서민금융기관의 대출금리, 비용구조 및 수익성에 관한 연구 : 해외의 영리 마이크로 파이넌스 기관을 활용한 실증분석

Structural issues in the lending market for low income and less creditworthy individuals could be summarized into two topics. The first is market inefficiency, that is, adverse selection and moral hazard caused by imperfect information, and the second is high unit cost of loans to service customers with small loan amount. Although there were many researches about financial access and availability of credit for less creditworthy individuals recently in Korea, they focused mostly on the institutional aspect of the lending market, and rigorous economic analyses were not performed enough to support policies in that area. The purpose of this paper is to provide an empirical knowledge of this special lending market, based on a theoretical model of lenders’ behavior. The theoretical model is developed by extending the historical research of Stiglitz and Weiss (1981) on credit rationing to the specific decision making process of lenders under imperfect credit information. Based on the lending market model, authors test the model using international commercial microfinance institutions, and interpret the results to enhance understanding of the lending market for low income and less creditworthy individuals. Economic models of loan loss rate, interest rate and profitability are theoretically derived, and they are tested as a simultaneous equation system. Also, cost structure of the loan is analyzed focused on the impact of economies of scale in operating cost. Although the empirical test is performed with data from international microfinance institutions, the analysis demonstrates a possibility of developing the special lending market based on commercial principles, and provides lessons for the related policy in the area, as well as a rigorous test of the theoretical model. In chapter 2, under the assumption of imperfect information, a theoretical model of lenders behavior is developed for lending interest rate, loan loss rate, and profitability. The models are developed by applying the original model of Stiglitz and Weiss (1981) on credit rationing. But it is extended by incorporating lender’s decision making process under imperfect information with specific consideration of important variables in banking, such as error rate in the credit decision making process, default rate, recovery rate, and market competition etc. Chapter 3 and 4 summarize previous researches in this area, and describe data and methodology of this research. Most of the previous researches came from microfinance industry in relation with its commercialization and outreach. But despite their significant contribution, they have limitation in some areas, such as the scope of data and method of empirical test. To address those issues, this paper adopts a differentiated approach. At first, different from the previous researches that estimated interest rate and loan loss rate as separate variables, this paper builds theoretical and empirical models as a simultaneous equation system, and adopts an appropriate econometric methodology. Secondly, previous researches on microfinance industry included all types of institutions including small NGOs. But, this paper uses only for-profit bank and nonbank financial institutions in order to focus on the economic behavior and opportunities in the market. Thirdly, a panel data is created to cover 182 microfinance institutions over six year period from 2005 to 2010 with Mix’s public data. Six year sample which is consistent with one credit cycle will provide an unbiased view of those financial institutions’ performance in the long term. Also, this data set makes possible a panel regression methodology, which is more useful than a simple cross sectional approach adopted by other researches. In Chapter 5, results of the empirical test are summarized. Using the panel data, two approaches are taken. The first is a regression analysis on the simply pooled data without consideration of uniqueness of each individual institution. The second is a random effect panel regression which incorporates unique values of each individual institution in each regression. The regression analysis of the simply pooled data shows an evidence of strong form of adverse selection, that is, a decreasing profitability of loans over a threshold interest rate, as well weak form adverse section, that is, an increasing loan loss rate with interest rate. But, in the random effect panel regression, the effect of interest rate on both loan loss rate and profitability is reduced to only marginally significant level. Also, the average loan loss rate of all sample turned out to be in the range of 1.5~2.5%, and average ROA was over 1%, both of which are better than general perception of the lending market for the poor. Based on those findings, authors contend that market inefficiencies caused by imperfect information is not such a serious barrier in extending financial access for the low income borrowers, at least in the observable range of lending interest rates. On the other hand, the most important factor in the high interest rates of small loans for low income borrowers was found to be the high operating cost, and the determinants of operating cost were analyzed focused on the impact of economies scales in two dimensions, that is, average loan size and number of outstanding loans. In addition to the average loan size, the number of loans outstanding was proved to have very significant impact of economies of scales, over a size of an institution which had been believed to have a maximum benefit of economies of scale. Also, market competition which was proxied by the ratio of total commercial bank credit over GNP was proved to reduce lending rate directly and indirectly through operating cost. Despite all the favorable findings to support development of credit market for low income borrowers based on market mechanism, very high volatility of loan loss rate and ROA should be seriously noted in the management and supervision of those financial institutions focusing on lending for less creditworthy and low income customers.