Consumer credit defaults in Russia: Credit scoring
Most of existing scoring systems are based on binary choice models with sample selection. This setting does not allow for up-to-date information about loans to be used and a lot of observations becomes lost. In the paper a model of binary choice with sample selection is extended to the case of many periods. This extension allows for defaults to be modeled for each period that solves the problem of lost observations. This setting also can be used to estimate the effectiveness of existing scoring system of a bank. The model is estimated using data granted by one of commercial banks of Nizhny Novgorod. Sample consists of observations from January 2009 to March 2012.
This conference proceeding includes selected full papers from the 11th EBES Conference – Ekaterinburg. We have accepted papers among resubmitted full papers after the conference ended. In this proceeding you will find a snapshot of topics that are presented in the conference. As expected, our conference has been an intellectual hub for academic discussion for our colleagues in the areas of economics, finance, and business. Participants found an excellent opportunity for presenting new research, exchanging information and discussing current issues. We believe that this conference proceeding and our future conferences will improve further the development of knowledge in our fields.
The paper presents the structural model of decision-making process on the residential mortgage market. We empirically estimates key drivers of mortgage borrowing, underwriting, and default process by jointly using market-level monthly data and loan-level data from regional branch of Agency of Home Mortgage Lending (AHML). The multistep estimation procedure allows correcting for sample selection bias and endogeneity and provides consistent parameter estimates. Obtained results shows that risk preferences are changing during the time and AHML borrowers are relatively high risky.
The given paper proposes a method to assess credit worthiness based on a continuous scale. This method in contrast to current methods that rely on a binary scale such as bad/good credit uses aggregated randomized indices. Its application may have certain practical benifits in real life, e.g. assessing the individual loan price of a particular person rather than setting a standard loan price for clients. The credit scoring model is based on set of private borrowers information that can be converted into quality function corresponding to a weighting coefficient. These weighting coefficients and quality functions then can be used to compute the quality of credit score on a continuous scale. Results based on data from German credit base have showed the feasibility of the approach. It was found that results of credit scoring with a different scales can not be correctly compared by probability of well classified borrowers.
The mortgage crisis that started in the U.S. in 2007 and lasted until 2009 was characterized by an unusually large number of defaults on the subprime mortgage market. As a result, it developed into a global economic recession and placed the stability of the world banking system in jeopardy. Therefore, the issues of credit risk modeling showed the shortcomings of the current credit risk practice. Truncation, or partial observability, and simultaneous equations bias causes sample selection bias. As a result, parameter estimates are biased and inconsistent. Firstly, we provide an overview of current approaches in the mortgage literature to control for the sample selection bias correction, such as the Heckman model and bivariate probit model with selection. Secondly, a review of the most significant mortgage studies discussing this problem is introduced. Specifically, different structural models, specific datasets and empirical results are regarded. In addition, we discuss such key credit risk determinants as borrower characteristics, terms of the mortgage contract, mortgage characteristics, and macroeconomic conditions. Finally, we conclude the discussion with possible research questions.
This paper analysis the application of artificial neural networks to tackle the problem of credit risk evaluation. They help to classify borrowers according its credit risk and to develop effective credit scorings systems on consumer credit market. We provide analysis related works, which based on using neural networks to credit risk modeling. The article emphasizes key advantages and disadvantages of artificial neural networks and to present ideas for future studies.
This paper analyzes the problems of credit risk modeling on the Russian residential mortgage market. The structural model of the credit risk evaluation, which controls for the sample selection bias and endogeneity, is presented. It estimates based on the regional mortgage data. Obtained results can be used to develop the effective risk management systems in credit organizations.