Font Size: a A A

The Surplus Forecast Model Industry Perspective: A Statistical Model & Gray Prediction Model

Posted on:2013-02-19Degree:MasterType:Thesis
Country:ChinaCandidate:J WuFull Text:PDF
GTID:2219330374962644Subject:Accounting
Abstract/Summary:PDF Full Text Request
Earnings forecasting reveals the problem that the intrinsic value of a companydepends on its future profitability. An accurate earnings forecasting can help manyinvestors of the capital market investigate the listed company reasonably and make awise judgment. In this paper, earnings forecasting of different industries will bepredicted from two aspects, one aspect is from those financial indicators which caninfluence the earnings, and the other aspect is from the internal structure of earnings,which means we only take the earning data into account.First nonferrous metals industry, pharmaceutical industry and automotiveindustry are considered as the representative industry in this paper. Principalcomponents of the three industries can be obtained by principal components analysismethod. The data used in the principal components analysis are all the financialindicators of the fourth quarter of2009and the earnings per share of the first quarterof2010, and these data are all from Guotai Junan database. Then the regressionequations of the three industries are built by respective principal components. Fromcomparing the actual value of the second, third and fourth quarter of2010with thepredictive value of the three quarters of2010, under this statistic method, theprediction of nonferrous metals industry is most effective. By this statistic method,two common problems can be avoided, one is that financial indicators are inefficientuse, the other one is multicollinearity in multiple regression.Then by using the quarter data of earnings per share of the three industries from2006-2009, quarterly Grey prediction model are set up after removing those datawhich are not in keeping with the request of building Grey prediction model. Aftercomparing the actual value of the four quarters of2010with the predictive value ofthe Grey prediction model, the conclusion is the prediction of nonferrous metalsindustry is most effective.Through the comparing of the predictive efficiency of the two methods, in thenonferrous metals industry, the predictive result of statistic method is more effectivethan Grey prediction. But in contrast to the he nonferrous metals industry, the predictive result of statistic method is less effective than Grey prediction. So thefinancial indicators should be emphasized in earnings forecasting for nonferrousmetals industry, however, the earnings data should be paid more attention in thepharmaceutical industry and automotive industry. Earnings forecasting can be moreaccurate from the aspect of industries.
Keywords/Search Tags:earnings forecasting, principal component analysis, multiple regression, Grey prediction, MATLAB
PDF Full Text Request
Related items