UTILIZING ARTIFICIAL NEURAL NETWORK FOR PREDICTION IN THE NIGERIAN STOCK MARKET PRICE INDEX

This paper utilize ANNs model to predict closing price of AshakaCem Security in Nigeria Stock Market price index. In this paper, AshakaCem Security historical technical data were collected for four years trading period (2005 – 2008), the data was partition into training, cross validation and testing set in the ratio 70%:10%:20% respectively. Architectural configuration of Feed Forward Artificial Neural Networks (FFANN) with parameters of three (3) layers, four (4) input nodes, one (1) hidden layer, eighteen (18) hidden nodes, one (1) output layer and one (1) output node was obtained and FFANN was build with the calibrations, trained with 268 exemplars, cross validated with 37 exemplars and tested on 76 exemplars and evaluated on four (4) performance indicators which include Mean Square Error (MSE), Correlation Coefficient (r), Normalize Mean Square Error (NMSE) and Mean Absolute Error (MAE). The technical data were subjected to sensitivity analysis. FFANN predictor was developed and modeled historical trading data of AshakaCem Security and pattern was captured in the historical data with a trend accuracy of 80%, Efficient Market Hypothesis was contradicted in this paper and sensitivity analysis shows that previous closing price (pclose) is the most significant input on FFANN predictor output. FFANN predictor build was evaluated and result shows that MSE = 3.59739909, NMSE = 0.039391504, MAE = 1.3407 and r = 0.981331937. It is possible to trained ANNs with controlled parameters using technical data of AshakaCem Security to capture pattern in the historical data and generalize well on unseen data.

[1]  Ramon Lawrence,et al.  Using Neural Networks to Forecast Stock Market Prices , 2000 .

[2]  Tak-Lam Wong,et al.  Design and implementation of NN5 for Hong Kong stock price forecasting , 2007, Eng. Appl. Artif. Intell..

[3]  Piotr Stalinski,et al.  The Determinants of Foreign Listing Decision: Neural Networks Versus Traditional Approaches , 2006 .

[4]  John Yearwood,et al.  Predicting Australian Stock Market Index Using Neural Networks Exploiting Dynamical Swings and Intermarket Influences , 2003, J. Res. Pract. Inf. Technol..

[5]  Goutam Dutta,et al.  Artificial Neural Network Models for Forecasting Stock Price Index in the Bombay Stock Exchange , 2006 .

[6]  E. Fama.  EFFICIENT CAPITAL MARKETS: A REVIEW OF THEORY AND EMPIRICAL WORK* , 1970 .

[7]  Gideon Kanji Damaryam.  Vision systems for a mobile robot based on line detection using the Hough Transform and artificial neural networks , 2008 .

[8]  Jason E. Kutsurelis.  Forecasting financial markets using neural networks: an analysis of methods and accuracy , 1998 .

[9]  Tong-Seng Quah,et al.  Improving returns on stock investment through neural network selection , 1999 .

[10]  Chi-Chung Lam,et al.  FINANCIAL TIME SERIES FORECASTING BY NEURAL NETWORK USING CONJUGATE GRADIENT LEARNING ALGORITHM AND MULTIPLE LINEAR REGRESSION WEIGHT INITIALIZATION , 2000 .

[11]  J. Moody.  Economic forecasting : challenges and neural network solutions , 1995 .

[12]  Dennis Olson,et al.  Neural network forecasts of Canadian stock returns using accounting ratios , 2003 .

[13]  Myungsook Klassen,et al.  Investigation of Some Technical Indexes in Stock Forecasting Using Neural Networks , 2007, WEC.

[14]  Ajith Abraham,et al.  Real stock trading using soft computing models , 2005, International Conference on Information Technology: Coding and Computing (ITCC'05) - Volume II.