Management in Hong Kong that do not have large IT departments. "A lot of traders at smaller shops like those often do the IT themselves. We can give them a head start on the competition" says Sentance.

In the next few months, Xenomorph will add interfaces that provide historic yield curve analysis using models from Tech Hackers, FinCAD and other vendors, and deriving implied volatility surfaces from Liffe's option data. Another key module is a trade-capture and risk management system that integrates the best aspects of instrument data storage in Xenomorph Databases with the storage of transaction and position data in the Microsoft SQL Server.

The Xenomorph system can also be used to access directly market data feeds such as Bloomberg, Datastream, FactSet and Reuters. Others are in development. It has interfaces to Excel, Visual Basic, C/C++ and Microsoft COM. The system runs on the Microsoft Windows NT operating system, versions 3.51 and 4.0, and will support all future versions of Windows NT.