Predicting the Rate of Inflation in Uzbekistan using seasonal ARIMA models

Abstract Adopting the univariate time series modeling, particularly the ARIMA and seasonal ARIMA models, the current study attempts to forecast the rate of inflation for the Uzbek economy. The current macroeconomic monetary policy regime in case of Uzbekistan is to contain the rate of inflation and the economic policies and reforms of the country are also targeting to bring the medium-term inflation to about 5% in near future. Utilizing the data on month-on-month inflation and year on year inflation published by the Central Bank of Uzbekistan the study tries to fit an appropriate model looking at the seasonal fluctuations and cyclical trends as embedded in the inflation data. An in-depth econometric analysis is carried out using the univariate time series modeling and appropriate statistical test such as augmented Dickey fuller test for unit root the auto correlogram function and the partial auto correlogram function. As per the results of the test an appropriate autoregressive integrated moving average model is fitted of the relevant order to forecast the month on month and year on year inflation rate for next five months commencing October 2023. The predicted values of the rate of inflations are also tested against the actual realized values and the result of the study shows that the predicted values are approximate to the actual realized inflation rates. This leads to strong policy directions which will be helpful for the monetary authorities to direct the policy towards achieving the objective of targeted inflation rate for the Uzbek economy.