Whitepapers & eBooks

Increased Predictiveness of SMB Risk through Configured Data

Issue link: https://resources.datadrivenmarketing.equifax.com/i/928505

Contents of this Issue

Navigation

Page 1 of 2

2 The Commercial Financial Network While models can be tailored with both internal and third-party data, data source configurability is only part of the value. Equifax can also configure on: ■ Target — Generic scores typically regard any account that goes over 90 days delinquent within a year as "bad." But with configurable scores, lenders can specifically determine what they consider "bad accounts," and adjust models to fit around those specifics. For example, a target can be 60 days delinquent. ■ Population — Generic scores are built around a generic population, such as a national footprint that covers several industries. However, when lenders only interact with certain industries within certain regions, they can create configured scores and models to meet those requirements. ■ Segmentation — Often, a lender's footprint spans different behaviors in different markets. Configured scores and models can help meet these varying needs across many diverse segments. For example, a lender can limit a model to existing customers versus non-customer segments, or small businesses versus large businesses. ■ Technique — Lenders can use non-linear neural networks, regression techniques, fusion and other models to get more granular and predictive. Generic scores and models provide one approach, but when a single approach doesn't match a lender's criteria, they can leverage configured models and scores for tailored solutions that address specific needs and improve performance. The CFN provides greater configurability through: Insights that Drive Configurable Solutions Configured models help lenders know more about small businesses to score more applicants and, ultimately, approve more loans and grow their business. Lenders choose the best strategy for their needs, and then tailor it to each situation. Equifax has the data and insights lenders need to fuel these models, including: ■ Financial, non-financial and consumer data; ■ Over 28 million business records; ■ Insight into $456 billion in small business credit obligations; and ■ Trended commercial data and attributes. In fact, one customer had a 50% lift in predictiveness of risk when evaluating their commercial card lending with configured models versus traditional models. Essentially, with increased configurability, lenders get the performance and flexibility of a custom score without the lead time to develop it. with unmatched modeling and analytical capabilities CONFIGURABLE DATA SOURCES that allow lenders to integrate their own data NEURODECISION TM TECHNOLOGY trended commercial attributes USE OF 500+ TRENDED DATA AND ATTRIBUTES that extend beyond traditional nancial information business records ACCESS TO 28+ MILLION in small business credit obligations INSIGHT INTO $456 BILLION $

Articles in this issue

view archives of Whitepapers & eBooks - Increased Predictiveness of SMB Risk through Configured Data