PREDIKSI HARGA DAN RISIKO KERUGIAN SAHAM BLUE CHIP MENGGUNAKAN GATED RECURRENT UNIT DAN VALUE AT RISK

Dana, Alvin Ryan (2025) PREDIKSI HARGA DAN RISIKO KERUGIAN SAHAM BLUE CHIP MENGGUNAKAN GATED RECURRENT UNIT DAN VALUE AT RISK. Undergraduate thesis, Universitas Pembangunan Nasional Veteran Jawa Timur.

[img] Text (Cover)
21083010035_Cover.pdf

Download (1MB)
[img] Text (Bab 1)
21083010035_Bab I.pdf

Download (100kB)
[img] Text (Bab 2)
21083010035_Bab II.pdf
Restricted to Repository staff only until 17 March 2028.

Download (566kB)
[img] Text (Bab 3)
21083010035_Bab III.pdf
Restricted to Repository staff only until 17 March 2028.

Download (363kB)
[img] Text (Bab 4)
21083010035_Bab IV.pdf
Restricted to Repository staff only until 17 March 2028.

Download (2MB)
[img] Text (Bab 5)
21083010035_Bab V.pdf

Download (15kB)
[img] Text (Daftar Pustaka)
21083010035_Daftar Pustaka.pdf

Download (126kB)
[img] Text (Lampiran)
21083010035_Lampiran.pdf
Restricted to Repository staff only until 17 March 2028.

Download (135kB)

Abstract

This study aims to predict stock prices of PT Mayora Indah Tbk (MYOR) and Tower Bersama Infrastructure (TBIG) using the Gated Recurrent Unit (GRU) model, while also assessing investment risk through the Value at Risk (VaR) approach. The issue addressed in this study is the high volatility of stock prices, which increases the risk of potential losses for investors. The GRU model is selected for its ability to predict fluctuating patterns in stock data. This study also incorporates univariate method to create a model with the target of close price. The data used includes daily data of stock prices from July 2019 to July 2024, sourced from the Yahoo Finance. An evaluation was conducted on three GRU model configurations with different parameter variations and the Adam optimizer to determine the best model. Model performance was measured using the Mean Absolute Percentage Error (MAPE). The results indicate that for MYOR stock, Model 2 with a configuration of 100 epochs and a batch size of 32 achieved the lowest MAPE value of 1.364710, making it the best-performing model. Meanwhile, for TBIG stock, Model 3 with a configuration of 100 epochs and a batch size of 32 demonstrated the best performance with a MAPE value of 1.583951. In addition to stock price prediction, this study also analyzed investment risk by calculating daily returns and Value at Risk (VaR) using historical simulation. The VaR calculation results show that the maximum potential loss for MYOR stock is RP 54.158 (5.42%), while for TBIG stock, it reaches Rp63.362 (6.34%). By considering both the prediction results and risk analysis, this study is expected to assist investors in making more optimal investment decisions, taking into account both profit opportunities and risk mitigation strategies

Item Type: Thesis (Undergraduate)
Contributors:
ContributionContributorsNIDN/NIDKEmail
Thesis advisorTrimono, TrimonoNIDN0008099501trimono.stat@upnjatim.ac.id
Thesis advisorIdhom, MohammadNIDN0010038305idhom@upnjatim.ac.id
Subjects: H Social Sciences > HA Statistics
Q Science > QA Mathematics > QA76.6 Computer Programming
Divisions: Faculty of Computer Science > Departemen of Data Science
Depositing User: Alvin Ryan Dana
Date Deposited: 17 Mar 2025 04:52
Last Modified: 17 Mar 2025 04:52
URI: https://repository.upnjatim.ac.id/id/eprint/35572

Actions (login required)

View Item View Item