Edited by Lawrence R. Klein
Chapter 9: Current Quarter Model for Turkey
Süleyman Özmucur INTRODUCTION 1. Forecasting of economic activity requires the use of all available information. However, key data are collected at different frequencies. This necessitates building models utilizing data at different frequencies, which was the starting point for high-frequency macroeconometric models initiated by Klein and Sojo (1989). Forecasts are useful not only for studying the short-term developments of the economy, but also for adjusting lower-frequency macroeconometric models so that they are solved from up-to-date initial conditions (Klein and Sojo, 1989; Klein and Park, 1993, 1995). We propose to form principal components of the monthly indicators whose periodic values appear at either different or similar time points of each month. Principal components analysis is based on our general point of view that a country’s (any country’s) economic growth is highly multivariate. No single measured economic activity can account for anything as complex as a modern economy. We examine many time series, and select those that seem to have a priori importance. In order to conserve degrees of freedom, we narrowed the list of explanatory variables in the relevant regression as much as possible. There have been important motivations in adopting the principal components methodology. The chapter is in six sections. The second section deals with the methodology of principal components. A brief summary of the basic features of the Turkish economy and available monthly data are presented in the third section. Section 4 is devoted to the estimation of the model. Model performance is presented in Section 5, and forecasts...
You are not authenticated to view the full text of this chapter or article.