ESTIMASI PARAMETER MODEL ROBUST AUTOREGRESSIVE CONDITIONAL HETEROSCEDASTICITY MENGGUNAKAN FILTER TAU (τ)

(Parameter Estimation of Robust Autoregressive Conditional Heteroscedasticity Model using Filtered Tau (τ))

Abstract

Time series data analysis can be modelling with autoregressive, moving average and smoothing. Autoregressive Integrated Moving Average (ARIMA) is a combination of autoregressive, integrated, and moving average process. Time series in the economic or financial sector often has a volatility effect.  Time series data that has high volatility can cause a heteroscedasticity effect, where the data has a non-stationary mean and variance. Model Autoregressive Conditional Heteroscedasticity (ARCH) are appropriate for modelling of time series data that there are heteroscedasticity effect. In a time series data, it is possible that there are outliers that can cause biased data analysis results. Outliers can be removed from the data, but if the outlier is a value that is not due to an error, it can eliminate information and change the sample size. The purpose of this study was to examine the ARCH model and estimate the parameters of the ARCH robust model. To overcome the outliers, a robust ARCH model is needed. The results of the study obtained a robust ARCH model with a tau filter estimate for data containing outliers.

Keywords: ARCH, filter tau estimation, heteroscedasticity, outliers, robust ARCH