Appendix D - Credit Scoring
Credit scoring is a sophisticated statistical tool utilised by most banks in their assessment and approval (or decline) of the vast majority of their personal loan business. Credit scores are derived from specially designed loan application forms and assessed against statistically approved templates derived from a detailed analysis of past lending experience. The aim is to identify characteristics that are powerful predictors of the risk of default. As such, credit scoring depends on large amounts of data; the greater the level of data and experience to draw upon, the less the statistical error involved in predictions.
While credit scoring has been successfully applied to the personal loan market, it is not currently used in the context of business (including SME) lending. Comments made by the banks indicate a certain amount of scepticism that credit scoring in the SME market can be successfully applied in New Zealand. There are a number of reasons for this:
- There is an awareness of the inherent danger in attempting to apply credit scoring to business lending. The concern is that good loans may be turned down by trying to "fit borrowers into boxes", especially given the diversity in operations, goals and behaviour of businesses.
- The banks have indicated an increased emphasis on relationship management. Credit scoring is not necessarily seen as something that would, or should, replace the judgment exercised by the relationship/branch manager.
- To work well, credit scoring depends on having a considerable amount of information upon which to develop measures that are reliable indicators of default risk. The size of the New Zealand market means that there may be issue around whether or not there is sufficient information to populate credit scoring models.
- The widespread use of collateral was not seen by the banks as a factor influencing their views on the use, or otherwise, of credit scoring.
Notwithstanding the cautionary remarks offered in relation to credit scoring, certain banks are currently testing credit scoring models with a view to their application to small commercial loans. It remains to be seen whether these models will be applied in practice.
We note anecdotal evidence from the United Kingdom where credit scoring is increasingly being used for loans up to ₤100,000. For small loans, credit scoring may be used as the primary decision tool but that as loan size increases toward ₤100,000, it is more likely that credit scoring will be used as a decision aid.43
It is worth noting that there is some international evidence that credit scoring can assist in overcoming the inherent benefit/cost trade-off that banks face when deciding whether or not to invest in obtaining information regarding a potential borrower. For example, one study44 estimated that the cost of evaluating micro loan applications in the US using credit scoring was reduced to around $100 compared to a range of $500-$1800 prior to the introduction of credit scoring. The time saving involved meant that banks could focus more time on marginal applications, existing loans that are showing signs of distress and processing more loan applications.
There is also some evidence that use of credit scoring means that the marginal benefits of taking and maintaining collateral are not justified for small loans. The Bank of England reports that there is some evidence of banks being more willing to lend on an unsecured basis when using credit scoring which potentially improves the access to bank finance for very small and start-up SMEs.45 Specifically, the Bank of England noted that Barclays estimated that 85% of borrowing applications from its small business customers are determined irrespective of whether security is available.46
Back to Top