Content area
Full text
Savings behavior has been investigated across many domains and times. In Malaysia, many Malaysians have trouble saving their income. One challenge Malaysians face is the inability to prepare for financial “shocks,” such as the loss of a job or facing illnesses. Coupled with many Malaysians’ unawareness of the use or benefit of subscribing to insurance policies points to the lack of awareness of the importance of saving.
While the rate of savings among Malaysians has been historically on the low side (Khazanah Research Institute, 2020), a finding in a recent survey showed that most working Malaysians have savings equivalent to less than 2–4 months’ worth of their monthly salary (Goh, 2020). This goes against the rule of thumb of having at least 3–6 months’ worth of a monthly salary as savings (Anong & DeVaney, 2010). While having adequate savings is important, it remains a challenge for many Malaysians. This brings forth a relevant need for policymakers to put increased focus on the savings behaviors of individuals rather than assessing the level of adequate savings for Malaysians.
Against this backdrop, this article proposes a machine-learning-based method to understand how different wealth and consumption categories framed under mental accounting affect savings behavior (i.e., among individuals’ savings decision-making). In addition, this study is envisaged to enrich existing literature on methods of analyzing savings behavior, where machine learning is a complementary method to standard statistical analysis methods in social science studies (Garibay et al., 2022). Moreover, given the limited number of predictive models of savings behavior constructed using machine learning, this study is also aimed at being foundational in nature given that it compares the performance of multiple machine learning models. While the relationships between income, expenditure, and household savings have been established, the influence of mental accounting is not clear from previous studies. A deeper look into how mental accounting influences savings behavior is one of the key objectives of this study. Financial counseling and planning professionals and policymakers have a stake in better understanding the factors influencing a person’s saving behavior. This is particularly important considering governments’ increasingly stretched capabilities in providing adequate social security infrastructure to individuals. Suppose a data-driven method in the form of a machine learning technique with an underlying basis in behavioral economics...