System Trading

1. Introduction to Technical Trading Systems

Computers and systems have added valuable structure to both individual and institutional trading. Indeed, multiple decision-making programs and expert sys- tems would not be possible without current technology. It has contributed to risk control and dynamic asset allocation. It also allows theories to be tested without real losses.


2. System testing at Cognitrend

The primary objectives of testing are to establish the statistical significance of the results achieved by the systematic following of pre-defined trading rules, and to evaluate the trading behaviour of the system.

2.1 The In-Sample-Test

2.2 Out-of-Sample-Testing

2.2.1 The Walk-Forward-Test

2.2.2 Walk-Across-Testing

2.3 The Stress-Test


3. Real-time trading

The trading system is now elaborated. It has survived the walk-forward and walk- across tests and is now ready for real-time trading. Fitted system had passed through the net. For as long as the behaviour remains typical, the system should not be abandoned. The greatest enemy of a good trading system is impatience.


4. Genetic Algorithms

Genetic optimisation is a particularly novel departure from the traditional methods of model building. In traditional mechanical models, a set of rules or algorithms is first defined, each modifiable according to the value of a certain variable. This variable can then be optimised in order to identify a zone of values that render the rule more efficient in a particular market. In genetic optimisation the rules themselves are optimised.