Ali Taiebnia, Hossein Amiri, Fatemeh Ravishi,
Volume 18, Issue 4 (1-2014)
Abstract
The harmful effects of chronic high inflation in the economy led the governments and country’s monetary authorities seek to reduce or eliminate this phenomenon. Therefore it’s very important to predict how inflation moves providing an appropriate economic model is a crucial factor to forecast inflation, so on. In this regard, in the present research, we attempt to generate a appropriate model for forecasting inflation using New Keynesian Phillips Curve. At first, purely forward-looking New Keynesian Phillips Curve and hybrid New Keynesian Phillips Curve to be estimated by the analysis of pricing models and the discussions of stickiness of wages and prices. This means that, we estimate purely forward-looking New Keynesian Phillips Curve, hybrid New Keynesian Phillips Curve by generalized method of moments and also estimate hybrid New Keynesian Phillips Curve by autoregressive distributed lag method and autoregressive moving average for the period Q - Q . For the evaluation of the performance of the different models over various forecast horizons, finally, we construct series of -step, -step and -step-ahead out-of-sample forecasts over the period . The results show that the hybrid New Keynesian Phillips Curve provides relatively more accurate forecast of inflation in Iran compared to the other models for middle forecast horizon while they are outperformed by the time series model (ARMA( )) only for the short forecast horizon and by ARDL( ) for long forecast horizon.
