Perbandingan Berbagai Model Conditionally Heteroscedastic Time Series Dalam Analisis Risiko Investasi Saham Syariah Dengan Metode Value At Risk

Abstract

Value at Risk (VaR) is one of the tools recommended Bank Indonesia to gauge the risk of an investment, the VaR approach tends to be more associated with the conventional assumption of a normal distribution, while contemporary empirical findings indicate the existence of patterns of abnormality in the nature of statistical data, especially on financial data. Up to this time shares in the Jakarta Islamic Index (JII) is still heavily influenced by the dynamics of market volatility which one, so the necessary in-depth analysis to help investors make the right decisions in investing. This research addresses the issue of risk analysis model using the VaR approach using a variety of model Heterokedastic Timeseris Conditionals (CHT) and find the best model. As for the data used is the daily closing stock index data-Sharia stocks (JII) post-crisis global 2008 (January 2009 – June 2011) and the software used is E-Views 5.1 and Excel 2007. The results obtained are of 16 (sixteen) model approach to VaR-CHT used, only 5 (five) a valid model on a confidence level of 99%, i.e. Approach (2.2) GARCH, GARCH M standard deviation GARCH (1,1), M Log (Variansi) (1,1), TARCH M Log (Variansi) (1,1), EGARCH and M Log (Variansi) (1,1). The VaR Model of the CHT are the best and recommended in analyzing the risks of stock investment is Shariah (JII) is a model that gives the value of the VaR model, i.e. the smallest VaR GARCH-M standard deviation (1,1) that gives the value of VaR is equal to 3.2396%.