Measuring Annual Real Exchange Rate Series for Turkey

This study deals with the problem of measuring the real exchange rate (RER). We consider four aspects of this measurement problem: (a) Using end-of-period or period averages of the nominal exchange rate. (b) Choosing price indexes. (c) In obtaining the real effective exchange rates (REER), deciding upon the number of trading partners in calculating the weights. (d) Deciding upon the formula to use in aggregation. Considering all these aspects together led to the calculation of a great number of alterative series. Our analysis of these series yielded the following conclusions: (1) The end- of- period based results reflected the dates of the major devaluations more accurately but the period average based results gave us a more conservative picture of RER behavior. (2) The consumer price index (CPI), the wholesale price index (WPI), and the GDP deflator (GDPD) were used as alternatives. When the same price indexes were used for both domestic and foreign prices, we found that the GDPD-based series appeared to overstate the depreciations and appreciations in the real exchange rate while the WPI-based results were the least volatile. When different price indexes were utilised, it was found that all series indicated changes in the competitiveness of Turkish tradables and nontradables to be in the same direction with a few periods of conflict. (3) We used four formulas to obtain the REER. In terms of the similarity in their results, we obtained two pairs. This pairing also showed itself in the sensitivity of these formulas to increases in the number of trading partners, which was chosen to be 5, 9 and 14. The sensitivity was observed when going from 5 to 9 trading partners. (4) When comparing the results from these pairs, no consistent disparity was obtained and conflicts were observed in very few cases. Hence, the choice within each pair or across pairs would be based on the assessment of the investigator as to which is easier to compute.