Comments on Default Probability by Employment Status in South Korea

The paper’s main objective is to examine the default rate in different groups of workers in South Korea. The maintained hypothesis is that the loan default rate is higher for the self-employed than the non-self-employed. The default rate is related to borrowers’ characteristics, loan types (unsecured, non-mortgage secured, and mortgages), and risk premiums. The authors define the risk premium as the interest spread between the interest charged by the purpose of borrowing, or by risk assessment (from credit rating agency), and the cost of fund index. Defaults are defined as loans overdue for more than three months. Bank of Korea provides Korea Consumer Credit Panel data from 2011:Q1 to 2017:Q2. The average debt of self-employed borrowers is three times larger than of the non-self-employed. The authors use a dynamic probit model with correlated random effects to avoid excessive standard errors and a biased estimator due to heterogeneity of observations over time. In their model esti-mation, there are 8.9 million observations: 7.9 million non-self-employed, and 923,985 self-employed. It is possible that the authors may obtain the empirical results that are highly overwhelmed by the enormous size of salary workers, which is more than eight times the number of self-employed workers. Loan size varying smallest largest loan size. The amount of loans should have been utilized directly without into the index. The actual loan size would provide more variations of the explanatory variable, rather than using a crude scale 1–10