Time Series Analysis and structural break detection: A case of Zambia’s CPI.

Abstract

Purpose: By empirically examining Zambia’s CPI between 2010 and 2020, the study attempts to determine the structural change in the time series. The CPI is one of the most important variables for analyzing inflation in macroeconomics, therefore any change in the dynamic must be determined. In this paper change points and dates are highlighted and statistical analysis methods have been employed to explore and discover the underlying patterns and trends of Zambia’s CPI for the past 10 years. Methodology/approach: Secondary Data from Zambia Statistics Agency (ZamStats.gov.zm) was used for the Study. From 132 elements of observations of time series for 10 years, the detection methods of structural change were employed. The Cumulative Sum Tests (CUSUM test) of Ordinary Least Squares (OLS), Andrew Sup F test, Bai and Perron test, and Chow test were used to detect the model stability and verify the hypothesis using P-value. Results: The results show that there were five (5) Structural changes or breaks in mean and variance and these were February 2012, February 2014, October 2015, October 2017, and May 2019. The structural breaks are highly suggestive as they appear to broadly coincide with readily identifiable macroeconomic events, increased stock of external debt following the issuance of Eurobonds in 2012, 2014, and 2015, rise increased food prices arising from the adverse impact of erratic rainfall on agricultural output and the pass-through from the depreciation of the Kwacha. Policy Implication: Based on the study, strong and sound macroeconomic policies are needed to be implemented: Such as debt management and diversification of foreign exchange sources, and increased earnings.