Five Ways to Make a Credit Score Model Work for You
Guest post written by Barrett Burns, President and CEO, VantageScore Solutions, LLC.
VantageScore Solutions, LLC is the NAFCU Services Preferred Partner for Credit Scoring.
Times have changed since a promise and handshake were all you needed to get a loan. Now credit scores speak to your character. Most credit unions primarily rely on credit scores to help make consumer lending decisions. Credit scoring models incorporate credit scores with other characteristics related to creditworthiness. In today’s market, there are dozens of different credit scoring models available, from generic models such as the VantageScore 3.0 model, to customized models that are generally expensive to build and maintain.
Even so, it’s a common misconception to think of credit scores as a commodity, or a “one-size-fits-all” risk management tool. A credit score is the numerical representation of the likelihood that a consumer within a specific population will become 90 days or more past due on a debt obligation in a two-year timeframe. It’s important to remember that this propensity to default is assessed within the context of the population being scored. The most effective credit scoring models incorporate other relevant information, such as current economic factors, over a greater population. Choosing the right model for your credit union can help you in ways you might not expect, from saving time and expense to improving accuracy and applicant pools.
Here are five features your credit scoring model should offer:
1) Expand your universe of applicants
According to Experian, 64 million U.S. adults have limited or no credit history. That’s an incredible opportunity for credit unions! Based on research commissioned by VantageScore Solutions and conducted by SourceMedia Research, almost 60% of lenders surveyed either never or rarely respond to loan applications from consumers without credit scores.
These consumers could become your members. A model that can provide scores to tens of millions of otherwise “unscoreable” consumers means they can be underwritten using your standard automated procedures instead of a manual underwriting approach, which can be costly.
2) Take advantage of new data
It’s intuitive, but it’s a point worth driving home: A model built on post-Recession data will be more reflective of current conditions and more predictive of future credit behaviors. The credit market is entirely different than it was just a few years ago.
Indeed, developing a model over the extended window reduces the model’s sensitivity to behavioral volatility so be sure the model you are using has this advantage. For example, the data sample used to build the VantageScore 3.0 model was developed on consumer behavior from two different two-year timeframes: 2009–2011 and 2010–2012. Each performance timeframe contributed 50 percent of the model’s development. Developing the VantageScore 3.0 model over the extended window reduces the model’s sensitivity to consumer behavioral shifts over different volatile periods.
3) Help your members understand risk-based pricing
As you are no doubt aware, credit unions are required to provide notices to any member who is denied a loan or offered credit terms other than the best available. As part of these notices, four or five “reason codes” must be also be provided to help explain how certain items in the member’s credit file impacted their credit score.
If your credit union uses the VantageScore model, take advantage of our new website, ReasonCode.org. We built a search engine that lets members type in reason codes to find more detailed explanations of factors that lower their credit scores, and information about how they can improve their scores.
4) Take the guesswork out of inconsistent scores
It can be frustrating for you as a lender, and equally if not more frustrating for the member, when a member’s credit score fluctuates when obtained from more than one credit reporting company (CRC), and there are no errors or other discernible differences in their credit files. The VantageScore model is the only credit scoring model that avoids score deviations inherent in the use of models built specifically for each CRC database. The VantageScore model is identical for all three major CRCs, and it uses a patented process called characteristic leveling to help provide more consistent scores across all three CRCs. If the member’s credit file is the same across all three CRCs, then the VantageScore credit score will be the same.
This translates into more consistent scores across Equifax, Experian and TransUnion, and more confidence that the credit score you’re looking at reflects a member’s true risk profile.
5) Predictiveness – where the rubber meets the road
None of items 1 through 4 means a hill of beans unless the credit score you pull is highly predictive. Across the entire population, in 99% of tests, the VantageScore 3.0 model improves upon the already strong performance of the VantageScore 2.0 model and outperforms all other benchmark models provided by the three credit reporting companies (Equifax, Experian, and TransUnion). Importantly, the model achieves up to 25% lift among Prime and near-Prime consumers, the population of borrowers that is often most desired by mainstream lenders, where lenders traditionally focus their credit score cut-offs.
On May 22, 2013 VantageScore Solutions will host an online webinar to provide lenders, regulators, and media an in-depth explanation of our newly introduced credit scoring model, VantageScore 3.0.
Register here.
VantageScore Solutions, LLC is the NAFCU Services Preferred Partner for Credit Scoring.
For contact info and more educational resources, visit: www.nafcu.org/vantagescore.