Analyzing the nonlinear dynamics of the Turkish textile and apparel industries

Abstract

Developments in the Turkish textile and apparel industries carry a high influence on the development of Turkish economy with their substantial shares in the total industrial production. This is the first study that documents the cyclical economic asymmetric behavioral dynamics of the Turkish textile and apparel industries by utilizing the Markov regime switching models to the mean and variance at monthly frequencies by using Expectation Maximization algorithm. The results provide a further detailed insight about the asymmetric dynamics of these sectors by examining nonlinearity, determining the number of regimes, identifying the regime dependent variances, calculating the transition and smoothed probabilities, and documenting the regime classifications with respect to each identified regime. The study also employs the estimated transition probabilities to define and quantify the duration and persistence of staying in each particular regime for each of the sectors. Revealing these results about the Turkish textile and apparel industries gives a better understanding for policy makers in terms of designing effective policies to increase competitiveness in the global markets.