Comparison of ARIMA and ARIMA Error Regression Models: Evidence from the Russian Consumer Price Index

Abstract

Consumer Price Index (CPI) is regarded as a common approach for measuring inflation. The present study examines the Russia CPI in the context of political and social upheavals, especially under the wars and the COVID-19. Due to the unstable political situation, the inflation rate in Russia sharply grow in 2014 and 2022 which creates two of the largest increase over the past decade after the Crimean war and the Ukrainian war were announced, thus it is crutial to make both long-term and short-term trends prediction. The paper aims to choose the best fitting model to estimate the future value of Russia monthly CPI data, and ultimately provides a suggestion and reference for monetary and fiscal authorities in Russia when making the policies to reduce inflation risk. The study applies the ARIMA (Autoregressive Integrated Moving Average) model to forecast Russia CPI, and uses regression with ARIMA errors to analyze the Russia CPI according to the ARIMA fitted values of the crude oil price in Europe. According to the result, ARIMA (2,3,1) is the best fitting model that can make comparatively sensible prediction for the future values.