It is often difficult, if not impossible, to understand how a CCC evaluates the grant of an application, for they usually guard their process through trade secret laws.
But one major CCC, FirstUSA, for reasons known only to them, decided to deviate from trade secret practice, and secured a patent back in 2001 on their credit card approval algorithms. A view into the mysterious world of consumer “bucketing” was thus revealed, in much detail.. Securing a patent requires the full disclosure of the process, and they chose to do this. So it is now public domain.
The patent number is 6,202,053, and can be secured free on the Patent and Trademark website (uspto.gov).
But for those of you who simply want the quick and dirty, here is my summary of the patented process disclosed by FirstUSA to evaluate applicants for a CC.
They ultimately divide the total population of applicants down into eight different scorecards/scoring algorithms.
Their first cut is to weed out applicants that have no current bankcard account, and do not have at least three existing trade lines (TLs). Those who do not make the intitial cut by not having a current bankcard, or if they do, not also having at least three current TLs, are put in scoring algorithm 8. This is apparently the kiss of non-approval. Low history, high risk.
For those remaining applicants who have at least three TLs and at least one current bankcard, these applicants are then subdivided into two primary scoring categories, based on prior payment history (this parallels the FICO emphasis on payment history).
Category one applicants have had no recent delinquencies, or only minor (30-day) prior delinquency. Category two applicants have either had a recent delinquency, or any prior moderate/serious delinquency (60 or 60+ days late, or a BK, PR, CO, or CA).
Category one applicants are their priority, and are then first broken down into four different algorithms for evaluation based on their current revolving credit utilization.
Category one applicants with current revolv %util of 25% or less are run through scoring algorithm 1, and are prime for approval based on low util, and thus low risk.
Category one applicants with current revolv %util of 25% to 70% are put into scoring algorithm 2, which looks at more factors.
Category one applicants with a current revolv %util of 70% or higher are put through a more demanding risk review, with closer scrutiny, and are subject to two different scoring algorithms, based upon the time of accounts on file. Those with a credit history of more than ten years are assigned to scoring algorithm 3, and those with ten years or less account history are assigned to scoring algorithm 4.
Category two applicants are those who have either a recent delinquency, or have had a moderate or severe delinquency in the past. Three different algorithms score these applicants.
The first cut on these applicants is %util of existing revolving debt. Those with a current %util of 40% or less are scored under scoring algorithm 5.
Those with a current %util of over 40% are them divided into two categories, based on credit history. Those with a credit history of less than ten years are assigned to credit scoring algorithm 6, and those with a credit history of more than ten years are assigned to scoring algorithm 7.
Message Edited by RobertEG on 10-26-2008 04:41 AM
