Please use this identifier to cite or link to this item: https://repository.uksw.edu//handle/123456789/32702
Title: Kinerja Pencocokan Model Realized GJR-CJ pada Data Aset Keuangan
Other Titles: Performance of Realized GJR-CJ Model Matching on Financial Asset Data
Authors: Wulandari, Nadya Putri
Keywords: lompatan;Realized Volatility;volatilitas;Realized Glosten-Jagannathan-Runkle;jump
Issue Date: 1-Mar-2024
Abstract: Volatilitas merupakan indikator utama dalam menilai risiko ketika membuat keputusan investasi. Di dunia pasar keuangan volatilitas mencerminkan tingkat fluktuasi nilai aset keuangan selama periode tertentu. Cara paling umum untuk mengukur potensi kerugian di masa depan dari suatu investasi adalah melalui volatilitas. Studi ini mempelajari model volatilitas Realized Glosten-Jagannathan-Runkle (RGJR) dengan memperluasnya menjadi RGJR-CJ dengan cara mendekomposisi variabel eksogen menjadi variabel-variabel kontinu dan lompatan. Sebagai kerangka sederhana, model-model yang dipelajari diasumsikan berdistribusi Normal. Sebagai ilustrasi empiris, model diaplikasikan pada data Tokyo Stock Price Index (TOPIX ) Jepang yang mencakup dari Januari 2004 hingga Desember 2011. Variabel eksogen yang diamati yaitu Realized Volatility (RV) yang dihitung dalam intra-harian dengan interval waktu 1 menit, 5 menit, dan 10 menit. Metode Adaptive Random Walk Metropolis (ARWM) dikerjakan dalam algoritma Markov Chain Monte Carlo (MCMC) untuk mengestimasi model. Berdasarkan empat kriteria pemilihan model, hasil empiris menunjukkan bahwa model RGJR-CJ memberikan pencocokan data yang lebih baik daripada model RGJR, yang mana pengadopsian data RV 1 menit menyediakan pencocokan terbaik.
Volatility is a key indicator in assessing risk when making investment decisions. In the world of financial markets, volatility reflects the degree to which the value of a financial asset fluctuates over a given period. The most common way to measure the future loss potential of an investment is through volatility. This paper studies the Realized Glosten-Jagannathan-Runkle (RGJR) volatility model by extending it to RGJR-CJ by decomposing the exogenous variables into continuous and jump variables. As a simple framework, the models studied are assumed to be Normal distributed. As an empirical illustration, the models are applied to an index in the Japanese stock market, namely Tokyo Stock Price Index (TOPIX) data covering from January 2004 to December 2011. The observed exogenous variable is Realized Volatility (RV) calculated intra-day with time intervals of 1 minute, 5 minutes, and 10 minutes. Adaptive Random Walk Metropolis (ARWM) method was employed in Markov Chain Monte Carlo (MCMC) algorithm to estimate the model. Based on four model selection criteria, the empirical results showed that the RGJR-CJ model provides a better data fit than the RGJR model, of which the adoption of 1-minute RV data provides the best fit.
URI: https://repository.uksw.edu//handle/123456789/32702
Appears in Collections:T1 - Mathematics

Files in This Item:
File Description SizeFormat 
T1_662020011_Judul.pdf3.51 MBAdobe PDFView/Open
T1_662020011_Isi.pdf
  Until 9999-01-01
1.17 MBAdobe PDFView/Open
T1_662020011_Daftar Pustaka.pdf463.26 kBAdobe PDFView/Open
T1_662020011_Lampiran.pdf
  Until 9999-01-01
628.35 kBAdobe PDFView/Open
T1_662020011_Formulir Pernyataan Persetujuan Penyerahan Lisensi Nonekslusif Tugas Akhir dan Pilihan Embargo.pdf434 kBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.