Abstract
Stock market prediction is among the significant factors determining the decision
to invest in the stock market. The motive of earning more and getting maximum
profits from the investments in the stock market has led the investors, researchers
and financial analysts to devise the methods to predict the price change of the
commodities which are in circulation in the market, and the need for
understanding forecasting techniques became even more crucial as the level of
trading and investment in the stock market grew. The desire for optimizing gains
and minimizing risk factor can also be considered as a motive of applying and
devising the new techniques for forecasting. This study uses a totally different
method to forecast and predict stock returns in Karachi Stock Exchange. The
study has applied three years data on monthly basis of Karachi Stock Exchange100 index for forecastimg. The data range is June 2011 to July 2014 consisting
of 36 observations. A Multilayer Perceptron Neural Network technique is applied
for prediction. The study concludes that the Karachi Stock Exchange 100 index is
not efficient and that returns can be forecastedby applying other methods for
prediction and Multilayer Neural Network is among the methods. In the
circumstances, buying and holding stocks is the most appropriate strategy.
Nosheen, Amjad Amin, Danish Alam. (2016) Neural Networks and Stock Market: Efficiency Hypothesis: A Case of Pakistan, Putaj Humanities And Social Science, Volume-23, Issue-1.
-
Views
621 -
Downloads
61
