Application of Regression and Four Weeks Moving Average Methods in NSE Secondary Market
In the highly volatile Indian Stock Market prediction of the stock price movements is a challenging task for the
investors. Investors are often in search of a suitable forecasting indicator to identify the movement in the stock price in order
to invest for good returns. In this study, 10 companies are selected from the top five highly volatile sectorial indices based on
beta computations and then two mid performing companies from each of these volatile sectors are selected based on alpha
computation from the National Stock Exchange. 53 weeks stock price data of equity shares is used for the analysis. Two
statistical techniques namely, regression analysis and 4 weeks moving average methods are applied to this data and error
terms such as Mean Absolute Deviation (MAD), Mean Squared Error (MSE) and Mean Absolute Percentage Error (MAPE)
are computed to find out the accuracy of the forecast. t-test is used to identify significant difference in error term values
between regression analysis and 4 weeks moving average methods. Correlation coefficients are computed to identify the
degree of relationship between the actual and forecasted values for both regression and 4 weeks moving average methods.
The results indicated that the 4 weeks moving average is a better forecasting method for 8 companies and for two companies
both forecasting methods are suited as their error term values and correlation coefficients are not significantly different.
Keywords - Stock Price, Regression, 4 Weeks Moving Average, MAD, MSE, MAPE.