FICO, once again, is not the only, or sometimes even the primary, factor used in credit decisions by lendors.
Lendors often use their own algorithms to guide in their credit decisions that are not just simple FICO based. One major CCC obtained a patent on their scoring algorithm a few years ago, and it sheds public light into their algorithm process. These are widely published, not speculation.
Many lendors rely upon a “segmentation” algorithm that is not pegged to FICO. Any basic text in credit scoring sets forth this as being a decades-old approach. Sure, FICO may also be a part of the decision, but they have also used a conventional “segmentation tree” approach within their own scorecards. First segmentation cut is usually separating those with “dirty” as opposed to “clean” credit histories. Applicants with no delinquencies or very old minor delinquencies are generally refereed to as having a clean credit history. Those with moderate (60 day) or severe (90+) delinquencies, or recent delinquencies, are generally referred to as having a dirty credit file. The severity index used to determine whether a credit history is clean or dirty is taken as a combination of severity and recentness of the delinquency.
Next cut on the segmentation tree is usually whether the credit history is thin or thick. The term “thin” and “thick” credit files are industry-recognized terms, although still at the discretion of the CCC in their algorithm.
A thin credit file usually encompasses credit histories of people that have not had a bank card or too short of a credit history. An applicant with a thin file generally refers to an applicant with fewer than three trades. Thick credit files usually means the history has data on three or more trades.
So, from the first branch of “clean” and “dirty” segmentation of applicants comes a second division into “thin” and “thick.”
Both primary factors. Then within each of the now two branches on each of the two trees comes the next factor, which is usually %util of revolving TLs. Another grouping, or segmentation, of applicants. From those branches then spring twigs that further segment into factors such as overall length of credit history.
Notice the clear parallel between the “segmentation tree” approach to credit scoring and the FICO approach. The first cut, “dirty” vs “clean,” parallels payment history in FICO, which is given the highest FICO weighting of 35%. The next cut, “thin” or “thick,” is a combination of both credit mix and credit utilization. The next cut, %util, looks almost exclusively at % util of revolving accounts, and is the prime factor used in that tree branch for then separating high and low risk applicants. At the end of each tree branch, a risk is then assigned (high, medium, low) that is highly weighted according to the major cut of %util. From the end of each branch comes an assessment of risk level. and thus approval criteria.
I do not mean to infer that lendors, such CCCs rely upon their own screening at the exclusion of FICO, but be aware that they may use their own in-house scoring. A CCC making a decision on a $200 CL card may want to cuts approval costs, and thus rely primarily upon a simple FICO score that they can readily purchase. But many use more than that. Many use their own algorithms to generate their mailings of offers for “pre-approval” of their credit offers. So, thinking that one is “pre-approved,” a consumer then applies. Then the CCC, in their final decision, may do an authorized “hard pull,” and then use your FICO as part of their final credit decision. Some even ask for residency or income info, which is clearly far beyond their evaluation using only FICO score.
Major creditors other than CCCs often have their own preliminary, or even final, scoring algorithms, and approval criteria, that are NOT based solely upon the Holy FICO! Hmm, maybe they consider income, or residency status? Family income? Total debt to income?
To think that obtaining a certain FICO score is the industry recognized standard that ensures approval or offer of a certain loan rate is simply not true.
!! I makes me feel better. I fall into the stable, but not high Fico. Thanks for giving me hope
