A computer software company that develops credit scoring models, generating credit scores used by a majority of lenders called FICO credit scores.
FICO is a way to consolidate the complexity of credit-related financial transactions that any one participant in the credit economy might encounter, in such a way as to present a quick summary of whether those many (or few) interactions have been
1) very positive over a long period of time, giving a lender comfort the borrower is a good risk
2) are average
3) are below average.
Based on this initial response, lenders are able to efficiently and inexpensively make initial credit determinations, primarily for new customers, which ultimately leads to a lower cost of operation for the financial institution. These are often not final determinations, but it can be final if the FICO is far above average, or far below average.
@Anonymous wrote:A place with computers working, not Humans with Brain..
Humans (with brains) use computers and their increased power to develop better credit scoring which has gradually improved over time.
Here's a fairly thorough document on using "R" (a powerful statistica analytic package) to create credit scoring algorithms. FICO is, of course, proprietary but this document provides info on techniques used worldwide for developing credit scoring as well as various metrics the industry uses to evaluate how good they are. Banks are always in search of better algorithms because it increases their ability to lend money to more people at the same risk. As models improve more people with poor scores move up than people with good scores move down. This is because most people with below average scores will repay a debt and imroved models identify more of these folks.
http://cran.r-project.org/doc/contrib/Sharma-CreditScoring.pdf
