SAI Global’s general manager, product and marketing, Benjamin Balk, said the root of the problem is that banks have to manually source and interpret a mass of complex data to assess whether a borrower could repay a loan.
Mr Balk said this typically involves credit teams poring over hundreds of pages of government extracts, performing separate ASIC, PPSR and bankruptcy checks, as well as other searches.
“Decisions about whether or not to approve credit applications are often based on these findings, which, despite best efforts, can be highly subjective, prone to error and overlook vital information tucked away in difficult-to-read reports, particularly in relation to commercial lending," he said.
“The fallout from this is that loans can be granted without the lender seeing the full picture.
“For financial institutions, this can mean accruing unnecessary bad debt and lost opportunities, as a result of time wasted on prolonged data reviews.”
Mr Balk told Mortgage Business that there is a common misconception that commercial lending assessments are highly automated.
He said that while the less complex residential mortgage sector is highly automated, this is harder to do with commercial finance, which can involve complex webs of organisational structures, directors, security interests and lending relationships.
One of the consequences of relying on manual assessments is that different organisations will often draw different conclusions about the same information, according to Mr Balk.
The potential scale of the problem can be seen by comparing the commercial sector with the less complicated retail sector, which also suffers from a subjectivity problem.
APRA chairman Wayne Byres revealed last month that the prudential regulator had encountered this problem during a recent borrower survey in which lenders were asked to provide serviceability assessments for four hypothetical borrowers.
“Mortgage lending is often thought of as a fairly commoditised product, but in reality there are wide differences in how lenders assessed the risk of a given borrower,” Mr Byres said.
“The first surprising result from our review was the very wide range of loan amounts that, hypothetically, were offered to our borrowers.
“It was not uncommon to find the most generous ADI was prepared to lend in the order of 50 per cent more than the most conservative ADI.”