AI Summary & Key Takeaways
AI InsightConsidering the AIC and SIC criteria, it was found that the best performing model out of four alternatives was the weighted equation according to the Almon polynomial distributed lags method.
📌 3 Key Takeaways
Abstract
Forecasting the short-term price movements is especially important in terms of developing adequate monetary policies during inflationary periods. For countries such as Turkey where inflation targeting policy were adopted and relatively high inflation rates are observed, making short term forecasting using daily data will allow decision processes to react more rapidly. In Turkey, several methods are used by the Central Bank and academicians for estimating the inflation rate. However, in all these methods, covariates are used from the same frequency (mostly monthly) in modelling the inflation rate. In this study, it has been tried to develop a model which can be used in the forecasting of inflation rate by using MIDAS method which allows the series to be used in the same regression equation from different frequency. In the set regression equation, commercial credit interest rate (weekly), TL / US Dollar parity (daily), gold gram price (daily) and oil price (daily) data are used as variables which have the potential to determine the monthly producer price level (PPI) by increasing the input costs. Considering the AIC and SIC criteria, it was found that the best performing model out of four alternatives was the weighted equation according to the Almon polynomial distributed lags method. The in-sample predictive success of the model was found satisfactory.