Modeling Wave-Like Patterns of Consumer Price Index of the USA

The Consumer Price Index (CPI) indicates whether the economy of a country is experiencing inflation, deflation, or stagflation. Changes in the CPI are used to assess price changes associated with the cost of living. There is an increasing trend of CPI in the USA. This study design forecasts the CPI of the USA. Monthly CPI data of the USA for the period of January 2011 to October 2021 were obtained from the International Monitory Fund (IMF) database. Sama Circular Model (SCM), Seasonal Autoregressive Integrated Moving Average (SARIMA), and Holt's-Winters three-parameter additive and multiplicative models were tested to forecast the CPI of the USA. The model assumptions were tested by Ljung-Box Q (LBQ) test, Anderson Darling test, and Auto Correlation Function (ACF). The forecasting ability of the model asses by both relative and absolute measurements of errors. Results of the study revealed that the SCM is not suitable due to the autocorrelation of residuals. SARIMA and the Holt-Winters’ models can be used to forecast the CPI of the USA. But the Holt-Winters additive model of alpha; 0.999, gamma;0.900, and delta; 0.999 is the most suitable model to forecast CPI in the USA. The rate of change of CPI was identified as 6 months. It is strongly recommended to design more studies on capturing the behaviors of the CPI for other countries