Consumption credit performs an extremely crucial function in aiding consumption and allows customers to smooth consumption. On top of that, several merchants independently give various forms of credit choices. In spite of this, lending is related with risks and it is for that reason necessary to be able to accurately forecast credit defaults. This document looks into exactly what elements are crucial to look out for when making credit default estimations by calculating a probit regression model using 170.000 approved consumption credits. Even though many conventional rating techniques largely evaluate financial and demographic variables this report illustrates that behavioural variables are at least as significant when generating default predictions…..
Contents
1 Introduction
2 Theoretical Framework and Previous Research
2.1 Credit Risk Management
2.2 Credit Scoring
2.2.1 General Purpose
2.2.2 Regulatory requirements: Basel II
2.2.3 Application and development of scoring models
2.3 Credit Scoring Method
2.3.1 General
2.3.2 Review of credit scoring methods in use
2.3.3 Our regression
2.4 Framework for analysis
3 Data
3.1 Origin
3.1.1 General
3.1.2 Credit process
3.1.3 Complementary data
4 Hypotheses
4.1 Direct financial ability19
4.2 Indirect financial ability
4.3 Moral hazard
5 Methodology
5.1 Methodology
5.1.1 Econometric Model
5.1.2 Regressions
5.1.3 Definition of default
5.1.4 Natural logarithm of stochastic variables
5.1.5 Deriving demographic data
5.1.6 Treatment of missing variables
5.1.7 Multicollinearity
6 Empirical Findings
6.1 Regressions
6.1.1 First regression
6.1.2 Second regression
6.1.3 Third regression
6.2 Discussion
6.2.1 Direct financial ability
6.2.2 Financial ability
6.2.3 Moral hazard
6.3 Limitations
6.3.1 Sample selection bias
6.3.2 Evaluation of the model
6.3.3 Lack of information on profitability
6.3.4 Different models for different applicants
7 Conclusion
References…………..
Source: Stockholm School of Economics