Pemodelan Saham Sektor Energi Menggunakan Non-Homogeneous Markov Switching Autoregressive Models (NHMS-AR)

Fandisyah, Adam Fahmi (2026) Pemodelan Saham Sektor Energi Menggunakan Non-Homogeneous Markov Switching Autoregressive Models (NHMS-AR). Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6003202011_Master_Thesis.pdf Download (5MB)

Abstract

namun diiringi dengan tingkat risiko dan ketidakpastian yang cukup besar. Dalam beberapa tahun terakhir, saham-saham pada sektor energi menunjukkan tren penguatan. Dinamika tersebut tidak terlepas dari pengaruh faktor internal dan eksternal perusahaan. Oleh karena itu, diperlukan suatu model yang mampu menangkap perubahan struktur dan dinamika return saham secara lebih adaptif. Penelitian ini bertujuan untuk menganalisis dinamika return saham sektor energi di Indonesia, khususnya PT Perusahaan Gas Negara Tbk (PGAS), PT Alamtri Resources Indonesia Tbk (ADRO), dan PT Medco Energi Internasional Tbk (MEDC), menggunakan pendekatan Non-Homogeneous Markov Switching AutoRegressive (NHMS-AR). Model NHMS-AR merupakan pengembangan dari model MS-AR. Model ini memungkinkan probabilitas transisi rezim dan/atau persamaan emisi dipengaruhi oleh variabel eksogen. Dalam penelitian ini, dipertimbangkan beberapa spesifikasi model, yaitu NHTMS-AR (Non-Homogeneous Transition MS-AR), NHEMS-AR (Non-Homogeneous Emission MS-AR), dan NHTEMS-AR (Non-Homogeneous Transition and Emission MS-AR). Variabel eksogen yang digunakan meliputi Climate Risk Index (CRI), Geopolitical Risk Index (GPR), dan Global Economic Policy Uncertainty Index (GEPU), yang merepresentasikan faktor ketidakpastian global dan makroekonomi yang relevan terhadap sektor energi. Hasil analisis menunjukkan bahwa model NHMS-AR mampu menangkap perubahan rezim return saham sektor energi dengan lebih baik dibandingkan model HMS-AR. Evaluasi kinerja dilakukan menggunakan Symmetric Mean Absolute Percentage Error (SMAPE), dan hasilnya menunjukkan bahwa model NHTEMS-AR menghasilkan nilai SMAPE paling kecil dibandingkan dengan model HMS-AR, NHTMS-AR, dan NHEMS-AR. Dengan demikian, penggunaan variabel eksogen pada transisi dan emisi secara simultan memberikan kinerja model yang lebih baik dalam menggambarkan dinamika return saham sektor energi.
============================================================================================================================
Stock investment is a financial instrument with high potential returns, but it also comes with a significant level of risk and uncertainty. In recent years, stocks in the energy sector have shown a strengthening trend. This dynamic is inextricably linked to the influence of both internal and external factors within the company. Therefore, a model is needed that can capture changes in the structure and dynamics of stock returns more adaptively. This study aims to analyze the dynamics of stock returns in the energy sector in Indonesia, specifically PT Perusahaan Gas Negara Tbk (PGAS), PT Alamtri Resources Indonesia Tbk (ADRO), and PT Medco Energi Internasional Tbk (MEDC), using the Non-Homogeneous Markov Switching AutoRegressive (NHMS-AR) approach. The NHMS-AR model is an extension of the MS-AR model. This model allows the probability of regime transitions and/or emission equations to be influenced by exogenous variables. In this study, several model specifications were considered, namely NHTMS-AR (Non-Homogeneous Transition MS-AR), NHEMS-AR (Non-Homogeneous Emission MS-AR), and NHTEMS-AR (Non-Homogeneous Transition and Emission MS-AR). The exogenous variables used include the Climate Risk Index (CRI), the Geopolitical Risk Index (GPR), and the Global Economic Policy Uncertainty Index (GEPU), which represent global and macroeconomic uncertainty factors relevant to the energy sector. The analysis results indicate that the NHMS-AR model is more effective in capturing changes in the energy sector's stock return regime compared to the HMS-AR model. Performance evaluation was conducted using the Symmetric Mean Absolute Percentage Error (SMAPE), and the results indicated that the NHTEMS-AR model yielded the smallest SMAPE value compared to the HMS-AR, NHTMS-AR, and NHEMS-AR models. Thus, the simultaneous use of exogenous variables in the transition and emission mechanisms provides better model performance in describing the dynamics of energy sector stock returns.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	AutoRegressive, Energi, Markov Switching, Non-homogeneous, Time series,Energy
Subjects:	Q Science > QA Mathematics > QA274.7 Markov processes--Mathematical models. Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49101-(S2) Master Thesis
Depositing User:	Adam Fahmi Fandisyah
Date Deposited:	06 Mar 2026 01:16
Last Modified:	06 Mar 2026 01:16
URI:	http://repository.its.ac.id/id/eprint/132692

Actions (login required)

View Item