Probability models enable Alliance & Leicester to identify potential loan defaulters on its books
Alliance & Leicester is to use data analytics software to improve credit risk processes within the bank.
The financial services firm’s credit risk management division will use software to analyse the performance of personal loan products and to identify trends associated with customers defaulting on repayments.
The division is using analytics software from SAS at the moment, but will spend six months rewriting standard reporting methods and training staff before introducing new software from IT firm SPSS.
‘We will be using the software for analysis of our portfolios and for financial modelling,’ said Simon Baum, director of credit and risk at Alliance & Leicester. ‘With our personal loan portfolio we do analysis, so if we see any strange trends we will look at it.’
The SPSS analysis package will use databases at the bank’s head office in Leicester, allowing credit risk managers to carry out ad hoc analysis and produce weekly and monthly reports.
Analysts will also be able to drill down into the research to see how different financial products are performing and compare them on a geographical
basis. Other socio-demographic data, such as home ownership status and customer debt levels, can also be examined.
Alliance & Leicester says it can also use historical information held on its databases to create risk models.
The bank will be able to identify characteristics of previous high-risk customers that have seriously defaulted on loan repayments, and create a model that could be used when deciding whether to approve new applications, says Baum.
‘We build probability models to gauge the risk levels of whether someone will default. Once we build the model we can then add new data to see if changes are needed,’ he said.
Other parts of the bank will continue to use SAS software for business analysis projects, but Baum says the credit risk management division has switched to SPSS because the software offers improved pricing and flexibility.
Alliance & Leicester is to use data analytics software to improve credit risk processes within the bank.
The financial services firm’s credit risk management division will use software to analyse the performance of personal loan products and to identify trends associated with customers defaulting on repayments.
The division is using analytics software from SAS at the moment, but will spend six months rewriting standard reporting methods and training staff before introducing new software from IT firm SPSS.
‘We will be using the software for analysis of our portfolios and for financial modelling,’ said Simon Baum, director of credit and risk at Alliance & Leicester. ‘With our personal loan portfolio we do analysis, so if we see any strange trends we will look at it.’
The SPSS analysis package will use databases at the bank’s head office in Leicester, allowing credit risk managers to carry out ad hoc analysis and produce weekly and monthly reports.
Analysts will also be able to drill down into the research to see how different financial products are performing and compare them on a geographical
basis. Other socio-demographic data, such as home ownership status and customer debt levels, can also be examined.
Alliance & Leicester says it can also use historical information held on its databases to create risk models.
The bank will be able to identify characteristics of previous high-risk customers that have seriously defaulted on loan repayments, and create a model that could be used when deciding whether to approve new applications, says Baum.
‘We build probability models to gauge the risk levels of whether someone will default. Once we build the model we can then add new data to see if changes are needed,’ he said.
Other parts of the bank will continue to use SAS software for business analysis projects, but Baum says the credit risk management division has switched to SPSS because the software offers improved pricing and flexibility.
Share this page
Email this page
Back to Top
Home
Edit Post
0 comments: