It is important for policymakers to forecast inflation correctly to struggle with it. There is a wide literature about forecasting inflation with minimum error created by central banks and economists. In this paper, the forecast performance of inflation with different tecniques are examined for the Turkish Economy for the period of 1995-2014. The forecast performance of different tecniques are evaluated by root mean square error, mean absolute error, mean absolute percent error and Theil U coefficient. The inflation forecasts are done by ARIMA, Augmented Phillips Curve, New Keynessien Phillips Curve and Triangle Model. The NAIRU (non-accelerating inflation rate of unemployment) variable in Triangle Model is produced by using two different methods which are Hodrick-Prescott Fitler and Time Varying Parameters.
Key words: Inflation Forecast, Phillips Curve, ARIMA, Hodrick-Prescott, triangle model, Kalman filter. JEL Classification: C53, E37 Article Language: EnglishTurkish
|
