During a long career in banking I never lent a penny. It is not that I was a skinflint – it is just that I never worked as a lender. However, I often marvelled at those who did lend money. The general public frequently do not appreciate the level of correct decisions that a lender must make. If you look at the bad debt lines of a bank, the typical level of bad debt in normal times is a lot less than 1%. In other words, on average the lender must be repaid more than 99% of all that he lends. Consider your own decision-making – are you correct on over 99% of all your decisions? Therefore, a typical bank lender has a good number of years of experience under his belt – and his time is both scarce and expensive.

My background is primarily in statistics and I developed and introduced credit scoring for the bank that I worked with many years ago. Credit scoring was commonplace both in the US and the UK and among some finance companies offering car loans. Credit scoring is a statistical technique which is applied to lending. Whereas the lender measures the client’s ability to repay and then applies his skill and judgment to a lending decision, credit scoring assesses the statistical likelihood of repayment based upon the characteristics of the applicant by reference to the statistical experience of previous applicants. If you happen to have the characteristics associated with a higher risk of default among previous customers, then you will not get the loan.

How credit scoring works

Credit scoring works as follows. The bank selects those customers with which it has had bad experiences. A bad experience might be default or it might be that the customer’s repayment performance was such that the bank had to chase him quite considerably for repayment. The bank looks back at the characteristics of that customer at the time of application for the loan. The bank also picks out a random sample of good customers and picks out the characteristics of their applications. A statistical analysis is then done to determine what the best points system would be, which when applied to the application information would have differentiated the good from the bad.

Credit scoring provides banks with a statistical estimate of the risk of default. If the bank rejects you it is not that you are considered a likely fraudster. It is because you belong to a group sharing certain characteristics which statistically puts you in a group with a risk level higher than the bank has chosen to accept. This is similar to the message on the side of a cigarette pack saying “smokers die younger”. Many smokers live a happy life into old age, but statistically they present a higher risk of cancer.

For some banks age may be a major factor – they may have had several bad experiences with younger borrowers. For another the risk may lie with the married couple with children in expensive education. Each bank has its own risk profile.

It is interesting to see what happens when an experienced lender meets credit scoring. He expects that the credit scoring will accept/reject the same applicants which he does. However, this is not always the case. It actually works from a statistical basis. The lender, experienced as he may be, is working from a different mindset.

Over the years credit scoring has become the dominant method of assessing loan applications for small/medium loans. Another technique known as behavioural scoring is used as an early-warning system on loans after they have been booked.

For more information about credit scoring, you can contact Tom Conlon of Bankhawk Limited at This e-mail address is being protected from spambots, you need JavaScript enabled to view it .