Using TGARCH, TGARCH-M, EGARCH, EGARCH-M, PGARCH and PGARCH-M models Gaussian and non-Gaussian for modeling (EUR/USD) and (GBP/USD) Exchange Rate

Abstract

This paper aims to study characteristics of exchange rate volatility of (EUR/USD) and (GBP/USD) using daily closing prices for two time periods, sub-period form (1 January 2015 until 15 may 2020) and full period from(1 January 2010 until 15 may 2020). This is done by studying (TGARCH,EGARCH, PGARCH,TGARCH-M, EGARCH-M, and PGARCH-M) models. The error term assumed five distributional (Gaussian, Student-t, Student-t with fixed df ,(GED) and (GED) with fixed parameter. The results showed that the EGARCH(1,1) has been the best model selected for the return series of ( EUR/USD) exchange rate in the both full and sub period series. The best model selected for the return series of (GBP/USD) exchange rate it was EGARCH(1,1) model in the full period and TGARCH(1,1) model in the sub-period. All four models selected are based on the Student's t distribution to provide the leverage parameter of the EGARCH model of (EUR/USD) in the full time period and sub-period and the EGARCH model of (GBP/USD) in the full time.