نوع مقاله : مقاله پژوهشی
نویسندگان
1 استادیاراقتصاد، گروه اقتصاد، واحد تهران غرب، دانشگاه آزاد اسلامی، تهران، ایران
2 دکترای اقتصاد دانشگاه آزاد اسلامی
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
Purpose: One of the most important economic problems in Iran during the last few decades is the phenomenon of high and double-digit inflation. So, improving the conditions caused by high inflation has always been one of the important goals of the country's development programs. Achieving this goal requires the creation of a precise and targeted mechanism in the economic policy-making process. In its standard form, it includes forecasting, goal setting and policy analysis. The general purpose of this study is to predict the inflation rate using economic variables that affect it. In this research, the Bayesian averaging method is used to investigate the best estimation model that can predict the inflation in Iran. In this regard, the previous studies conducted in this field are first reviewed, and then the most important economic variables affecting the inflation are identified and used to predict the inflation rate.
Methodology: Friedman believes that inflation is always and everywhere a monetary phenomenon. Monetarists believe that inflation comes from the disproportionate growth of nominal money supply. So, the higher this growth, the higher the inflation rate is. There is a direct and proportional relationship between money growth and inflation. According to this theory, changes in money supply have no effect on real variables such as production, employment and real wages; they only affect nominal variables such as prices and nominal wages proportionally. Monetarists consider the real growth of the economy in the long term to be independent of the change in the money supply and generally believe that this growth is determined by factors such as production capacity, increase in labor force due to population growth, advancement of technical knowledge and natural resources.
In order to control or curb inflation, the influencing factors must be identified. The results of the studies on the factors that cause inflation are different or even inconsistent, because it is based on the researcher's specific attitude. In this article, to avoid falling into such a vortex, the Bayesian averaging method is used to predict inflation. The seasonal data of 1990-2022 have been used to predict the inflation rate in Iran.
Results and discussion: Forecasting inflation is one of the most important but difficult issues in macroeconomics. Many different approaches have been proposed in this field. Perhaps the most popular of these approaches are those based on the Phillips curve. However, the general framework includes a dependent variable such as inflation (or change in inflation) and explanatory variables such as inflation breaks, unemployment rate and other predictive factors. Meanwhile, return and regression-based methods have been somewhat more successful.
The results show that dynamic model averaging leads to significant improvements in forecasting compared to other approaches such as OLS, ARMA, and ARDL. Also, among the variables influencing inflation, the most influential for predicting the inflation rate relates to household consumption expenditure, unemployment rate, and workers' wage rate.
Conclusions and policy implications: The general purpose of this study was to predict the inflation rate using the economic variables that affect it. In this research, the Bayesian averaging method served to investigate the best estimation model that can predict the inflation in Iran. In this regard, the previous studies conducted in this field were first examined and then the most important economic variables affecting the inflation were identified and used to predict the inflation rate. For this purpose, the seasonal data of the variables during the period of 1990-2022 were used. In this research, the methods of ordinary least squares (OLS), auto regression moving average (ARMA), auto regression with distributed lag (ARDL), Bayesian dynamic averaging (DMA) and Lasso regression were used to predict the inflation and evaluate the prediction accuracy.
Therefore, based on the results obtained in this research, attention should be paid to the behavioral economic parameters of households when choosing the optimal policy. Factors such as the increase in household food prices, the high increase in workers' wages, the increase in money supply, the increase in interest rates and the increase in residential rental rates have definitely caused the formation of inflationary expectations in the society and can cause instability in the future and make the inflation deviate from equilibrium. Thus the measures to take include controlling the household food market, controlling the housing market, reforming the wage pattern in the country, controlling the interest rates in banks, and using contractionary monetary policies. These help to control and reduce inflationary expectations among the people.
کلیدواژهها [English]
)