What's in a Credit Score? More than You Might Think, and It's Constantly Changing

Editor’s note:  DataQuick®, a provider ofrnadvanced real estate information solutions recently announced that GordonrnCrawford, Ph.D. had joined their firm as vice president of Analytics.  MND doesn’t normally note such appointments,rnbut it caught our eye that Crawford who until recently was vicernpresident of Mortgage Research at Fannie Mae, isrnan expert on credit scoring performance models and spoke about thatrntopic at a recent Mortgage Bankers Association Risk Conference.  He was kind enough speak with us on the topicrnwhile preparing for that meeting.</p

While a lot of us can now actually singrnthe words, the original screwball television commercials about credit scores markedrnthe first time many people ever heard the words.  Dr. Crawford told us that until recently creditrnscoring was sort of a “black box” for consumers.   Comparedrnto credit reports, scoring is a relatively young credit tool, and even after consumersrnlearned they had a score, many had no real idea of what that meant. </p

In the last few years FICO, the mainrnpurveyor of credit scoring technology, and other providers have worked torneducate borrowers, providing information about what goes into credit scores andrnconsumer guides on how to improve them.  Anotherrnrecent consumer innovation is the ability to do credit scorernsimulations online.  </p

Using one of these simulators a consumerrncan, for example, test the effect of paying off one debt entirely in contrastrnto paying a portion of another or whether opening a new line of credit willrnhave a positive or negative effect. rnThis, Crawford said, is a double edged sword.  Credit scoring companies now know that thesernsimulators have given consumers the ability to game the system and must takernthis into account whenever they make changes to their models.  “People should know what is in their score andrnhow to improve it,” he said, “but being able to manipulate it can invalidaternthe scorer’s model.”  </p

In two respects credit scoring is not arnstatic science.  From the consumerrnstandpoint, Crawford said that what were considered acceptable credit scores beforernthen started to decline in 2002.  A scorernof 700 used to be considered good, then 660; then suddenly it was 580.  That trend flipped in 2008 and since thenrnscores have ramped up to what is now a 760 to 780 average for what isrnconsidered a prime score.  “We now have arnstrange situation,” he said, “with very low interest rates but very tightrnaccess to credit.  What used to bernconsidered a good score, 720, is now one where lenders have to proposernalternative programs.  We have swung fromrna standard that was way much low to one that is probably too high.”</p

From a lender standpoint credit scoringrndoesn’t stand still either.  Crawfordrnsaid that vendors frequently update scoring models and the industry isrnchallenged by this because it is not easy to adapt.  All credit scoring is predicated on a FICOrnmodel and changing one necessitates changing or at least tweaking others.  Existing pricing and underwriting models arernalso already calibrated to old scoring, making it costly to put new ones inrnplace.  Consequently lenders want to see substantivernchanges or ones that provide measurable benefits before they undertake the troublernand expense of training their people and changing their systems in order to accommodaternthem. </p

A case in point is a new model recentlyrnintroduced by FICO which introduces data from a new CoreLogic credit reportrninto the scoring and thus gives weight to such information, according to FICO “asrnproperty transaction data, landlord/tenant data, borrower-specific public data,rnand other alternative credit data.  FICOrnmaintains that the new model raises the scores of many borrowers – about 3rnpercent more individuals would score above the current 715 median – and has arnpredictive value of risk performance 7-1/2 times greater than current models.</p

Crawford said “Newer measures to beef uprnthinner files need to be proven out.  Ifrnone lender adopts a new model and others don’t follow then it is hard for himrnto transfer a loan or its servicing, sell the loan on the secondary market, orrncommunicate about that loans and its risk. rnHe risks losing common measures with other lenders.”</p

Besides, Crawford said, lenders arernhappy right now with the tighter underwriting standards.  Behind their concerns about any loosening of themrnis an uncertainty both about the capital levels that will be required forrnholding mortgages and about the future of Fannie Mae and Freddie Mac.  “Until they have a higher comfort level theyrnwon’t be willing to accept more exposure to risk.”  </p

He said that lenders, of course, userncredit scoring as one in a whole arsenal of tools when they evaluate risk;rnappraisals, owner occupancy status, and other factors all enter into thernequation.  In addition, Crawford said,rncredit scores are intended to predict success with a loan, not necessarily tornpredict success with a mortgage loan.  Forrnthis reason many lenders beef up scores by taking specific fields from them -rnfor example the utilization rate of revolving debt – and wrapping them intorntheir underwriting, thus multiplying the weight they carry.</p

Crawford’s role at DataQuick will involverndeveloping all models and analytics, including home price indices, automatedrnproperty valuation models and loan performance models.  While at Fannie Mae he worked on performancernmodels used to determine loss allowance measures and pricing.  Crawford earnedrna Bachelor of Arts in Economics from Brigham Young University and both a Masterrnof Arts and a doctorate in Economics from the University of Rochester.  

All Content Copyright © 2003 – 2009 Brown House Media, Inc. All Rights Reserved.nReproduction in any form without permission of MortgageNewsDaily.com is prohibited.