Pramudito, Samuel Cahyo (2024) Peramalan Volatilitas Nilai Indeks Harga Saham Gabungan (IHSG) Menggunakan Model Integrasi Hybrid GARCH-LSTM. Other thesis, Institut Teknologi Sepuluh Nopember.
Text
5006201065-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (2MB) | Request a copy |
Abstract
Investasi merupakan penempatan aset dengan harapan nilainya akan meningkat dan menghasilkan pendapatan pasif. Investasi saham di pasar modal penting bagi perusahaan global karena dampaknya signifikan terhadap ekonomi negara. Saham memiliki risiko tinggi namun dapat memberikan return besar. Investor harus mempertimbangkan pilihan saham sesuai toleransi risiko dan indikator seperti Indeks Harga Saham Gabungan (IHSG), yang mengukur kinerja portofolio dan pergerakan harga saham di pasar modal. IHSG berbanding lurus dengan harga saham. Volatilitas IHSG dipengaruhi berbagai faktor, seperti suku bunga dan nilai pertukaran mata uang. Faktor-faktor tersebut dapat meningkatkan risiko dan ketidakpastian investasi, sehingga analisis tren IHSG perlu dilakukan sebelum investasi. Peramalan volatilitas IHSG bertujuan memprediksi pergerakan harga saham di masa mendatang. Penelitian ini menggunakan metode hybrid GARCH-LSTM untuk meramalkan volatilitas IHSG dari Januari 2018 hingga Desember 2023. Imputasi menggunakan metode linear imputation dilakukan untuk mengatasi 33 missing value pada data. Orde GARCH yang digunakan pada penelitian ini adalah GARCH(1,1). Pemodelan volatilitas pada model GARCH(1,1) menghasilkan nilai volatilitas dengan rentang 0,0000360984 sampai 0,0025139193. Model LSTM yang digunakan untuk melakukan prediksi residual adalah model dengan proporsi split data 80:20 dan menghasilkan nilai MAE sebesar 0,000015339655. Hasil prediksi volatilitas IHSG menggunakan kombinasi hybrid GARCH-LSTM pada penelitian ini menghasilkan rentang nilai 0,0000319 sampai 0,0000741. Hasil prediksi volatilitas dengan menggunakan metode hybrid GARCH-LSTM memiliki MAPE sebesar 30,38%, yang menunjukkan bahwa prediksi dengan metode ini termasuk dalam kategori reasonable prediction. Hal ini menunjukkan bahwa metode hybrid GARCH-LSTM masih dapat dipertimbangkan untuk melakukan prediksi volatilitas IHSG. Hasil peramalan nilai log return untuk 20 periode selanjutnya dengan mempertimbangkan volatilitas IHSG menunjukkan bahwa peramalan nilai log return cenderung konstan dan tidak mengalami perubahan yang signifikan.
===========================================================
Investment is the placement of assets with the hope that their value will increase and generate passive income. Stock investment in the capital market is important for global companies due to its significant impact on a country's economy. Stocks have high risk but can provide high returns. Investors must consider stock choices according to risk tolerance and indicators such as the Composite Stock Price Index (IHSG), which measures portfolio performance and stock price movements in the capital market. The IHSG is directly proportional to stock prices. The volatility of the IHSG is influenced by various factors, which could increasing the risk and uncertainty of investments, hence the need to analyze IHSG trends before investing. The aim of forecasting IHSG volatility is to predict future stock price movements. This study uses the hybrid GARCH-LSTM method to forecast IHSG volatility from January 2018 to December 2023. Imputation using the linear imputation method was performed to address 33 missing values in the data. The GARCH order used in this study is GARCH(1,1). Modeling conditional variance in the GARCH(1,1) model produced volatility values ranging from 0,0000360984 to 0,0025139193. The LSTM model used for residual prediction is a model with an 80:20 data split ratio and resulted in an MAE value of 0,000015339655. The IHSG volatility prediction results using the hybrid GARCH-LSTM combination in this study produced a value range of 0,0000319 to 0,0000741. The volatility prediction results using the hybrid GARCH-LSTM method have a MAPE of 30,38%, indicating that predictions with this method fall into the category of reasonable prediction. This shows that the hybrid GARCH-LSTM method can still be considered for predicting IHSG volatility. The forecast results for the log return values for the next 20 periods, considering IHSG volatility, show that the forecasted log return values tend to be constant and do not experience significant changes.
Item Type: | Thesis (Other) |
---|---|
Uncontrolled Keywords: | GARCH, Forecasting, LSTM, Volatility, Volatilitas, Peramalan |
Subjects: | H Social Sciences > HA Statistics > HA30.3 Time-series analysis H Social Sciences > HG Finance > HG4529 Investment analysis H Social Sciences > HG Finance > HG4915 Stocks--Prices Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. 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) > Actuaria > 94203-(S1) Undergraduate Thesis |
Depositing User: | Samuel Cahyo Pramudito |
Date Deposited: | 02 Aug 2024 06:16 |
Last Modified: | 02 Aug 2024 06:16 |
URI: | http://repository.its.ac.id/id/eprint/111157 |
Actions (login required)
View Item |