“I’ve seen more customers go to 100% in last six to eight months” when it comes to pre-close audits, said Riddell, whose company provides automation that assists with, among other things, data integrity reviews aimed at managing risks that could affect loan quality.
When asked what the catalyst for the shift has been, he said pressure to compete with peers who have increased their sample sizes has been part of it.
While there is a trend toward more thorough reviews, Riddell said it is not something all players are doing.
“Many don’t look at enough of their loans,” he said, noting that some are still using sample sizes as low as 10-15%, which “if not well orchestrated” may mean some risks will be overlooked.
“Our recommendation probably north of 20% prior to close examinations,” Riddell said. “I think is healthy. I think if there is less than that, there is some exposure.”
When asked if 100% reviews are necessary, he said, “The more, the better.”
If the pre-close/quality assurance review process is strong, then quality control, or post-close reviews can be done with sample sizes that should be no less than 10%, he said, noting that “more and more” sampling in this area tends to be even higher, in the 15% to 20% range.
While to an extend post-close reviews can be likened to trying shut the doors after the horse has already gotten out of the barn, making pre-close reviews more important, post-close reviews still are necessary even if the pre-close review is strong. “Life happens to consumers after closing,” Riddell said. In the manufacturing of loans, one needs to ensure there is creditworthiness both at the pre and post-closing stage, he said.
One challenge in doing these both types of reviews is that they need to “impartially validate production without negatively impacting consumers’ experience,” he said.
“It takes a little time to get procedures in place,” Riddell said. “Provided expectations are set early on with customers, you can do this with a positive approach that works for everyone’s benefit.”
Whatever means one uses to conduct reviews, he advises being sure that personally identifiable information is secure.
“The PII factors are extraordinary,” Riddell said. “What is found on a 1003 is a life story.”
He said automation that provides controlled access to documents and protects personally identifiable information is one way to tackle the issue, although he warns that developing this from scratch involves considerable expenditure. “Not everyone is willing to make this investment,” he said.
In managing loan quality risks, he suggests a focus “first and foremost” on data integrity throughout the loan process. Riddell said this should include a review and record of the numerous “pit stops” along the ways, as well as who is involved in the process, what information they receive and what they do with it.