Dublin-based trading and risk managment technology vendor Prediction Dynamics has launched Crucible 1.0, a new financial time-series modeling platform at the Seventh European Conference on Hedge Fund Investments in Geneva, Switzerland.
Designed for high-end users, including hedge funds and investment bank proprietary trading desks, the system can be used to build and maintain a large number of predictive models across a variety of financial markets and asset classes. Built around two core predictive engines, the package has applications in systematic trading, active risk management and quantitative research, says Prediction Dynamics.
The Crucible's FactorEngine enables quantitative researchers to analyse and identify the optimal factor sets that drive a market. The ModelEngine enables users to create stable nonparametric predictive models based on the factors defined by FactorEngine.
John Carney, chief technology officer at Prediction Dynamics, says the system has been designed to build models that consistently generate stable and accurate predictive signals, even during sustained periods of market stress.