الفهرس 
Back groundOnly 14 pages are availabe for public view
 |  | from | 143 |  |  |
| | | from | 143 | | |
Abstract
Forecasting of exchange rates has been an extremely challenging and important task for both academic and business researchers. These forecasts are essential for central banks, corporations, investors and even individuals to hedge exchange rate risks and to generate profits. This study aims to propose a statistical model to forecast the exchange rate of the Egyptian pound per U.S dollar (LE/$US) for both short-term and long-term periods. Several economic variables that were expected to have an impact on the LE/$US exchange rate were investigated. The study covered the period from February, 2003 to July, 2014 using monthly and daily LE/$US exchange rates.
Different statistical models were applied, for short-term forecasting, univariate autoregressive integrated moving average (ARIMA) model together with a hybrid model that combines the ARIMA model with the generalized autoregressive conditional heteroscedasticity (GARCH) model were applied using the daily exchange rate series. For long-term forecasting, univariate ARIMA model together with a dynamic regression model that combines the multiple regression analysis with the ARIMA model were applied using the monthly exchange rate series. Several economic variables were included in the dynamic regression model. The forecasting performance of all models estimated were evaluated using different forecasting accuracy measures that were based on both in-sample and out-of-sample forecasts.
The results showed that for the short-term forecasting model, the hybrid ARIMA-GARCH model outperformed the univariate ARIMA model in terms of forecasting accuracy. As for the long-term forecasting model, the dynamic regression model outperformed the univariate ARIMA model.
These models may aid the Egyptian authorities to deal with the disequilibrium in the foreign exchange market. Policy makers, corporations, banks, individuals and foreign currency dealers, may also use these models to generate forecasts of the LE/$US exchange rate and to hedge against exchange rate risk.
Keywords: Exchange rate forecasting, Autoregressive integrated moving average, ARIMA, Dynamic regression, Regression with ARIMA errors, Generalized autoregressive conditional heteroscedasticity, GARCH, Hybrid ARIMA-GARCH, Technical analysis, Fundamental analysis.