ANALISIS INDEKS HARGA SAHAM GABUNGAN DAN FAKTOR PENGARUHNYA MENGGUNAKAN PEMODELAN REGRESI SEMIPARAMETRIK KERNEL DILENGKAPI GUI-R

Abstract

Composite Stock Price Index (IDX) shows the movement of stock prices used by investors to determine their investment strategy. IDX movement is influenced by macroeconomic factors such as money supply and inflation, so regression analysis is used to determine the relationship between the variables. Based on the scatterplot, money supply is known as a parametric predictor variable as it has a linier line patterned scatterplot and inflation is a nonparametric predictor variable as it has a random patterned scatterplot, so semiparametric regression modelling is used for the analysis. Kernel regression was chosen to analyze the nonparametric component based on the random patterned scatterplot of inflation. This study aims to obtain the results of semiparametric kernel regression modelling analysis and to create a GUI to be applied to the analysis as a development of previous similar studies that still done based on CLI. This study uses monthly data from January 2013 to December 2020 with the proportion of in sample and out sample data distribution 87,5%:12,5%. Based on the smallest MSE value as the best model criteria, semiparametric regression model with triangle kernel function is the best model obtained with optimal bandwidth=3.24, which means the model is strong and which means that the forecasting results are very accurate. GUI has been created according to the needs of the modelling analysis implementation.