Overview
Neural networks are naturally leading the way with the perspective offering predictive indicators and new trading strategies.Trade forecasts and improved future executions are a result of insight gleaned from past performance and complex scenario, and based on machine learning.
Our product experts at Horizon Software are continuously improving strategic features of our trading platform so that our clients can easily script and run their own algorithms and launch streamlined back-testing strategies. Beyond transparency the trend is to leverage financial reporting requirements and information overload to reach a model for more intelligibility and enhanced decision-making.
This is why Horizon Software undertook major R&D work to design and implement an innovative, deep neural network architecture within the Horizon trading platform to predict the future evolution of the VWAP in very short time frames via the analysis of the first 10 limits in the order book and the last events on the underlying.
The aim of the project is to provide the platform with compressed and pre-trained neural networks to meet the load and speed requirements of high frequency trading calculations and to update predictions with real-time market data.
Key features
Top benefits
- API integration to multiple external data sources
- Open and flexible neural network to integrate your own AI strategies
- Predict trends and the evolution of VWAP and other trading strategies over different lengths of time
- Liquidity analysis, Yield Curve analysis
- High minus low predictions
- Analyse and forecast markets with greater accuracy
- Trade at the optimal price
- Efficiently mitigate risks for higher returns
- Best-in-class models with the highest levels of accuracy, performance and stability
- Attain previously impossible levels of accuracy