A Multivariate Time Series Analysis of Household Debts during 2007-2009 Financial Crisis in South Africa: A Vector Error Correction Approach

Abstract

In this study, we adopted multivariate time series approach to investigate the effect selected macroeconomic variables on household debts using South African data collected from 1990 Q1-2013 Q1. SAS 9.3 was used for the analysis. Johansen multivariate cointegration and the Toda-Yamamoto causality testing approaches were adopted. The ADF and KPSS unit root test results confirmed that all the variables under study were stationary. Cointegration results allowed the acceptance of four cointegrating vectors. The ECM revealed a negative sign implying the existence of a long-run equilibrium relationship among the variables. The coefficient of the error term implied that 13.8% of disequilibrium due to negative and positive shocks was previously corrected in one period per quarter. This adjustment provided the model with all the signs concurring with economic theory and literature on household debts. Toda-Yamamoto causality results revealed a unidirectional relationship running from GDP, UR and ER to HHD. A feedback relationship was discovered between the remaining variables confirming they are not weakly exogenous in South African household debts systems. This study recommends the formulation and implementation of policies that will effectively help in dealing with household indebtedness in SA and also to complement the NCR. More emphasis should be placed on threatening determinants such as household consumption expenditure, gross domestic product and house prices. DOI: 10.5901/mjss.2014.v5n7p107