Credit scoring and credit scores in the Australian context aren't the same as the US concept and reflect far more variables than simply the CRAA report.
Essentially, they are predictive modelling tools that take all the known data of portfolios/borrowers and look for variables that, individually or in concert with others, generate odds of a loan going into default. As a consequence (a) they change over time and (b) it isn't generally possible to say "this bit or that bit" was the specific issue.
A given number of Veda entries over a short time may be a non-issue for one borrower but "appear" to be a problem with the another when, in fact, it's the Veda data in combination with a bunch of others. By way of example, I was looking at some output the other day whereby the probability of default for a given combination of borrower variables got worse when the income went above a certain number.
And I've mentioned before, these things work.
Essentially, they are predictive modelling tools that take all the known data of portfolios/borrowers and look for variables that, individually or in concert with others, generate odds of a loan going into default. As a consequence (a) they change over time and (b) it isn't generally possible to say "this bit or that bit" was the specific issue.
A given number of Veda entries over a short time may be a non-issue for one borrower but "appear" to be a problem with the another when, in fact, it's the Veda data in combination with a bunch of others. By way of example, I was looking at some output the other day whereby the probability of default for a given combination of borrower variables got worse when the income went above a certain number.
And I've mentioned before, these things work.