ِAnalysis of the Exchange Rates Using the ARCH Model

Mahmoud N. Murad, Hussain Badran


In this research, we performed a deep analysis of two main variables concerning the Euro/Dollar and Yen/Dollar exchange rates (daily data that cover the period from 1-1-2003 to 31-7-2004), as well as analyzing several variables based on them. To reach our objective, we benefited from the different techniques and tests that permit us to decide about the normality, linearity, asymmetry and, particularly, the use of the evolutive skewness and kurtosis. We have used the methods associated to the homoscedastic Autoregressive models (AR model) and Autoregressive Conditional Heteroskedastic model (ARCH model). A numeric comparison between the conditional variance and the unconditional variance has clearly been achieved. We accomplished some forecastings for the August of 2004 and we estimated the forecast intervals taking into account the presence of a volatility. The quality of forecasts has been measured using the MAPE criteria. The results of our research show the importance of the applications of the AR-ARCH techniques on the time series taken in first difference. Finally, we proposed a pursuit of this research acting the technical multivariate GARCH model.


Exchange Rate, Volatility, ARCH Model, Forecasts.

Full Text:



  • There are currently no refbacks.