MODELLING JAKARTA COMPOSITE INDEKS USING SPLINE TRUNCATED

Abstract

Regression analysis can be done by parametric and nonparametric approach. The nonparametric approach does not assume an assumption compared to parametric. One nonparametric approach is the spline truncated. Spline is a polynomial piece that provides high flexibility. Spline modeling requires spline and knots. To determine the knots using General Cross Validation (GCV). In this study modeled the value of Jakarta Composite  Index (JCI). JCI provides benefits to know the overall stock price in the stock exchange Indonesia. In this study the best spline model is linear with three knots with R square is 94.34%.
Keywords: Jakarta Composite’s Index, Spline truncated, GCV.