Trading System Platform

Mechanical trading systems are increasingly gaining popularity in the financial sector. They originate from the chart analysis techniques and are in use since the beginning of the last century to predict price developments at stock markets. Based on market experience, individual traders or firms have their own rule systems for profitability and they use mechanical trading systems for the prediction of market trends and to increase profits. The traditional trading system development has not changed much ever since. Optimization of parameters for some fixed criterion/expected results by performing parameterized calculations on the available data is the traditional way of using mechanical trading systems for profitability.

However, over recent decades, different developments have been evolving in the domain of Artificial Intelligence. It involves:

• Fuzzy Logic as a method of "Diffuse Conclusion"
• Neural Networks to emulate neurological structures
• Neuro-Fuzzy as the combination of Fuzzy Logic and Neural Networks
• Genetic Algorithms to simulate evolutionary selective processes, etc.

At First Futures, we use the Artificial Intelligence (AI) based modern approach included in our Trading System Platform. This approach provides much flexibility for modeling, optimizations and analysis of signals and strategies. The modeling based on AI techniques does not reduce complexities and often results into huge search spaces. Developers need to understand about processes they perform in the enormously huge search spaces. Our Trading System Platform aims to provide optimization mechanism for such huge search spaces.

Many trading software products - if ever they support some Artificial Intelligence techniques - mostly focus on just a single technique, while the full power could be exploited by using them in combination with each other. The success of our Trading System Platform is in the simultaneous application of several hybrid AI techniques.

Key functionalities

• Use of Artificial Intelligence
• System designer, a highly extensible and flexible tool to design trading signals using AI, just as a human developer imagines them in regards to the process and data flow
• A high-performance back-testing engine
• Modeling and elaboration of optimization hierarchies
• Modeling and elaboration of optimization paths
• Dynamic recombination of signals and other strategy components like filters and position sizers within the paths and hierarchies
• Optimized selection of training, test and control data series
• Flexible optimization targets and target combinations for each step
• Robustness and distribution analysis with 3D graphs, and also with Monte-Carlo Simulations and data modulation
• Result set persistency for knowledge bases and reusability
• Automated execution through our FF.FIX Engine and Adapters for different Exchanges and brokers

Future versions of our Trading System Platform also include:

• Correlation analysis for the optimization of input selection
• Data mining on knowledgebases

System designer

Click the image below to see some screenshots.