Goal: Trading System Enhancement

Regency Stocks and Commodities Fund, LP (www.regencyfund.com) is a mixed fund. It trades stocks, options and futures, both discretionarily and mechanically. Regency’s customers are accredited investors and qualified eligible participants (QEP). Mike Barna (michaelb@regencyfund.com) is Regency’s Vice-President in charge of Research and Trading. His responsibilities include managing Regency’s mechanical trading systems used in various financial markets. As part of his job he evaluates different software tools to improve and enhance his work. In this endeavor, he chose WizWhy.

On an on-going basis, Regency takes data from large databases and pre-processes it based on domain expertise. This results in a larger database. Various modeling and database engines are then used to perform predictability studies on various elements in the database. In essence, Regency takes theoretical models and financial company data and applies them to the actual trading data to make a product. Regency’s product is a rule-based trading system used in actual trading operations for the benefit of their investors, to manage a certain allocation of the fund.

Because Regency is always challenged to take their data and produce a better product, they looked for a data-mining tool, which would produce better modeling and rules. For this reason they chose WizWhy.

The deterministic nature of WizWhy’s resulting rules was the main reason WizWhy was chosen for Regency’s day-to-day work. During the initial use of WizWhy, Regency discovered a deeper or hidden knowledge, which was difficult to find in other data mining products. They also found it easy to manipulate and import data. In addition, Regency liked the ease at which WizWhy’s generated rules could be exported into other trading systems’ platforms for future analysis.

When Regency first used WizWhy, they passed a flat file through WizWhy and noticed that some features and rules were more significant. They knew this to be very important. But a big issue for Regency was to know if these rules were profitable for their trading system. They were delighted to find that even the simpler rules were very good based on their measurement standards.

When using WizWhy, the hypothetical results of Regency’s model increased by 30%!