신용회복제도의 도덕적 해이 유발가능성에 관한 실증연구

When the credit recovery plan offers big advantages such as delays in repayment of debts and exemption of interests, debtors with moral hazard will not make repayments but intend to take benefits from the plan. As consumers with such a behavior increase, the society will have to bear the massive economic losses. In this sense, the plan is controversial due to the possibility to stimulate moral hazard. The plan does not always give benefits to its applicants. Applicants for the plan have to take disadvantages as their information is submitted and strictly managed by all financial organizations. In addition, they have to bear the direct expenses including the legal counseling fee for application, and the indirect expenses of losing social reputations. Accordingly, introduction of the plan is not always regarded as a stimulator of voluntary defaults. The plan should be considered from the long-term perspective and in the context of the national economy, since the plan serves as the social safety net allowing financially isolated citizens to normally make economic activities again. Most of the previous literatures, studied mostly in the US, found the empirical evidence that the plan increased voluntary defaults demand, and expressed concerns over moral hazard of delinquents. This study aims to verify whether the credit recovery plan, which has been introduced or being introduced in Korea, has increased the number of delinquent cardholders. For this, we compared the three cases before and after introducing the plan specified in <Table 1>, in order to verify the influences of the plan on the voluntary delinquency demand. As classified in Chapter 2, the three cases included the work-out plan for multiple debtors (announced in July 2002), the overall rescue programs for credit delinquents (announced in March 2004), and the credit recovery plan for financial delinquents with low income (announced in March 2005), and the cases were analyzed based on the announced month. Using logistic regression model, we control other factors influencing the delinquency rate used in Min et al.(2007) and Gross-Souleles (2002), and add the dummy variable implying the plan enforcement (Six months before the introduction : 0, and six months after the introduction : 1). If the introduction of the plan increases voluntary delinquents, the coefficient of dummy variables will show a significant positive value. Also, we analyzed the influences of the plan on voluntary delinquents with the differences between the estimated delinquency rate and the actual delinquency rate. This approach was added to improve robustness of the results of the formula (3.1). First, I excluded dummy variables and calculated the estimated delinquency rate through the logistic regression model. I calculated the difference between the estimated delinquency rate and the actual delinquency rate, and then verified the hypothesis by finding out the significance of the difference. As empirical data for the study, we used data of a large Korean credit card company. Not all applications for the plan are resulted from the use of credit cards. However, most of studies including major researches on the plan such as the one of Fay, Hurst, and White (2002), and the one of Gross and Souleles (2002) depended on data of credit card users. In addition, according to the surveys in Korea, 67.3% of respondents answered that credit card debts resulted in application for the credit recovery plan, proving that the credit card data is representative. In this study, credit card users were classified into three grades to analyze the influences of the plan on each group of users. According to the credit rating, users from grade one to grade four were classified as “High grade Users,” those from grade five to grade six as “Potential group,” and those in grade seven as “Marginal group.” The finance limit utilization rate was used for detailed classification of groups. Those of grade five or higher show the utilization rate of over 1. The rate of over one means that the card users make most of their consumptions through borrowings. It is expected that card users with high limit utilization rates will have strong consumption tendency. As they have already used most of the credit limit, they have high potentials to become delinquents when adverse events occur. The results of the analysis are as follows. As neither overall groups nor the potential groups showed meaningful empirical evidences on possibility of moral hazard in all the three cases, the hypothesis I that the plan increases voluntary delinquents demand was not supported. In other words, the empirical evidence of the possibility of moral hazard caused by the credit recovery plan has not been discovered during the analysis made in this study. This conclusion does not mean that Korean delinquents have the lower level of moral hazard than foreign delinquents, for previous foreign studies were made in late 1990’s when the credit card delinquency rate and the number of bankrupt individuals skyrocketed, while this study was conducted at the point where the delinquency rate was showing a steady decline. Furthermore, it is likely that the aggressive risk management of credit card companies may limit the increase of voluntary delinquents. Failure to control such factors is a limit of this study model. This study is differentiated with previous foreign studies in a sense that it used the case methodology enabling direct measurement of the effect of the credit recovery plan. This study is also meaningful in that it is the first study in Korea aimed to analyze the impact of the financial plan on decision-making for financing and delinquency of consumers, the key member of the economy. we hope this study will serve as a catalyst for studies on diverse functions provided by the credit recovery plan. we also hope that the domestic academic circle will realize the significance of the consumer financial market, and in particular, the studies on the financial markets mainly used by low-income citizens like the credit card market will be further stimulated.