Krom Üretim Miktarı ve Fiyatının Panel Veri Kullanılarak Modellenmesi

Abstract

Within the scope of the thesis, by providing the information about the chrome mine, the factors affecting the quarry sales price, quarry sales amount and the right of government have been examined thanks to the data obtained the General Directorate of Mining and Petroleum Affairs. Panel Data Analysis is performed with Stata-12 program. First, it is investigated whether there is a cross-section independence between variables. After finding the cross-section dependence between variables, the Pesaran Unit Root Test, one of the second generation unit root tests, is performed. The effects of the unit and time are also investigated. The models, having the effects of unit and time, are investigated by applying Hausman Test and it is examined whether the fixed effect model or the random effect model should be used. According to the Hausman test statistic, it is decided to use the fixed effect model and then the assumptions are tested. Heteroscedasticity problem is tested by the Wald test and the problem of autocorrelation is tested by the Bhargava-Franzini-Narendranathan Durbin-Watson Test and Baltagi-Wu Test. The parameters of the models are obtained by using the Arellano-Froot-Rogers Estimator as there are heteroscedasticity and autocorrelation problems in the model. In addition, multi-collinearity problem is also investigated and it is found that there is no multi-colinearity in the models despite series are related. Finally, according to the results of the panel data analysis, performed to the data between 2005 and 2015, it is determined that the impacts of average dollar exchange rate on quarry sales price, the prices of the right of government on the sales volume are significant and negative; besides, the stock amount has a remarkable and negative impact on the quarry sales price; whereas, the production quantity and the quarry sales price have a remarkable and positive effect on the quarry sales quantity.

Keywords: Panel Data Analysis, Arellano-Froot-Rogers Estimator, Unit Root Tests, Pesaran Cross Section Independence Test, F Test, Maximum Likelihood Ratio (LR) Test, Hausman Test, Wald Test, Bhargava-Franzini-Narendranathan Modified Durbin-Watson Test, Baltagi-Wu Test, Stata, Chrome