Prediksi Harga Saham Bank Himbara Menggunakan Model LSTM Multivariat dengan BI-Rate sebagai Variabel Eksternal

Firmansyah, Wildan Hafiz (2026) Prediksi Harga Saham Bank Himbara Menggunakan Model LSTM Multivariat dengan BI-Rate sebagai Variabel Eksternal. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) cover.pdf Download (1MB)
	Text (Bab 1) BAB 1.pdf Download (208kB)
	Text (Bab 2) BAB 2.pdf Restricted to Repository staff only until 4 September 2028. Download (375kB)
	Text (Bab 3) BAB 3.pdf Restricted to Repository staff only until 4 September 2028. Download (286kB)
	Text (Bab 4) BAB 4.pdf Restricted to Repository staff only until 4 September 2028. Download (2MB)
	Text (Bab 5) BAB 5.pdf Download (142kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf Download (155kB)
	Text (Lampiran) Lampiran.pdf Restricted to Repository staff only until 4 September 2028. Download (230kB)

Abstract

HIMBARA bank stock prices (BBNI, BBRI, BBTN, and BMRI) are influenced by internal company conditions and external macroeconomic factors, including Bank Indonesia's benchmark interest rate (BI-Rate). This study aims to measure the accuracy of a multivariate Long Short-Term Memory (LSTM) model in predicting daily (t+1) stock prices of BBNI, BBRI, BBTN, and BMRI using closing prices (Close) and BI-Rate as an external variable. Secondary data for January 2016–December 2025 were obtained from Yahoo Finance and Statistics Indonesia (BPS), processed through preprocessing, Min-Max normalization, and sequence construction using a 30-day sliding window, then tested across eight hyperparameter scenarios. Model performance was evaluated using MAE, RMSE, and MAPE, and compared with LSTM without BI-Rate and with XGBoost. Results show MAPE values of 1.46% for BBNI, 1.70% for BBRI, 2.21% for BBTN, and 1.50% for BMRI, averaging 1.72%. The choice of BI-Rate has a theoretical basis, as the benchmark interest rate affects the cost of funds and credit distribution, yet empirically BI-Rate only improved accuracy for BBNI, while for BBRI, BBTN, and BMRI the model without BI-Rate performed better. Overall, the multivariate LSTM model can predict HIMBARA bank stock prices with relatively low error, while the contribution of BI-Rate to accuracy varies across stocks.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Hadiwiyanti, Rizka	NIDN0727078602	rizkahadiwiyanti.si@upnjatim.ac.id
Thesis advisor	Pahlawan, Muhammad Reza	NIDN0716059801	muhammad_reza.si@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Information Systems

Depositing User:

Wildan Hafiz Firmansyah

Date Deposited:

07 Sep 2026 02:12

Last Modified:

07 Sep 2026 02:12

URI:

https://repository.upnjatim.ac.id/id/eprint/59977