About VantageScore

The nation's three major credit reporting companies (CRCs) -- Equifax, Experian and TransUnion -- collaborated to develop VantageScore in response to credit grantors desire for a model that could more reliably and predictively, score more people. The development team designed a model that accomplishes these goals:

• A highly predictive, generic consumer credit risk model
• Improved risk assessment
• For the first time, a consistent algorithm across the three CRCs
• Expanded the universe of people who can be scored
• Scaled logically for ease of understanding

VantageScore at a Glance

• Predicts the likelihood of future serious delinquencies (90 days late or greater) on any type of account
• Returns a score range of 501-990 (higher scores represent a lower likelihood of risk)
• A consumer score is based primarily on a 24-month review of a consumer's credit file
• Includes up to four score factor codes and a fifth FACTA reason code (Spanish version available)
• Can be accessed from all three major credit reporting companies
• Ignores "authorized-user" tradelines to assess credit risk

VantageScore Development

VantageScore was developed during 2005 from a national sample of approximately 15 million anonymous consumer credit files - five million files from each CRC. This wide-ranging sample selection resulted in a model that reflects the diversity of the country. The credit information included public record information, tradeline data and inquiries.

• 15 million anonymous consumer credit files
  • Five million files per CRC
  • Completely random sample
  • 50% of the files were used for development; the remainder were used for validation after development
• Credit Files used in development reflect recent shifts in consumer behavior
  • Files were selected from the June 2003 through June 2005 time frame and used as a performance window
  • Two populations were identified from these files:
    • "Existing" - Files with tradelines opened prior to June 2004
    • "New" - Files with tradelines opened between July - September 2003

Characteristic Leveling

The development team's earliest task was to address disparities in the CRCs' individual definitions of credit characteristics, a foundation of consumer credit risk assessment. Credit characteristics are defined as consumers' payment and credit management behavioral data supplied by the lenders to the CRCs.

Inconsistent definitions in characteristics across the three CRCs result in different risk perspectives when decisioning. This is the case for many consumers whose credit file is housed by more than one of the CRCs. Through a process known as characteristic leveling, consistent and equitable characteristic definitions were established across multiple sources of data so that the data is interpreted in the same manner when present at multiple sources.

For credit grantors, characteristic leveling increases confidence in consistent credit decisions regardless of the data source. For consumers, it means a comparable risk assessment regardless from which CRC their financial institution selects to obtain the consumer's VantageScore. While there will still be score differences among the three CRCs, the variance is minimized because the difference is due to data differences in each credit file at each of the CRCs rather than inconsistencies in the characteristic definitions. Learn more about VantageScore's characteristic leveling efforts and resulting benefits in our Newsroom.

Segmentation Methodology

VantageScore was developed using a combination of key attribute and score-based segmentation methodology, resulting in 12 scorecards, including previous bankruptcy, thin file and thick file. There are several beneficial characteristics that are a part of this approach:

• Scores developed specifically for bankruptcy and thin file populations
• Score-based segmentation facilitates population stability and reduced score volatility
• Ability to identify the "best of the worst" in each segment
• Better identification of risk segments
• Segmentation score captures multiple elements of risk behavior

Learn more about VantageScore's patent-pending segmentation approach by reviewing our white paper Segmentation for Credit Based Delinquency Models.

Score Range

The numerical range used by VantageScore is 501-990, with higher scores representing a lower likelihood of risk.