CQG, the financial markets technology solutions provider, has announced the successful completion of testing of its artificial intelligence (AI) predictive model for traders, boasting an impressive 80% accuracy rate in forecasting the future movements of the E-mini S&P 500 futures contract.
The newly developed machine learning (ML) toolkit is designed to provide retail traders and institutional clients, such as proprietary trading firms and hedge funds, with cutting-edge tools for identifying trading opportunities, guiding trading strategies, and managing positions effectively.
CQG’s started developing the model in early 2023, which leverages CQG’s repository of historical trade data and analytics, and addresses several real-world challenges, including the management of large volumes of data, integrating CQG’s Python-based ML infrastructure with the financial industry’s C++ frameworks, and refining the ML training pipeline for time series prediction.
“We intentionally steered clear of the widespread natural language processing technologies, impressive as they are,” Ryan Moroney, CQG’s CEO, tells TradingTech Insight. “When we set up our AI lab last year, our focus was to capitalise on our expertise in handling time series and market data, along with the analysis of specific trading patterns. Our next tick predictor, which forecasts market movements with 80% accuracy, serves as a tangible validation of our efforts.”
The model was rigorously tested it in a multi-platform laboratory environment before its capabilities were validated in a live trading scenario last week, where it mirrored the 80% predictive success rate it had previously achieved in back-testing environments.
“One practical application of our technology is to enhance the effectiveness of our algorithms, particularly for those who frequently trade in derivatives contracts and are looking to minimise slippage,” says Kevin Darby, Vice President of Execution Technologies at CQG. “By integrating this technology, our algorithms become significantly more efficient in accumulating futures contracts for users.”
“We help people use time series data to make trading decisions,” says Moroney. “We don’t tell people what those decisions should be. We don’t tell them what they should do next. We don’t give advice. We’re a technology company that gives users tools to make better decisions. Predicting the next tick is one application of that. Ultimately, our goal is to place our robust infrastructure – encompassing advanced math engines, AI models, and comprehensive market data – into our customers’ hands, enabling them to innovate and develop their own solutions.”
The company is now exploring additional applications of its AI toolkit in collaboration with key partners.
Subscribe to our newsletter
