Household indebtedness and non-performing loans in Malaysia / Theong May Jin

• Main Author: Theong , May Jin
• Format: Thesis
• Published: 2021
• Subjects: HA Statistics; HG Finance
• Online Access: http://studentsrepo.um.edu.my/14265/1/Theong_May_Jin.pdf; http://studentsrepo.um.edu.my/14265/2/Theong_May_Jin.pdf; http://studentsrepo.um.edu.my/14265/

Aggregate non-performing loan (NPL) ratios in Malaysia have been decreasing steadily. However, there has been increasing awareness concerning predicting banking vulnerability and systemic risk in the financial market when debt grows faster than the economy. At the end of 2018, Malaysia’s household debt fell to 83% of the GDP ratio. In fact, in value, total household debt grew from RM1.08 trillion in 2016 to RM1.18 trillion in 2018. Besides, household borrowing has been increasing over recent years. It comprised about 57% of Malaysian banks' total lending, as of 2018, which has exposed the banks to ex-post credit risk. Aggregate NPLs have been considered the transmission channel of macroeconomic shocks to the banks' balance sheets. However, another strand of literature on analysing NPL by category has been less noticed by the researchers. At times of stress, these relationships may be nonlinear. For these reasons, first, the present study examined the relationship between four common household credit facilities and Malaysian household NPLs, controlled by a set of indicators. Narrowing down to household NPLs by economic purpose, this study examined the chosen sets of economic indicators that determine the former movements and their relative strength in their impact. The study also compared the appropriate approaches in capturing the dynamics (linear versus nonlinear) between the tested variables. Using an available monthly dataset of macroeconomic and monetary variables from January 2006 to December 2018, linear and nonlinear autoregressive distributed lags (ARDL) were employed. The findings showed that finance-constrained households were more likely to default on credit card loans than other types of loans during times of financial distress. Household NPLs and personal loan debt were found to be better explained in a linear specification, in a negative direction, assuming consumers tend to pay off other debt using easy to access loans at any time. There was no long-run relationship between them—properties outstanding loans and household default using the ARDL approach. However, NARDL revealed a possible relationship between loans and NPLs. Both positive and negative property loan changes were significant, whereas negative changes significantly impacted household NPLs. In another direction, the results showed that macroeconomic shocks affected consumer and mortgage loans differently; however, asymmetrical changes of the overnight policy rate (OPR) did not contribute to any type of NPLs. A hike or cutback in the OPR was a response toward the economic outlook and price changes; hence the effect of the OPR was muted, comparatively. In addition to the macroeconomic determinants, this study assessed how the loan portfolio affected each type of NPL. When other debts were factored in, credit card loan borrowers with another loan (s) tended to default on their credit card debt, compared to those without credit cards. Personal loan defaults were not linked to household debt, as personal loan borrowers usually covered their debts by taking up additional personal loans. Lastly, high housing loans and high vehicle loans contributed to the high default rate of residential properties and transport vehicles debts, respectively.