Forecasting Inflation in Emerging Markets by Using Phillips Curve and Alternative Time Series Models

The aim of this paper is to investigate the performance of Phillips curve to forecast inflation in a high inflation emerging market country by taking Turkey as a case. For this purpose we compare the forecasting performance of Phillips curve with alternative time series models namely univariate ARIMA model, vector autoregression and vector error correction model and a naive no change model. The data pertains to the quarterly inflation rate in Turkey for the period of 1987-2001. The results show that inflation forecasts obtained from the Phillips curve are found to be more accurate than forecasts based on other macroeconomic variables. The remaining models outperform "no change model" in most of the cases