The analysis of J and S curve and Harberger-Laursen-Metzler effects in Turkey's bilateral trade

Relationship between exchange rates and trade balance is explained by three different approaches to the balance of payments (BOP): elasticity approach, absorption approach, and the monetary approach. In this study, in addition to the J curve effect initially introduced by Magee (1973), Junz and Rhomberg (1973), and Meade, E. E. (1988) through the inclusion of time into the elasticity framework, which claims that as a result of the devaluation the trade balance would deteriorate prior to an improvement in the short run, the S curve effect formulated by Backus, Kehoe, and Kydland (1994a) which states that the trade balance is negatively correlated, in general, with current and future movements in the terms of trade (TOT) (or alternatively the real exchange rate), but positively correlated with past movements, is also investigated in the bilateral trade between Turkey and her 16 selected trading partners.These dynamics are examined through the analysis of the behavior of cross-correlation functions computed between the annual bilateral trade balance and real exchange rate series for the period 1980-2010. The series are detrended by using Hodrick-Prescott filter (Hodrick and Prescott, 1997) before the computations. Besides the J and S curve effects, this procedure enables to investigate the Harberger-Laursen-Metzler (HLM) effect introduced by Harberger (1950) and Laursen and Metzler (1950) which puts forward by a Keynesian perspective that trade balance would deteriorate due to a deterioration in the TOT.It is hypothesized in this dissertation that Turkey may be able to improve her bilateral trade balance in the short run by using exchange rate adjustments but can not affect foreign prices and that a deterioration in the TOT may lead to an increase in her bilateral trade deficits. Empirical findings suggest that J curve and HLM effects are present between Turkey and 9 out of 16 trading partners. Moreover, there is an inverse J curve effect between Turkey and 4 out of 7 remaining trading partners and a delayed J curve effect for the last 3 remaining trading partners. Last but not least, I was unable to find an S curve effect in foreign trade neither in the bilateral nor in the aggregate levels. This finding is consistent with the fact that Turkey is a small country and can not affect foreign prices. By taking these findings into account, it can be put that exchange rate policies play a key role in reducing Turkey?s bilateral trade deficits especially in the short run and that a deterioration in TOT deepens these deficits.