Classification Analysis to Predict a Closing Price Condition of the Stock of UNVR using the Gaussian Copula Model

Copula introduced first by Abe Sklar at 1959 in Sklar's Theorem that represented bond or tie. Copula Modeling mostly used in multivariate analysis as alternative if multivariate normal distribution assumption not fulfilled. In this research, will do classification analysis based on Gaussian Copula Model to classify or to predict increase or decrease the stock price of PT Unilever Indonesia (UNVR) based on increase or decrease of LQ-45 Index (LQ45), Consumer Good Index (KMSI), Jakarta Composite Index (IHSG), and stock price of PT Wismilak Inti Makmur at Indonesian Stock Exchange during 111 days of exchange (January 5, 2016-June 14, 2016). Prediction process do with simulation data that built using couple information in research data that one of them is distribution for each variable. As for the results achieved are the value of the biggest increase and decrease stock price UNVR successively is 0.0529 and-0.0473. The identified distribution with estimate parameter for each variable is UNVR ∼ Bernoulli (0.46), LQ45 ∼ Normal (0.000370, 0.009267), KMSI ∼ Logistic (0.000841, 0.006491), IHSG ∼ Normal (0.000370, 0.009267), and WIIM ∼ Cauchy (−0.000735, 0.008259). While the result of prediction has prediction accuracy rate and error prediction rate successively is 68.47% and 31.53%.