ANALYSIS OF FLUCTUATIONS IN ECONOMIC ACTIVITY IN TURKEY: A DYNAMIC FACTOR MODELING WITH ALTERNATIVE FILTERS

Abstract

This study is analyzed fluctuations in economic activity in Turkey with the help of dynamic factor models. Economic activity, which cannot be observed based on the observed variables such as industrial production, employment, exchange rate, interest and price movements, current balance and change in stock market index, has been estimated. A two-step forecasting process was adopted and the period between 2005 and 2020 was analyzed with monthly data. Final factors and factor loads were calculated with alternative filters using the initial values calculated by the principal components method. In this context, Kalman filter and Kalaba and Tesfatsion?s flexible least squares estimator are used. These estimators were found to have very close estimation performances. Factor estimation reflecting economic activity showed that after the Global Crisis in 2008 and the exchange rate shock in August 2018, two important contractions occurred. It was found that the economy reached the highest expansion level in the sample period in early 2011.