Question

Section 1: True/False, & explain why in two or three sentences:

4. You are trying to forecast the quarterly US dollar-Euro exchange rate as a function of the difference between US & EU benchmark interest rates in this quarter and the previous quarter. This is an autoregressive distributed lag model.

Answer 1

TRUE,

Literature Review and Implementation of ARDL Approach to Co-integration

The steps followed will be elaborated when using the ARDL technique with co-integration according to the ARDL/Bounds Testing Methodology and after accessing the Error Correction Model (ECM). In the applied research field, there is considerable research leading ARDL with co-integration. Kwofie and Ansah [6] examined the effect of exchange rate and inflation on stock market returns in Ghana, Damane [7] analysed the private consumption in Lesotho, Baig et al. [8] studied the causal relationship between natural disasters and economic growth in Pakistan, Ghumro and Karim [9] dealt with thereal broad money supply as a dependent variable and the gross domestic product discount rate and inflation rate in Pakistan, Sunge et al. [10] established the relationship among the annual time series associated to the household final consumption expenditure as a dependent variable, the gross domestic product, the government expenditure, the government tax revenue, total government debt, that is, the size of public debt and interest payments on outstanding government debt, Mourad [11] analyzed the deposits in Lebanese commercial banks, Dritsakis [12] modeled the real money supply in Hungary with the economic and financial variables composed from the real income at constant prices, the nominal exchange rate and the inflation rate with the main objective was to ensure a long-run equilibrium relationship among these variables, Shahbaz et al. [13] established a strong relationship between the development of the stock market and the economic growth in Pakistan, Tang [14] investigated the money demand function for five Southeast Asian countries and his findings revealed that real M2 aggregate, real expenditure components, exchange rate, and inflation rate were co-integrated for Malaysia, Philippines and Singapore, Oskooee and Wing [15] detected an equilibrium relationship among demand for money, interest rates and exchange rates using quarterly data in Hong Kong. Using the Engle-Granger two-stage procedure to co-integration, Mourad [16] accepted the null hypothesis of no co-integration between the gross domestic saving and each of the gross national expenditure and the household final consumption expenditures as a proportion of the gross domestic product in Jordan, finally Mourad [17] studied the long-run equilibrium between the GDP annual growth rate and the annual growth rate of GDP per capita in each of the GGC countries.In this context, the first question that needs to be answered is: How do we use the PSS technique to construct the ARDL model with co-integration between variables that can be either I(1) or I(0)? The PSS procedure aims to test the null hypothesis that there exists no co-integrating relationship between variables versus the alternative hypothesis which recommends a long-term equilibrium relationship between variables, i.e. co-integration. The PSS procedure tests the existence of this equilibrium relationship regardless of whether the underlying variables are I(1), I(0) or I(d), (0<d<1( (Fractionally integrated). This means that if one variable is I(2), then the PSS procedure becomes invalid. Below we present the important steps needed to implement PSS procedure.Step 1: The stationarity of the variables is tested to confirm statistically the existence of time series that follow I(0) or I(1) and without variables integrated of order two in the model.Step 2: Since we have only two variables, the Unrestricted Error Correction Model (UECM)will be considered: