(an `Intelligent’ decision support tool for investors)

The Trading Rule Generator is a decision support tool that processes a large fan of financial data, and generates simple linguistic statements, explaining hidden relationships between seemingly unrelated stocks, bonds and exchange rates.

The idea behind such a system is that certain time series may be linked to a number of unsuspecting other time series. If this hypothesis is true, then such relationships can be extracted, and used in market timing and portfolio management.

Let us assume that the Trading Rule Generator is asked to find rules to explain when certain "stock X" would go up. Input data typically would cover a number of time series. For example:

Using the information provided, the system automatically categorises each field into their fuzzy groupings of DOWN, NORMAL, UP, then it puts forward and tests thousands of hypotheses to explain - when stock X would go UP using the past data from all available data fields. Depending on the size of the dataset the system spends some time on generating, testing and analysing rules. In the end it reports a list of high scoring rules in terms of their validity in the given dataset. A typical rule would state:

More data will improve the quality and the validity of rules but with the impact of slowing down the computational process. The system is not limited to the kind of inputs presented above, and it can handle symbolic or numeric information.

Trading Rule Generator generates a number of most significant rules accompanied by a numerical analysis of associated with a confidence value. The system is not an automatic portfolio manager, it is a decision support tool mining historic data and discovering and reporting relationships which may be used by traders and portfolio managers.

The system is a state of the art simulator designed as a decision support tool, ideal for medium to large size investors and fund managers who deal with a large number of securities. The system (Figure 1) contains the following modules:

• Intelligent Data Miner
• Graphical User Interface
• Time Series Database

The intelligent engine is a hybrid of Fuzzy Logic and Genetic Algorithms. It generates and tests thousands rules, and reports significant relationships in the form of fuzzy logic linguistic statements.

GUI uses any Java-enabled internet browser, allowing the selection of data fields, starting and monitoring the data mining process.

A fast database and hardware platform which holds and updates large amounts of financial data. For a strategic trading system a set of ASCII files are sufficient. But for an online trading system the Intelligent Data Miner needs to be linked to internet data feeds for a prompt reaction. 

Trading Rule Generator is hardware platform independent, and runs on a PC or a Unix workstation. Its Java version runs on any Internet browser regardless of the operating system. The software needs to be integrated with current databases/data feeds, and potentially it can run as a background process continuously generating and reporting rules.

The main strength of the system over classic decision support tools is that it generates linguistic statements combining a large number of financial time series. Significant relationships can then be used in trading and portfolio management to reduce risk.

As the system uses a computationally intensive algorithm it requires high performance computers. This is important particularly if intra-day trading rules are required.

As it is, the system is designed as an interactive decision support tool requiring the manual selection of inputs and the assessment of results by marketing analysts. We are currently working for fully automating the system so that it runs continuously as a background process, and reports only the significant rules in appropriate times.

We envisage the following possible options:

This option allows both parties to work in their area of expertise. The client provides input and target datasets, we generate relationships, and report to the client at agreed intervals.

The system can be fully or partly licensed and run by the client at the client’s premises. This option requires a familiarisation and training period on the use of the software. There are three possibilities in licensing the system:

This option covers the delivery of the following modules:

a - Intelligent Data Miner

b - Graphical User Interface

c - Integration with databases

This option includes the Intelligent Data Miner and Graphical User Interface as described above without integration with databases, the system will work as demonstrated from a distributed network operating on ASCII files across the internet/intranet.

The code for the Intelligent Data Miner is available in Java, C or C++. It reads ASCII flat files and writes the outputs into similar files. It needs to be customised to the specific requirements of the client. If the client wishes, help is provided in order to link the Intelligent Engine to current databases.