Is There a J-curve for Azerbaijan? Evidence from Industry-Level Analysis

This paper has attempted to estimate the J-curve phenomenon for Azerbaijan using quarterly industry-level data over the period 2000-2009. A weighted average of the production indexes of Azerbaijan’s four major trading partners, which account for 70% of Azerbaijan’s total trade turnover, was chosen as a proxy for foreign income. Ten non-oil industries, which amount to 97% of Azerbaijan’s non-oil foreign trade, have been analyzed. A recent technique in cointegration has been employed, which allowed for a simultaneous short-run and long-run analysis of the trade balance model. Empirical results suggest that for 3 of 10 cases there is strong evidence for the fulfillment of the Marshall-Lerner condition, as the trade balance in each of those three industries improves in the long run in response to a currency devaluation. This study does not find any systematic presence of the J-curve effect in the short-run, although a few industries do show certain resembling patterns. The price effect is strong and present in almost all industries. All 10 cases exhibit long-run cointegration and are stable according to the CUSUM and CUSUMSQ stability tests.