TY - JOUR
TI - PEMODELAN RETURN INDEKS HARGA SAHAM GABUNGAN MENGGUNAKAN THRESHOLD GENERALIZED AUTOREGRESSIVE CONDITIONAL HETEROSCEDASTICITY (TGARCH)
AU - Saida, Maidiah Dwi Naruri; Sudarno, Sudarno; Hoyyi, Abdul
IS - Vol 5, No 3 (2016): Wisuda periode Agustus 2016
PB - Departemen Statistika FSM Undip
JO - Jurnal Gaussian
PY - 2016
SP - 465
EP - 474
UR - http://ejournal-s1.undip.ac.id/index.php/gaussian/article/view/14702
AB - ARIMA model is one of modeling method that can be applied on time series data. It assumes that the variance of residual is constant. Time series data, particularly the return of composite stock price index, tend to change rapidly from time to time and also fluctuating, which cause heteroscedasticity where the variance of residual is not constant. Autoregressive Conditional Heteroscedasticity (ARCH) or Generalized Autoregressive Conditional Heteroscedasticity (GARCH) can be used to construct model of financial data with heteroscedasticity. Besides of having inconsistent variance, financial data usually shows phenomenon where the difference of the effect between positive error value and negative error value towards data volatility, called asymmetric effect. Therefore, one of the GARCH asymmetric models, Threshold Generalized Autoregressive Conditional Heteroscedasticity (TGARCH) is used in this research to solve heteroscedasticity and asymmetric effect in stock price index return. The data in this research is stock price index return from January 2nd, 2013 until October 30th, 2015. From the analysis, TGARCH models are obtained. ARIMA([3],0,[26])-TGARCH(1,1) is the best model because it has the smallest AIC value compared to other models. It produces the forecast value of stock price index return nearly the same with actual return value on the same day.Â Keywords: Return, Heteroscedasticity, Asimmetry effect, ARCH/GARCH, TGARCH.