Use this URL to cite or link to this record in EThOS:
Title: Credit risk in the context of European integration : assessing the possibility of Pan-European scoring
Author: Andreeva, Galina
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2004
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Credit scoring is the collection of techniques used for risk assessment in consumer credit. Traditionally a credit scoring model is constructed to fit a specific credit portfolio when normally consists of residents of one country (customised models). But the political desire for further integration of the European Union into a single internal market opens the possibility for the lenders to compete across the national borders. Therefore the necessity arises to assess the risk of mixed heterogeneous population consisting of residents of several European countries. This thesis shows how a single generic model can be used to credit score the applicants for a revolving store card from three different European countries. First, the EU harmonisation process is reviewed with the aim to establish its likely impact on the credit scoring practice. In particular, the legal restrictions on the information used in credit scoring models are examined and the effect of such restrictions for both lenders and borrowers is investigated. The comparison of credit regulations is provided between the USA and the EU, and for the latter the differences in the national legislation of the EU member states are presented. Second, several generic models are developed using logistic regression and survival analysis and their predictive accuracy is benchmarked against the performance of equivalent national (customised) models. Whilst logistic regression is the most established approach in credit industry, survival analysis is a relatively new application that offers an advantage of predicting time to the event of interest and therefore, lays the foundation for estimating the applicant’s profitability. Predicting profitability requires estimates of both the probability of default and the likely usage of the store card. Whereas modelling default is the traditional task of credit scoring, estimation of usage is far less common. Time to the second purchase is considered as the measure of the card usage and dependencies in application and behavioural data are examined that can be used for predicting the customer’s future behaviour. Generic models are found to perform well across three countries under different modelling approaches and in different applications. In predicting default they are competitive with the national models, whilst in other applications generic models demonstrate marginally inferior results. However, harmonisation of the data available for the analysis is likely to further enhance the predictive power of generic models and expand the possible scope of their application.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available