Empirical Results of Modeling EUR/RON Exchange Rate using ARCH, GARCH, EGARCH, TARCH and PARCH models

• Main Authors: Andreea – Cristina PETRICĂ, Stelian STANCU
• Format: Article
• Language: English
• Published: Romanian National Institute of Statistics 2017-03-01
• Series: Revista Română de Statistică
• Subjects: Exchange Rate Volatility; Heteroscedasticity; Symmetric GARCH Models; Asymmetric GARCH Models; Fat-tails; Volatility Clustering; Leverage Effect
• Online Access: http://www.revistadestatistica.ro/wp-content/uploads/2017/03/A4_RRS1_2017.pdf

The aim of this study consists in examining the changes in the volatility of daily returns of EUR/RON exchange rate using on the one hand symmetric GARCH models (ARCH and GARCH) and on the other hand the asymmetric GARCH models (EGARCH, TARCH and PARCH), since the conditional variance is time-varying. The analysis takes into account daily quotations of EUR/RON exchange rate over the period of 04th January 1999 to 13th June 2016. Thus, we are modeling heteroscedasticity by applying different specifications of GARCH models followed by looking for significant parameters and low information criteria (minimum Akaike Information Criterion). All models are estimated using the maximum likelihood method under the assumption of several distributions of the innovation terms such as: Normal (Gaussian) distribution, Student’s t distribution, Generalized Error distribution (GED), Student’s with fixed df. Distribution, and GED with fixed parameter distribution. The predominant models turned out to be EGARCH and PARCH models, and the empirical results point out that the best model for estimating daily returns of EUR/RON exchange rate is EGARCH(2,1) with Asymmetric order 2 under the assumption of Student’s t distributed innovation terms. This can be explained by the fact that in case of EGARCH model, the restriction regarding the positivity of the conditional variance is automatically satisfied.