FORECASTING CONVENTIONAL AND ISLAMIC STOCKS ON SUB SECTOR OF PLANTATION USING WAVELET DAUBECHIES DISCRETE TRANSFORM METHOD.

• Main Author: WIJAYANTO (NIM : 10214088), RESTU
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/30346
• Institution: Institut Teknologi Bandung

Econophysics is one of the studies of the physics of complex systems that apply and offer ideas, methods, and models in statistical physics and complexity to analyze data from economic phenomena. One of the studies of econophysics is forecasting a time series data. Forecasting aims to predict what will happen in the future by doing the analysis of previous data. Forecasting method is often to predict the stock price so it can be used as a reference to make the right investment strategy. In general, there are two types of stocks namely conventional stock and islamic stock, the difference between the two types of stocks are islamic stock apply the principles of islamic in all transactions while the conventional stock is not. The stocks studied are conventional stocks (Gozco Plantation, Jaya Agra Wattie, Provident Agro, and Sinar Mas Agro Resource and Technology) and islamic stocks (Astra Agro Lestari, Austindo Nusantara Jaya, Eagle High Plantations and PP London Sumatra Indonesia) in the plantation sub-sector. The data studied are the closing prices of each stock in the period March 2017 to February 2018 as data training and March 2018 as test data. Used Daubechies Discrete Wavelet Transformation to perform time series data forecasting with the help of Rstudio software. Discrete wavelet transforms are used to convert data in time domain into wavelet domains. The selection of thresholding function is using soft thresholding and hard thresholding with Minimax and Adaptive parameters. From the function and thresholding parameters used, the best estimation is searched from the mean square error (MSE) <br /> <br /> value obtained for forecasting. The experiment on Gozco Plantation stock got the best estimation on the first Adaptive level parameter and obtained the data for the <br /> <br /> period of March 2, 2018 - March 23, 2018, with the value of MSE of 0.6634 which is quite close to the actual data. The results obtained by conventional stocks have a higher average increase of 5.04% compared to islamic stocks which has an increase of 3.29%.