2. 2
Scott Brackin
Vice President, Auto
sbrackin@factortrust.com
WELCOME TO FACTORTRUST WEBINAR
Dallas Munkus
Director, Customer Analytics
dmunkus@factortrust.com
3. 3
AGENDA
We will discuss:
• The value and types of alternative credit data
• Key components of a comprehensive data set for testing
• Best practices for an effective implementation strategy
4. 4
VALUE OF ALTERNATIVE DATA
M A R K E T O P P O R T U N I T Y T O S TAY C O M P E T I T I V E
Underserved
by Big 3
300-499
6%
500-549
8%
550-599
10%
600-649
10%
650-699
13%
700-749
16%
750-799
18%
800-850
19%
“Big 3” ~$8.5B in 2014 Revenue – Serving the “Prime/Banked”
Credit and Consumer Data Services Market
For Underbanked / Non-prime Largely Untapped
Massive Underserved Market Population
113M
of US adults,
have a FICO score
under 700
FICO Score Distribution of US Population (240M)
With ~$8.5B in combined revenue, the Big 3 credit reporting agencies (“CRAs”) are focused on
covering prime US consumers, leaving a significant market opportunity in the underbanked
population that is non-prime.
From CFPB (*May 2015) – 26 million U.S. adults have no credit history with the Big 3 bureaus, and further
19 million U.S. adults credit data is so limited or out of date with the Big 3 bureaus, that they are
unscoreable. In total, 45 million U.S. adults are living without credit scores.
47% OF U.S. ADULTS HAVE BELOW-AVERAGE CREDIT SCORES
6. 6
VALUE OF ALTERNATIVE CREDIT DATA
N O N - T R A D I T I O N A L V S T R A D I T I O N A L
7. “I am not sure how to effectively utilize alternative data.”
8. 8
COMPREHENSIVE DATA TESTING
E S TA B L I S H C L E A R G O A L S
Know the endgame and the more likely
your objective will be met.
Considerations:
• qualify more consumers
• eliminate credit losses
• improve conversion
• risk adjust pricing
9. 9
COMPREHENSIVE DATA TESTING
K E Y E L E M E N T S
The key to a successful data analysis
involves starting with:
1. Proper input file
– Relevant sample size
– Performance Indicator(s)
– Matured accounts
2. Comprehensive data set, including
̶ Bookings
̶ Rejects
̶ Approved Not Booked
3. List of potential alternative data
providers to a manageable group
10. 10
Measurements
Hit Rate
DisparityImpact
COMPREHENSIVE DATA TESTING
M E A S U R E M E N T: H I T R AT E
Hit Rate – Number of
applicants who would have
been present in the alternative
data file at the time of
application
Separation – Lift provided
by presence in the
database
Impact – Overall
results of cost
benefit analysis
12. 12
• There are large collections of public and proprietary data from thousands of
sources with identity intelligence and verification available that allow you to
make sure that the borrower’s:
– Input data is current and accurate
– SSN was issued and belongs to the name of the borrower
– Name verifications, e.g. marriage licenses
– Physical addresses and date verification
– Birth date verification
IMPLEMENTATION STRATEGY
V E R I F Y O R VA L I D AT E A N A P P L I C A N T ’ S I D E N T I T Y
13. 13
• Knowing that a consumer is not present in an
alternative database is also telling
• Not being present in a short-term loan database
usually indicates an overall lower default rate
• If you simply segment on hit or no-hit, you will
potentially miss additional ranking power from
alternative data attribute suites or scores
IMPLEMENTATION STRATEGY
VA L U E O F A “ N O H I T ”
14. 14
IMPLEMENTATION STRATEGY
O V E R L AY Y O U R R I S K S C O R E
KS is a standard measure for model effectiveness in the credit risk space and most auto
finance companies would consider the 40% lift in KS, as with this case study, statistically and
strategically significant.
15. 15
• Alternative credit data may be used to automatically monitor accounts for
changes or events.
• Triggers alert finance providers on where and when to take action – Events
include:
– New loans or inquiries on file
– Change in salary
– Recent or late payments
– Change in debt-to-income
– Most recent address
– Most recent phone number
IMPLEMENTATION STRATEGY
A L E R T S T O R I S K A C T I V I T Y
17. 17
QUESTIONS AND NEXT STEPS
To inquire about data tests and analysis,
contact:
Scott Brackin
Vice President, Auto
E: sbrackin@factortrust.com
Twitter: @underbankedstew