In this paper genetic programming is used to investigate a number of long time series of price data for a stock exchange quoted share, in order to discern whether there are any patterns in the data which could be used for technical trading purposes. This extends the work done by the authors in a previous paper (Fyfe et al. 1999) which suggested that, although it was possible to find a rule which did outperform simple buy and hold, there were insufficient grounds for the rejection of the efficient market hypothesis. The purpose of the present paper is to investigate the robustness and generalisability of the conclusion reached by Fyfe et. al.
|Title of host publication||Computing in Economics and Finance 2000|
|Publisher||The Society for Computational Economics|
|Publication status||Published - 2000|