Icing on the Cake: How the FICO Score and alternative data work best together

Significant attention has been paid recently to the potential for use of alternative data (i.e., consumer data not present in the traditional credit bureau files) to enhance the predictiveness and inclusiveness in credit scoring. At FICO, we have been on the frontlines of this challenge for decades, working tirelessly to identify sources of safe and reliable alternative data to enhance the risk prediction of our scores, including our industry-leading FICO® Scores, such as the UltraFICO™ Score. Through our longstanding research efforts, we've built and developed newer iterations of our scores that incorporate new categories of safe and reliable alternative data, such as telco/utility, property/public record, and cash flow data. This is especially critical for the approximately 50 million consumers in the U.S. with little or no credit information in their traditional credit bureau files, who have traditionally struggled to access credit. Due to their lack of traditional credit history, it is particularly important that these consumers have opportunities to demonstrate their readiness for credit through these alternative data sources.

While FICO may be the foremost champion of using safe and reliable alternative data, we also recognize both the opportunities and limitations of incorporating alternative data into credit risk assessment. Despite its name, alternative data is not and should not be seen as a true alternative to the data from the traditional credit bureau file that has been used in the FICO® Score to offer predictive and reliable credit assessments for decades. While there are clear benefits to both consumers and lenders from incorporating alternative data into credit risk assessment, FICO's research shows that models, such as the core FICO® Score, that are based solely on traditional credit bureau data offer a more reliable/accurate assessment of consumer credit risk than using alternative data on its own.

Fortunately, often it doesn't have to be a question of choosing one or the other: Alternative data can and should be used in models to complement the use of available traditional credit bureau data, not serve as a replacement. By using a reliable credit risk model that combines traditional credit bureau data with alternative data, lenders can gain deeper predictive insights and consumers can increase their opportunities to prove their readiness for credit. One such model that uses this combined approach is the UltraFICO™ Score, which allows consumers to contribute cash flow data from their checking and savings accounts to enhance their scores.

More than 200 million U.S. adults have sufficient credit bureau history to receive a FICO® Score; for this group, FICO® Scores derived from traditional credit bureau data alone continue to provide lenders with reliable and predictive assessments of their credit risk. Even for this population, scores, such as the UltraFICO™ Score, which combines traditional credit bureau data with safe and reliable alternative data can improve predictiveness.

Our analysis of a subset of this population - the thin file/new-to-credit prime segment1 - shows that while cash flow data is in fact predictive of their credit risk in standalone fashion, FICO® Score 9, which relies entirely on traditional credit bureau data, is still approximately 10% stronger than a score relying on cash flow data alone. Yet despite the core FICO® Score's high level of predictiveness and accuracy, a score that combines the use of cash flow data and/or other alternative data sources with traditional bureau data represents a "best of all worlds" approach: For this segment of the population, a score derived from both traditional credit bureau data and cash flow data is approximately 5% more predictive than a model using bureau data alone.

This approach - combining traditional credit bureau data with alternative data sources - reflects FICO's commitment to continued innovation in service of financial inclusion and further expanding credit access. And to help ensure these new products are able to truly make a difference for consumers, we recently launched our Inclusion Accelerator Program and the Financial Inclusion Lab, which allows for testing of new credit decisioning tools, and further lender adoption of alternative data solutions like the UltraFICO™ Score.

Alternative data holds incredible potential to provide millions of consumers with an onramp to the mainstream credit system. By adding these new sources of consumer data to the proven and tested FICO® Score, lenders can make better informed decisions and consumers can receive the opportunities they need to achieve their financial dreams. FICO is committed to remaining a leader in identifying and incorporating safe and reliable alternative data into its credit scores to drive financial inclusion and expand credit access for more consumers.

[1] Thin/new-to-credit prime segment is defined on the traditional credit bureau data as those consumers with no serious delinquencies or collections or public records and with less than 4 accounts or less than 36 months of credit history.