Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/153677| Title: | Perbandingan Kinerja Model Geometric Brownian Motion dan ARIMA dalam Memprediksi Harga Saham BIRD |
| Other Titles: | Comparing the Accuracy of Geometric Brownian Motion and ARIMA Models in Predicting the BIRD Stock Prices |
| Authors: | Mangku, I Wayan Budiarti, Retno YULIANTY, SHERLY |
| Issue Date: | 2024 |
| Publisher: | IPB University |
| Abstract: | Saham merupakan salah satu instrumen investasi yang dapat dipilih untuk
memperoleh keuntungan di masa yang akan datang. Harga saham bergerak
fluktuatif mengikuti proses stokastik. Model matematika diperlukan dalam
membaca pergerakan saham, misalnya model Geometric Brownian Motion (GBM)
dan Autoregressive Integrated Moving Average (ARIMA). Penelitian ini bertujuan
untuk membandingkan tingkat keakuratan model GBM dan ARIMA dalam
memprediksi harga saham. Data yang digunakan adalah harga penutupan saham PT
Blue Bird Tbk periode 11 Februari 2023 – 11 Februari 2024. Data tersebut dibagi
menjadi data training sebanyak 191 data dan data testing sebanyak 46 data. Di
antara model ARIMA, model ARIMA(4,1,4) adalah model terbaik dengan nilai
AIC sebesar 2008.16 dan semua parameternya signifikan pada taraf nyata 5%.
MAPE dipilih menjadi alat dalam mengukur tingkat keakuratan model dalam
kinerjanya memprediksi harga saham karena data yang digunakan tidak memiliki
pencilan. Berdasarkan hasil penelitian diperoleh nilai MAPE model GBM sebesar
10.3634%, sedangkan model ARIMA(4,1,4) sebesar 7.7838%. Stocks are one of the investment instrumens that can be chosen to obtain profits in the future. Stock prices fluctuate following a stochastic process. Mathematical models, such as Geometric Brownian Motion (GBM) and Autoregressive Integrated Moving Average (ARIMA) models, can be used in reading stock movements. This study aims to compare the accuracy of GBM and ARIMA models in predicting stock prices. The data used are the closing price of PT Blue Bird’s Tbk shares for the period February 11, 2023 – Februari 11, 2024. The data are divided into 191 training data and 46 testing data. Among the ARIMA models, the ARIMA(4,1,4) was the best model with an AIC value of 2008.16 and all its parameters were significant at level 5%. MAPE was chosen as a tool in measuring accuracy of the model in predicting stock prices because the data used do not have outliers. Based on the results of the study, the MAPE value of GBM model was 10.36345%, while the MAPE value of ARIMA(4,1,4) model was 7.783817%. |
| URI: | http://repository.ipb.ac.id/handle/123456789/153677 |
| Appears in Collections: | UT - Mathematics |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| cover_G5401201017_961ca407cea240d49903d78affa7d22e.pdf | Cover | 2.44 MB | Adobe PDF | View/Open |
| fulltext_G5401201017_644481594e0d41769eda4df0e05a8df9.pdf Restricted Access | Fulltext | 2.86 MB | Adobe PDF | View/Open |
| lampiran_G5401201017_68089fe3acf74c728f6ab61ffdf2e6c7.pdf Restricted Access | Lampiran | 2.31 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.