Real Exchange Rate and Economic Growth : Turkey

This paper assesses the relationship with real exchange rate and growth using quarterly data of 1989:Q1-2005:Q2. Two groups of models are used in the study that is held. The first model, which was considered as the core model RER, PPI, and GDP, are involved whereas import and export are added on former variables in the expanded model. Integration level of the variables are investigated using with DF, PP, KPSS, Ng-Perron Tests and according to the test results, it is decided that all series are first order integrated. The empirical analysis is started with the application of bivariate data analysis held for RER and GDP variables to study the relationship between them. RER and GDP series are used together with different transformations of these series and seasonally adjusted version of these series are used so that cross correlation values of these variables are calculated as full sample and for a sub-sample. The attained results showed that 1989:Q1-2001:Q3 subsample and full sample had differentiations in values and in terms of statistically significance. Using Johansen Cointegration Test this paper finds evidence that one cointegration vector based on two groups of variables. Vector Error Correction Models were estimated that incorporates the long run behavior variables and short run adjustment dynamics for both two models. For both of these models ImpulseResponse Functions and Variance Decomposition Analysis studied. Formed impulse-response functions, a positive RER shock increases GDP in the core model for the first three periods but then decreases. In the other model on the other hand, it increases during the first four periods and after the observed decrease it continues its movement in the seasonal fashion. Before Variance Decomposition is started, series are aligned by using Block Exogeneity Test the alignment of the series in the model effects the results of this analysis. Impulse-response functions shows that the positive RER shock increases GDP in the core model for the first three periods but then decreases successor periods. In Variance Decomposition Analysis, it is evident that the sources of variance in output are the own shocks and also observed that RER's explanatory ratio on GDP does not disappear in the long run