Abstract. In this study, effects of real exchange rate, domestic and foreign income on the external trade balance has been analysed in the framework of the extended Marshall-Lerner condition and J-curve phenomenon using 1995:Q1-2015:Q4 period data for fourteen transition economy and Turkey by means of panel cointegration analysis with multiple structural breaks under crosssectional dependency. Cross-sectional dependence has been analysed with the bias-adjusted cross-section dependence Lagrange multiplier developed by Pesaran et al. (2008) and cross-section dependency has been observed among the countries. Stationary of the series has been tested with the panel unit root test with multiple structural breaks developed by Carrion-i-Silvestre et al. (2005) and it has been found that series are non-stationary in level. The existence of a cointegration relationship between series has been tested with the panel cointegration test with multiple structural breaks developed by Basher and Westerlund (2009) and a cointegration relationship has been observed between the series. Cointegration coefficients have been estimated with the Pesaran (2006) Common Correlated Effects method. According to the analysis result we found evidence that support the extended Marshall- Lerner condition in Belarus, Bulgaria, Croatia, Hungary, Latvia, Poland, Romania, Russia, Slovakia, Slovenia and Turkey. Furthermore, the J-curve phenomenon is valid in Belarus, Romania and Slovenia.

Keywords: trade balance, real exchange rates, cross-sectional dependence, panel data analysis, transition economies.