»
VISIT US ONLINE @ DSNEWS.COM
COVER STORY
INDUSTRY INSIGHT
Devising a Robust QC Plan
INSIDE THE BELTWAY
Ensuring compliance with GSE, FHA,
CFPB, and other investor and regulatory
requirements necessitates a robust quality
control program. The extent of these
regulations and requirements, and the
increased consequences of failing to meet
the requirements, has elevated servicing
quality control to a new level of importance.
The process starts with the development
of a servicing quality control plan. The plan will
document the scope, frequency, and sample size
of the quality control reviews. The scope of the
review should include, at a minimum:
» Default servicing, including loss mitigation
efforts
» Foreclosure processing
» Boarding of new loans and servicing transfers
» Customer service
»Fees
» Escrow administration, including compliance
with RESPA
» ARM adjustments
» Handling of prepayments and payoffs
The plan should detail what quality control
procedures will be performed in each of the
broad categories listed above. These procedures
should include testing of all of the GSE, investor,
and CFPB requirements that are applicable.
Quality control testing should generally be
conducted monthly, although smaller servicers
may consider quarterly reviews. FHA specifically
requires monthly reviews of the servicing of
delinquent loans.
Sample sizes are generally left to management's discretion, although again FHA is
specific, requiring reviews of 10 percent of
delinquent loans on a monthly basis and a review
of at least 10 percent of all loans quarterly. FHA
allows servicers with more than 3,500 loans to
use statistical sampling.
FHA's sampling requirements are thought to
represent the high end of a range of appropriate
sample sizes that will depend on the level
of controls already in place. If you review 10
percent of delinquent loans each month and the
population of delinquent loans is static (i.e. chronic
delinquent accounts), every delinquent loan will
be reviewed in less than 12 months, which makes
for a very large sample over the course of a year.
Management should develop sample sizes that
57
