Amelia, Noor (2015) Pemodelan Volatilitas Menggunakan Metode Constant Conditional Correlation Multivariate Garch Pada Pasar Modal Indonesia. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
1213201033-Master_Thesis.pdf - Accepted Version Download (4MB) | Preview |
Abstract
Volatilitas adalah ukuran variansi suatu harga saham maupun indeks harga
saham yang bergerak dalam suatu periode tertentu. Model ARCH dan GARCH
banyak digunakan untuk mendeskripsikan bentuk volatilitas suatu data time series
yang heteroskedastisitas. Dalam perkembangannya, model multivariat GARCH
(MGARCH) merupakan perluasan dari model univariate GARCH untuk
memodelkan gerakan searah (comovement) dari serangkaian waktu. Conditional
Correlation multivariate GARCH (CCC-MGARCH) merupakan salah satu model
MGARCH yang mengasumsikan matriks korelasi konstan dan mampu mereduksi
parameter sehingga estimasi model menjadi lebih mudah. Tujuan dari tesis ini
adalah untuk mendeteksi adanya gerakan searah dari volatilitas indeks harga
saham dengan model CCC-MGARCH. Pada penelitian ini digunakan data return
indeks harga saham pada pasar modal Indonesia yaitu IHSG dan JII. Langkah
awal penelitian adalah dengan memodelkan return IHSG dan JII dalam bentuk
model univariat GARCH. Model tersebut akan menjadi dasar untuk
pembentukkan model CCC-MGARCH. Kemudian dilakukan estimasi parameter
model MGARCH dengan menggunakan metode two-step estimation. Selanjutnya
model didiagnosa dengan AIC dan SIC mengecek kecukupan model. Model CCCMGARCH
yang telah valid digunakan untuk peramalan varians dan menghitung
akurasi model dengan RMSE. Hasil penelitian menunjukan bahwa terdapat
pergerakan bersama antara IHSG dan JII dengan nilai korelasi sebesar 0,891031
dengan menggunakan CCC-MGARCH (1,1).
=================================================================================================
Volatility is a measure of the variance of a stock price and the stock price
index which moves in a given period. ARCH and GARCH models are widely
used to describe the shape of the volatility of heteroscedasticity time series data.
During its development, the multivariate model GARCH is an extension of
univariate GARCH models to model the direction of movement (comovement) of
the time series. Conditional Correlation Multivariate GARCH (CCC-MGARCH)
is one of MGARCH models that assume constant correlation matrix and can
reduce estimation parameters so that the model becomes easier. The aim of this
thesis is to detect the direction of movement of the volatility of the stock price
index CCC-MGARCH model. In this research used the data return stock price
index in Indonesia capital market, that are JCI and JII. The initial step of this
research is build model return of JCI and JII in the form of univariate GARCH
models. The model will be the basis for the formation of the CCC model-
MGARCH. Then MGARCH model parameters were estimated using a two-step
method of estimation. Furthermore, the model was diagnosed with AIC and SIC
checking the adequacy of the model. CCC-MGARCH models that have been valid
used for forecasting variance and calculate the accuracy of the model with RMSE.
The results showed that there is a movement joint between JCI and JII with a
correlation value of 0,89103 by using CCC-MGARCH (1.1).
| Item Type: | Thesis (Masters) |
|---|---|
| Additional Information: | RTMa 519.535 Ame p |
| Uncontrolled Keywords: | Volatilitas, Return Indeks Harga Saham,CCC-MGARCH, IHSG, JII. |
| Subjects: | H Social Sciences > HG Finance > HG4012 Mathematical models |
| Divisions: | Faculty of Mathematics, Computation, and Data Science > Mathematics > 44101-(S2) Master Thesis |
| Depositing User: | Yeni Anita Gonti |
| Date Deposited: | 10 Feb 2020 08:10 |
| Last Modified: | 10 Feb 2020 08:10 |
| URI: | http://repository.its.ac.id/id/eprint/74792 |
Actions (login required)
 |
View Item |