Abstract
The economic development of a country can be seen based on the capital market that was growing and developing. One of the most popular capital market instruments is stocks. Stocks based on market capitalization groups include longitudinal data. One of the statistical methods for longitudinal data modelling is nonparametric regression which has no modelling assumptions requirement. This research models monthly stock prices using a nonparametric local polynomial method with the selection of the best model which has minimum value of Mean Square Error (MSE). The data was divided into 2 parts, namely in sample data from November, 2018 to June, 2021 to form a model and out sample data from July, 2021 to February, 2022 used for evaluation of model performance by Mean Absolute Percentage Error (MAPE) values. The best model is the local polynomial model with Biweight kernel function of degree 5, local point of 4, bandwidth of 37, and MSE value of 0.03481085. MAPE out sample of data value is 31.13%, which indicating that the model has sufficient forecasting. In this research arrange Graphical User Interface (GUI) by using R software with shiny package is built to make display output data analyzing more easy and more interactive.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
