Please rotate your device to landscape mode to view the charts.

On the Rise of FinTechs: Credit Scoring Using Digital Footprints

Journal: Review of Financial Studies

Date: 20200701

Author: Berg, Tobias; Burg, Valentin; Gombovic, Ana; Puri, Manju

Abstract:
We analyze the information content of a digital footprint--that is, information that users leave online simply by accessing or registering on a Web site--for predicting consumer default. We show that even simple, easily accessible variables from a digital footprint match the information content of credit bureau scores. A digital footprint complements rather than substitutes for credit bureau information and affects access to credit and reduces default rates. We discuss the implications for financial intermediaries' business models, access to credit for the unbanked, and the behavior of consumers, firms, and regulators in the digital sphere.

Link: Google Scholar

Background and Context

Study Scope

Analysis of 270,399 purchases from a German E-commerce company between October 2015-December 2016 to examine how digital footprints predict consumer default.

Digital Footprint Definition

Information that users leave online simply by accessing or registering on a website, including device type, operating system, email provider, and browsing behavior.

Methodology

Comparison of default prediction accuracy between credit bureau scores and digital footprints using area under curve (AUC) analysis.

Digital Footprint Variables Match Credit Bureau Score Performance

  • Digital footprint alone achieves 69.6% prediction accuracy, exceeding credit bureau score accuracy of 68.3%
  • Combining both sources improves accuracy to 73.6%, showing complementary value
  • Demonstrates digital footprints can be as effective as traditional credit scores

Default Rates by Device Type Show Clear Risk Patterns

  • Mobile users have 3x higher default rate (2.14%) compared to desktop users (0.74%)
  • Device type serves as proxy for customer income and financial stability
  • Shows how simple digital variables can predict creditworthiness

Email Provider Choice Indicates Default Risk

  • Premium email provider (T-Online) users have lowest default rate of 0.51%
  • Free email service (Yahoo) users have highest default rate of 1.96%
  • Email provider choice reflects customer's economic status

Impact of Digital Footprint on Default Rates Over Time

  • Default rates dropped by 53% after implementing digital footprint screening
  • Shows practical effectiveness of digital footprint in risk assessment
  • Demonstrates significant business impact of using digital information

Digital Footprint Effectiveness Across Customer Segments

  • Digital footprint works better for unscorable customers (72.2% AUC) than scorable ones (69.6% AUC)
  • Shows potential for expanding credit access to unbanked populations
  • Demonstrates value for financial inclusion initiatives

Contribution and Implications

  • Digital footprints provide an easily accessible alternative to traditional credit scoring
  • Can help expand financial inclusion for 2 billion unbanked adults worldwide
  • Enables FinTech firms to compete with traditional banks in credit assessment
  • Offers potential for reducing defaults while maintaining credit access

Data Sources

  • AUC comparison chart: Table 4 - Default regressions results
  • Device type default rates: Table 2 - Digital footprint variables and default rates
  • Email provider default rates: Table 2 - Digital footprint variables and default rates
  • Default rates over time: Table 9 - Development of default rates around introduction of digital footprint
  • Customer segment comparison: Table 13 - Default regressions for unscorable customers