Equity Forecasting a Case Study on the KLSE Index, Neural Networks.pdf

EQUITY FORECASTING A CASE STUDY ON THE KLSE INDEX JINGTAO YAO HEANLEE POH Department of Information Systems and Computer Science National University of Singapore Kent R idge Singapore Fax Email yao jingtiscsnussg p ohhliscsnussg ABSTRACT This pap er presents the research of neural networks as applied in equity forecast ing in an emerging market such as the Kuala Lumpur Sto ck ExchangeKLSE Backpropagation neural networks are used to capture the relationship b etween the technical indicators and the levels of the KLSE index over time The exp er iment shows that useful predictions can b e made without the use of extensive market data or knowledge In fact a signicant pap er prot can b e achieved by purchasing indexed sto cks in the resp ective prop ortions The pap er however also discusses the problems asso ciated with technical forecasting using neural networks such as the choice of time frames and the recency problems KEY WORDS Neural Network Financial Analysis Emerging Equity Market Prediction Intro duction Equity has long b een considered a high return investment eld The ma jor forecasting metho ds used in the nancial area are either technical or fundamen tal Due to the fact that sto ck markets are aected by many highly interrelated economic p olitical and even psychological factors interaction among these indi cators b ecame very complex Therefore it is generally very dicult to forecast the movement in the sto ck market Classical statistical techniques for forecasting reach their