Analisis Dependensi Return Logam Mulia Menggunakan Pendekatan ARIMA dan Copula

Kusuma, Tasya Nabila Setyadani (2025) Analisis Dependensi Return Logam Mulia Menggunakan Pendekatan ARIMA dan Copula. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5006211076-Undergraduate_Thesis.pdf - Submitted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Investasi merupakan aktivitas penanaman modal yang bertujuan untuk memperoleh keuntungan di masa depan. Logam mulia seperti emas, perak, platinum, dan paladium merupakan instrumen investasi yang banyak diminati karena sifatnya yang relatif stabil dan sering dianggap sebagai safe haven dalam menghadapi ketidakpastian ekonomi global. Namun demikian, harga logam mulia tetap menunjukkan volatilitas yang signifikan akibat faktor eksternal seperti inflasi, nilai tukar, kebijakan moneter, dan fluktuasi permintaan industri. Oleh karena itu, pemodelan pergerakan harga dan pemahaman terhadap struktur ketergantungan antar logam menjadi penting dalam manajemen risiko portofolio. Penelitian ini bertujuan untuk menganalisis hubungan dependensi return logam mulia menggunakan pendekatan ARIMA dan Copula multivariat. Data yang digunakan berupa harga penutupan harian dari empat logam mulia selama periode 1 Januari 2024 hingga 1 Januari 2025. Model ARIMA digunakan untuk memodelkan dinamika masing-masing logam secara individual, kemudian residual dari model tersebut ditransformasikan menjadi pseudo-observations. Selanjutnya, lima jenis copula diuji untuk menggambarkan ketergantungan antar logam, yaitu: Normal, Student-t, Clayton, Frank, dan Gumbel. Estimasi parameter dilakukan dengan metode Maximum Likelihood Estimation (MLE). Hasil estimasi menunjukkan bahwa Normal Copula merupakan model terbaik dengan nilai log-likelihood sebesar 158,85, AIC sebesar –305,70, dan BIC sebesar –284,52. Model ini mengungguli Student-t, Clayton, Frank, dan Gumbel Copula. Korelasi tertinggi pada Normal Copula ditemukan pada pasangan Platinum–Paladium (ρ = 0,6300), diikuti oleh Emas–Perak (ρ = 0,5697), yang menunjukkan adanya ketergantungan positif signifikan antar logam dalam portofolio.=====================================================Investment is an activity involving capital allocation with the goal of generating future returns. Precious metals such as gold, silver, platinum, and palladium are popular investment instruments due to their relatively stable nature and frequent designation as safe haven assets amid global economic uncertainty. However, despite their perceived stability, the prices of precious metals still experience significant fluctuations influenced by external factors such as inflation, exchange rates, monetary policies, and industrial demand. Therefore, risk estimation is essential for investors to understand the dynamics of these assets and optimize their portfolio strategies. This study aims to analyze the dependence structure among precious metals using the ARIMA and multivariate Copula approaches. Daily closing prices of four precious metals from January 1, 2024, to January 1, 2025, are used as the data source. After modeling each series with ARIMA, the residuals are transformed into pseudo-observations for Copula estimation. Five types of Copula models—Normal, Student-t, Clayton, Frank, and Gumbel—are evaluated using the Maximum Likelihood Estimation (MLE) method.The results show that the Normal Copula provides the best fit, with a log-likelihood value of 158.85, AIC of –305.70, and BIC of –284.52. It outperforms the other models, including Student-t, Clayton, Frank, and Gumbel. The highest correlation is observed between platinum and palladium (ρ = 0.6300), followed by gold and silver (ρ = 0.5697), indicating strong positive dependencies among precious metals within the portfolio.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	ARIMA, Copula, Logam Mulia, Dependensi, Pseudo Observations, Precious Metals, Dependence Analysis
Subjects:	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
Depositing User:	Tasya Nabila Setyadani Kusuma
Date Deposited:	04 Aug 2025 07:54
Last Modified:	04 Aug 2025 07:54
URI:	http://repository.its.ac.id/id/eprint/126761

Actions (login required)

View Item