Title: Analyzing the volatility of daily time series FOREX rates by using GARCH models: A case study from Albania Authors:  Altin Kulli - Qiriazi University College (Albania) [presenting]
Abstract: Time series properties and predictability of exchange rates in transition economies considered. A period 2009-2011, prior euro daily Albanian exchange rates, with respect to four major currencies (USD, DEM, drachma and the Italian lira), were considered. The forex movements, as the most important factors affecting sales, profit forecasts, capital budgeting plans and international investment value, are of particular significance for the Albanian economy as an emerging country. The time-varying characteristics of forex volatility were explored using the GARCH methodology, discovering that the GARCH models might adequately describe the conditional second moments of these exchange rates of Albanian LEK; furthermore, upon consideration of past information, the percentage changes of such FOREX rates be predictable. The first lagged percentage changes for dollar and drachma were both significant and approximately the same, which is to be expected for an emerging economy, indicating market inefficiency. The volatility is well represented by a GARCH (1, 1) (it might be predictable on past innovations basis, volatility measures). The conditional variance shock did not persist and quickly died out. The weekend effect on trade opening was found to be positive for all except the USD series, which had its value equal to zero. The volatility has increased from Friday to Monday for all, mainly for Mark and Drachma.