Rizal, Muhammad Rizal (2025) Perbandingan Model Mixed-Data Sampling Quantile Regression (MIDAS-QR) dan Mixed Frequency Quantile Vector Autoregression (MF-QVAR) untuk Nowcasting Produk Domestik Bruto Sub-Sektor Industri Pengolahan Nonmigas. Masters thesis, Institut Teknologi Sepuluh Nopember (ITS).
![[thumbnail of 6003231015-Master_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
6003231015-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy |
Abstract
Sektor industri pengolahan nonmigas merupakan bagian penting dalam perekonomian suatu negara. Penilaian dan prediksi terhadap pertumbuhan Produk Domestik Bruto (PDB) sub sektor ini secara tepat waktu (nowcasting) menjadi hal yang sangat krusial untuk mengantisipasi fluktuasi ekonomi dan menyusun kebijakan makroekonomi yang tepat. Pertumbuhan PDB dipengaruhi banyak faktor yang sebagian diantaranya terwakili oleh kondisi finansial yang disebut financial Growth at Risk (GaR). Penelitian ini penerapkan Mixed Data Sampling Quantile Autoregression (MIDAS-QR) dan Mixed Frequency Quantile Vector Autoregression (MF-QVAR) untuk nowcasting PDB sub-sektor industri pengolahan non-migas. Model ini dirancang untuk mengintegrasikan data runtun waktu yang berbeda frekuensi dan memprediksi distribusi kuantil pertumbuhan PDB menggunakan regresi kuantil, yang memungkinkan penilaian risiko dan pertumbuhan lebih akurat dibandingkan dengan metode tradisional. Penelitian ini menggabungkan informasi yang berasal dari dua bulanan indikator keuangan Financial Stress Index (FSI) dan Financial Condition Index (FCI) yang diperoleh dari analisis komponen utama beberapa indikator ekonomi. Parameter yang tidak diketahui pada model MF-QVAR diestimasi menggunakan Quasi Maximum Likelihood (QMLE) dengan bantuan algoritma Broyden-Fletcher-Goldfarb-Sanno (BFGS). Penggunaan algoritma ini memungkinkan optimasi parameter yang lebih efisien dan akurat. Hasil penelitian ini menunjukkan bahwa pada kuantil 50% model MIDAS-QR memiliki akurasi yang lebih baik dibandingkan dengan model MF-QVAR. Namun, MF-QVAR mampu menangkap lonjakan dan penurunan tajam yang terlewatkan oleh model MIDAS-QR, yang lebih terfokus pada data dengan fluktuasi stabil. Model ini diharapkan memberikan wawasan yang lebih dalam tentang dinamika sub-sektor dan membantu pengambil kebijakan serta pelaku industri dalam pengambilan keputusan strategis di tengah ketidakpastian ekonomi global yang terus berubah.
===================================================================================================================================
The non-oil and gas processing industry sector is an important part of a country's economy. Assessment and prediction of the growth of Gross Domestic Product (GDP) of this sub-sector in a timely manner (nowcasting) is very crucial to anticipate economic fluctuations and formulate appropriate macroeconomic policies. Economic growth (GDP) is influenced by many factors, some of which are represented by financial conditions called financial Growth at Risk (GaR). This research proposes a model for applying Mixed Data Sampling Quantile Autoregression (MIDAS-QR) and Mixed Frequency Quantile Vector Autoregression (MF-QVAR) for nowcasting GDP in the non-oil and gas processing industry sub-sector. The model is designed to integrate data of different frequencies and predict the quantile distribution of GDP growth using quantile regression, which enables more accurate risk and growth assessment compared to traditional methods. This study incorporates information derived from the bi-monthly financial indicators Financial Stress Index (FSI) and Financial Condition Index (FCI) obtained from principal component analysis of several economic indicators. The unknown parameters in the MF-QVAR model are estimated using Quasi Maximum Likelihood (QMLE) with the help of the Broyden-Fletcher-Goldfarb-Sanno (BFGS) algorithm. The use of this algorithm allows for more efficient and accurate parameter optimization. The results show that at the 50% quantile, the MIDAS-QR model has better accuracy than the MF-QVAR model. However, MF-QVAR is able to capture sharp spikes and dips that are missed by the MIDAS-QR model, which is more focused on data with stable fluctuations. The model is expected to provide deeper insights into the dynamics of the sub-sector and assist policy makers and industry players in making strategic decisions amidst the ever-changing global economic uncertainty.
| Item Type: | Thesis (Masters) |
|---|---|
| Uncontrolled Keywords: | Growth at Risk, MIDAS-QR, MF-QVAR, Indikator Kondisi Keuangan, Pertumbunan PDB Quarter-To-Quarter Growth at Risk, MIDAS-QR, MF-QVAR Financial Condition Indicator, GDP Growth Quarter-To-Quarter |
| Subjects: | Q Science 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) > Statistics > 49101-(S2) Master Thesis |
| Depositing User: | Muhammad Rizal |
| Date Deposited: | 23 Jul 2025 08:56 |
| Last Modified: | 23 Jul 2025 08:56 |
| URI: | http://repository.its.ac.id/id/eprint/120937 |
Actions (login required)
 |
View Item |