TIME SERIES MODELING OF STOCK RETURN USING EXPONENTIAL GARCH (EGARCH) MODEL
The EGARCH model is one of the asymmetric GARCH models. This model is capable of capturing residuals smaller than zero (bad news) and larger than zero (good news) in volatility. The EGARCH model has asymmetric coefficients to address the leverage effect by modeling leverage with heteroscedasticit...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/74487 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | The EGARCH model is one of the asymmetric GARCH models. This model is
capable of capturing residuals smaller than zero (bad news) and larger than
zero (good news) in volatility. The EGARCH model has asymmetric coefficients
to address the leverage effect by modeling leverage with heteroscedasticity and
asymmetric effects. The modeling is based on daily stock price data of PT. Astra
International Tbk. (ASII) from January 4, 2021, to October 31, 2022. ASII stock
traded on the stock exchange has fluctuating values and non-constant volatility
of stock returns (heteroskedastic). One risk measurement that can be used to
predict stock investment risk is Value-at-Risk (VaR). The research results indicate
that the best model is ARIMA(1,0,0)-EGARCH(4,5) with an AIC of -5.1082, BIC
of -4.9524, and MSE of 0.0003546861. Therefore, the predicted results using the
ARIMA(1,0,0)-EGARCH(4,5) model for the period of November 1, 2022, November
2, 2022, November 3, 2022, November 4, 2022, and November 7, 2022 are -
0.0019294, 0.0007587, 0.0003357, 0.0004022, and 0.0003918 respectively, with
VaR values of 0.0323619, 0.03334215, 0.03086995, 0.0329255, and 0.03222855 at
a 95% confidence level. |
|---|