Optimalisasi Pembentukan Portofolio Saham Berbasis Mean Shift Clustering untuk Prediksi Value at Risk

Putri, Shafira Amanda (2026) Optimalisasi Pembentukan Portofolio Saham Berbasis Mean Shift Clustering untuk Prediksi Value at Risk. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (COVER) 22083010008_COVER.pdf Download (1MB)
	Text (BAB I) 22083010008_BAB 1.pdf Download (193kB)
	Text (BAB II) 22083010008_BAB 2.pdf Restricted to Repository staff only until 6 March 2029. Download (617kB)
	Text (BAB III) 22083010008_BAB 3.pdf Restricted to Repository staff only until 6 March 2029. Download (404kB)
	Text (BAB IV) 22083010008_BAB 4.pdf Restricted to Repository staff only until 6 March 2029. Download (2MB)
	Text (BAB V) 22083010008_BAB 5.pdf Download (174kB)
	Text (DAFTAR PUSTAKA) 22083010008_DAFTAR PUSTAKA.pdf Download (174kB)
	Text (LAMPIRAN) 22083010008_LAMPIRAN.pdf Restricted to Repository staff only until 6 March 2029. Download (194kB)

Abstract

The Indonesian stock market is volatile, causing investors uncertainty in selecting stocks that can generate optimal returns with manageable risk. Such fluctuations can significantly change investment value in a short period, making portfolio construction require not only potential returns but also diversification, proper allocation, and quantitative risk assessment. An approach that integrates stock selection, portfolio optimization, and risk prediction is therefore needed. This study constructs an optimal portfolio using Mean Shift clustering on LQ45 stocks from January 2023 to December 2025. Variables used are expected return, variance, and average trading volume, standardized to ensure uniform measurement. Clustering produced three stock clusters, and selection was based on the highest Sharpe Ratio in each cluster, resulting in EXCL, BMRI, ANTM, and BRPT. Correlation analysis showed low return correlations, supporting diversification. The portfolio was optimized using Mean–Variance to determine allocation weights and evaluated for risk via Value at Risk (VaR) using Historical Simulation and Cornish–Fisher Expansion for different holding periods and confidence levels. Results showed weights of EXCL 43%, ANTM 24%, BRPT 33%, and BMRI 0%, with a Sharpe Ratio of 0,87. At a 95% confidence level and 1-day holding period, VaR values were 2,77% (Historical Simulation) and 2,85% (Cornish–Fisher). These results indicate the approach can produce a portfolio with relatively high returns and measurable risk.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Trimono, Trimono	NIDN0008099501	trimono.stat@student.upnjatim.ac.id
Thesis advisor	Muhaimin, Amri	NIDN2119950723270	amri.muhaimin.stat@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA76.6 Computer Programming

Divisions:

Faculty of Computer Science > Departemen of Data Science

Depositing User:

Shafira Amanda Putri

Date Deposited:

09 Mar 2026 07:43

Last Modified:

09 Mar 2026 07:43

URI:

https://repository.upnjatim.ac.id/id/eprint/50213