Optimasi Portofolio Utang Valas Pemerintah Indonesia Menggunakan Metode CVaR Berdasarkan Peramalan ARIMA

Pambudi, Bagus Rico (2026) Optimasi Portofolio Utang Valas Pemerintah Indonesia Menggunakan Metode CVaR Berdasarkan Peramalan ARIMA. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002221144-Undergraduate_Thesis.pdf Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Utang valas memiliki peran penting sebagai salah satu sumber pembiayaan untuk menutup defisit APBN, tetapi memiliki risiko akibat perubahan suku bunga global dan nilai tukar mata uang. Oleh karena itu, diperlukan optimasi portofolio agar diperoleh komposisi utang valas dengan ekspektasi perubahan nilai utang dan risiko yang minimum dengan mempertimbangkan potensi perubahan suku bunga serta nilai tukar mata uang pada periode mendatang. Model ARIMA dipilih karena mampu memodelkan dan meramalkan pergerakan suku bunga serta nilai tukar mata uang, sedangkan ukuran risiko CVaR dipilih karena mampu menangkap risiko pada kondisi ekstrem. Penelitian ini bertujuan untuk menghasilkan peramalan tingkat suku bunga dan nilai tukar mata uang menggunakan model ARIMA, memformulasikan model optimasi portofolio utang valas Pemerintah Indonesia menggunakan model Mean-Variance Optimization (MVO) dan model optimasi dengan kendala risiko Conditional Value at Risk (CVaR), serta menentukan komposisi portofolio utang valas optimal dan membandingkan kinerja kedua model tersebut. Data yang digunakan merupakan data sekunder deret waktu harian periode 3 Juli 2017 sampai 24 November 2025, yang meliputi tingkat suku bunga utang, nilai tukar mata uang, dan data portofolio utang valas Pemerintah Indonesia. Data harian diubah menjadi data bulanan, lalu suku bunga dan nilai tukar diramalkan selama 12 bulan ke depan menggunakan model ARIMA dengan bantuan perangkat lunak RStudio. Hasil peramalan digabungkan dengan data historis sebagai input optimasi MVO dan CVaR dengan bantuan perangkat lunak MATLAB. Hasil penelitian menunjukkan bahwa model MVO mampu menurunkan ekspektasi persentase perubahan nilai utang valas sekitar 19% dan menekan risiko sekitar 13% dari portofolio yang digunakan Pemerintah Indonesia saat ini, sedangkan model CVaR-90% mampu menurunkan ekspektasi persentase perubahan nilai utang valas sekitar 18% dan menekan risiko CVaR sekitar 0,3%. Model ARIMA mampu menghasilkan peramalan tingkat suku bunga dan nilai tukar mata uang dengan akurasi yang baik sehingga dapat digunakan sebagai dasar optimasi portofolio utang valas, sedangkan model MVO dan CVaR-90% sama-sama mampu mengoptimalkan kinerja portofolio, dengan MVO lebih unggul dari sisi standar deviasi dan CVaR-90% lebih sesuai untuk pengendalian risiko ekstrem.
===================================================================================================================================
Foreign-currency debt plays an important role as one source of financing to cover the state budget deficit, but it carries risks arising from changes in global interest rates and exchange rates. Therefore, portfolio optimization is required to obtain a foreign-currency debt composition with minimum expected changes in debt payment costs and risk while considering possible changes in interest rates and exchange rates in the future period. The ARIMA model is used because it can model and forecast interest rates and exchange rates, while the CVaR risk measure is used because it can capture risk under extreme conditions. This study aims to forecast interest rates and exchange rates using the ARIMA model, formulate the foreign-currency debt portfolio optimization model of the Government of Indonesia using the Mean-Variance Optimization (MVO) model and the optimization model with Conditional Value at Risk (CVaR) constraints, and determine the optimal foreign-currency debt portfolio composition by comparing the performance of both models. The data used in this study are secondary daily time series data from 3 July 2017 to 24 November 2025, consisting of debt interest rates, exchange rates, and the foreign-currency debt portfolio data of the Government of Indonesia. The daily data are converted into monthly data, and the interest rates and exchange rates are then forecasted for the next 12 months using the ARIMA model with the aid of RStudio. The forecasting results are combined with historical data as the input for MVO and CVaR portfolio optimization using MATLAB. The results show that the MVO model can reduce the expected percentage change in foreign-currency debt payment costs by approximately 19% and reduce risk by approximately 13% compared with the current portfolio used by the Government of Indonesia. Meanwhile, the CVaR-90% model can reduce the expected percentage change in foreign-currency debt payment costs by approximately 18% and reduce CVaR risk by approximately 0.3%. The ARIMA model provides accurate forecasts of interest rates and exchange rates, so the forecasting results can be used as the basis for foreign-currency debt portfolio optimization. Both the MVO and CVaR-90% models can improve portfolio performance, with MVO being superior in terms of standard deviation and CVaR-90% being more suitable for controlling extreme risk.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	ARIMA, CVaR, MVO, optimasi portofolio, utang valas
Subjects:	H Social Sciences > HA Statistics > HA30.3 Time-series analysis H Social Sciences > HB Economic Theory > Economic forecasting--Mathematical models. H Social Sciences > HG Finance > HG3881 Foreign exchange. H Social Sciences > HG Finance > HG4012 Mathematical models H Social Sciences > HG Finance > HG4529.5 Portfolio management
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Bagus Rico Pambudi
Date Deposited:	30 Jul 2026 02:14
Last Modified:	30 Jul 2026 02:14
URI:	http://repository.its.ac.id/id/eprint/140291

Actions (login required)

View Item