Peramalan Volatilitas Return Emas Dunia dengan Model ARIMA-GARCH

Febriko, Muhammad Iqbal (2025) Peramalan Volatilitas Return Emas Dunia dengan Model ARIMA-GARCH. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003211074-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Perkembangan teknologi informasi telah memberikan perubahan besar pada aktivitas kehidupan masyarakat, termasuk dalam aktivitas ekonomi yang mendorong minat terhadap berbagai instrumen investasi yang ada di Indonesia, emas digital menjadi salah satu instrumen populer karena kemudahan transaksi, dengan XAUUSD merepresentasikan harga emas per troy ounce dalam dollar AS, dan fluktuasi harga yang dipengaruhi oleh kondisi ekonomi global, kebijakan moneter, inflasi, serta nilai tukar dollar AS menyebabkan volatilitas tinggi sehingga peramalan volatilitas return emas sangat diperlukan untuk mengantisipasi risiko dan memaksimalkan peluang investasi. Penelitian ini menerapkan model ARIMA-GARCH di mana ARIMA memodelkan komponen mean return emas, lalu model GARCH dipakai pada residual model ARIMA untuk mengatasi heteroskedastisitas pada residual ARIMA serta memodelkan volatilitas. pada karakteristik data harga mingguan emas dunia periode 2000–2023 menunjukkan tren jangka panjang kuat (harga terendah ≈ 257,95 USD pada awal 2000-an hingga puncak 2.070,9 USD pada September 2011, lalu terdapat koreksi lalu mencapai rekor di atas 2.000 USD pada paruh kedua 2023, dengan mean harga 1.085,91 USD dan median 1.211,3 USD sedangkan, return mingguan bervariasi dengan mean 0,18 % dan median 0,26 % serta rentang -8,64 % hingga +14,22 %, pada beberapa model dugaan terpilih AR([16,49])–GARCH(1,1) sebagai yang terbaik (semua parameter φ₁₆, φ₄₉, ω, α₁, β₁, dan derajat kebebasan distribusi t signifikan p < 0,05, lulus uji diagnostik Ljung–Box, ARCH LM, dan asimetri), serta pada uji kebaikan model memiliki nilai AIC 4.3838 serta nilai QLIKE 2.435704, peramalan 26 minggu ke depan memperkirakan return mingguan akan berosilasi di sekitar nol persen, dengan nilai minimum sebesar -3,43% pada minggu ke-21 dan maksimum sebesar 8,05% pada minggu ke-2. Sementara itu, volatilitas kondisional (σₜ) menunjukkan tren meningkat secara gradual dari ≈ 2,26% pada minggu pertama menjadi ≈ 2,32% pada minggu ke-26, sehingga prakiraan ini diharapkan dapat membantu investor dan manajer risiko dalam merumuskan strategi investasi yang lebih adaptif terhadap dinamika pasar emas.
==========================================================================================================================================
The advancement of information technology has significantly transformed societal activities, including economic activities, which in turn has increased interest in various investment instruments available in Indonesia. Among them, digital gold has become one of the most popular instruments due to its transactional convenience. The XAUUSD pair represents the price of gold per troy ounce in US dollars, and its price fluctuations are influenced by global economic conditions, monetary policy, inflation, and the US dollar exchange rate. These factors lead to high volatility, making the forecasting of gold return volatility essential for managing risks and optimizing investment opportunities. This study applies the ARIMA-GARCH model, where ARIMA models the mean component of gold returns, and the GARCH model is applied to the ARIMA residuals to address heteroskedasticity and model volatility. The characteristics of weekly world gold prices from 2000 to 2023 show a strong long-term trend (with the lowest price ≈ 257.95 USD in the early 2000s and a peak of 2,070.9 USD in September 2011, followed by a correction and then reaching a new record above 2,000 USD in the second half of 2023). The mean price during this period was 1,085.91 USD, with a median of 1,211.3 USD, while weekly returns varied, having a mean of 0.18% and a median of 0.26%, ranging from -8.64% to +14.22%. Among several estimated models, the AR([16,49])–GARCH(1,1) specification was selected as the best, with all parameters (φ₁₆, φ₄₉, ω, α₁, β₁, and the degrees of freedom of the t-distribution) being statistically significant at p < 0.05, and passing diagnostic tests including the Ljung–Box test, ARCH LM test, and asymmetry test. The model also showed good performance in model fit metrics, with an AIC value of 4.3838 and a QLIKE value of 2.435704. The 26-week ahead forecast projects weekly returns oscillating around zero percent, with a minimum of -3.43% in week 21 and a maximum of 8.05% in week 2. Meanwhile, the conditional volatility (σₜ) is projected to increase gradually from approximately 2.26% in the first week to about 2.32% in the 26th week. These forecasts are expected to assist investors and risk managers in formulating more adaptive investment strategies in response to the dynamic nature of the gold market.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	ARIMA, GARCH, INFLASI, VOLATILITAS , XAUUSD, FLUKTUASI ARIMA, GARCH, INFLATION, VOLATILITY , XAUUSD, FLUCTUATIONS
Subjects:	H Social Sciences > H Social Sciences (General) H Social Sciences > HB Economic Theory H Social Sciences > HB Economic Theory > Economic forecasting--Mathematical models.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Muhammad Iqbal Febriko
Date Deposited:	01 Aug 2025 06:47
Last Modified:	01 Aug 2025 06:47
URI:	http://repository.its.ac.id/id/eprint/125801

Actions (login required)

View Item