Scott Horwitz

advertisement
Product Showcase:
FICO Roadmap for Insurance Fraud
Analytic Innovations To Reduce Loss Rates
Michael Betron
Product Management
FICO
Scott Horwitz
Insurance Consulting
FICO
© 2014 Fair Isaac Corporation. Confidential.
This presentation is provided for the recipient only and cannot be reproduced or shared without Fair Isaac Corporation’s express consent.
Insurance Fraud Roadmap
One third of insurers recently surveyed
pegged their fraud loss rates at 10–20% of
claims volume.
Are you one of them?
2
© 2014 Fair Isaac Corporation. Confidential.
Agenda
►P&C
Insurance Market Overview and
Fraud Trends
►FICO®
Claims Fraud Solution
►Demo
►Business
Value
►Questions
3
© 2014 Fair Isaac Corporation. Confidential.
P&C Market Overview and Claims Fraud
4
© 2014 Fair Isaac Corporation. Confidential.
The Insurance Fraud Problem
► United
States
► $30–$40
billion lost to fraud each year
►
Higher incidences of opportunistic fraud
► Increase in organized crime rings
► UK
(ABI Statistic)
► Undetected
general fraud totals £1.9B per year
According to the FBI, fraud costs
property/casualty insurers more than $40
billion annually. In 2012, those losses
exceeded the industry’s profits by about 20
percent.
Insurance Network News
May 2014
5
© 2014 Fair Isaac Corporation. Confidential.
The Scope of Fraud Schemes
Fraud Schemes Cross a Wide Range of Perpetrator Intention,
Potential Impact and Planning
Opportunistic
►
6
Eric’s home is broken into.
How does it hurt if he claims
he had an 80” television?
© 2014 Fair Isaac Corporation. Confidential.
Premeditated
►
Marissa knows most insurers
will give her a quick buck for
whiplash to settle a claim
quickly before it escalates to
higher payments. She slams
on her brakes in front of a
tailgating policyholder
Organized
►
Michael takes his job
seriously. He knows the
more people he can pack
into the two junk vehicles he
is staging a crash with, the
higher the payout
The Problems We Solve
Profitability
Reduce fraud losses—and drive revenues, improving capital position and profits
Customer Satisfaction
Reduce false positives in balance with risk appetite
Loss Reduction
Detect more fraud—and emerging fraud schemes—before claims are paid
Operational Efficiency
Prioritize work for more efficient use of customer service resources
7
© 2014 Fair Isaac Corporation. Confidential.
FICO Claims Fraud Solution
► Powerful
combination for finding
maximum fraud
► An
integrated environment for
detection and investigation
Predictive
Analytics
Business
Rules
Social
Network
Analysis
► Analytic
feedback loops for
continuous learning
► Mechanisms
for prioritization for
efficient processing
► Finds
new, emerging fraud AND
fraud rings
8
© 2014 Fair Isaac Corporation. Confidential.
Case Management and Investigation
Business Rules/Model Authoring Tools
► Utilize
detection results and business
rules to assign the high risk claims or
entities to the right resource
► Guide
investigators to the right next
action, based on:
► Likelihood
of fraud (rank ordered by
fraud score)
► Reason codes
► Control
score thresholds for claim
referrals
► Create
9
Flash Fraud Rules
© 2014 Fair Isaac Corporation. Confidential.
Analytic Capabilities
Fraud Trained
Outlier Detection
SNA
Learn complex relations from identified
fraud in historical data, then leverage
this to predict fraud
Uncover new types of fraud
by identifying outlier and
aberrant claims
Social Network Analysis for identity
and entity resolution
Fraud
►
Use tags to differentiate the
innocent from the guilty
10
© 2014 Fair Isaac Corporation. Confidential.
►
Learn patterns and identify
aberrance
►
Connects fraud rings across
seemingly unconnected data
by detecting and linking identity
information
Improving Claims Fraud Analytics with SNA
Analytic Capabilities
Customer/account level
indicators
Simple schemes
Manual referrals
Entity/network level
indicators
Subtle, complex cases
Previously unknown patterns
Fraud Rings
Early Detection
Claim Level
Entity/Network Level
Association with Known Bads
Individual on a hot list appears as a claimant
Multiple cases with fraud outcomes and
recoveries exist on this network
Domain Specific Rules
Delayed claim report date
Incident date is close to policy inception
This network shares injured third party
claimants across otherwise unconnected
incidents
Statistical Outliers
Medical billing is unusually high given typical
claim type profile
High velocity of claims on this network
compared to most networks
Predictive Analytics
Claims containing a high number of soft tissue
injuries is correlated with a fraud outcome
Networks which contain repeated incident
descriptions across claims are correlated with
fraud outcomes
11
© 2014 Fair Isaac Corporation. Confidential.
Case Management/Investigation
12
© 2014 Fair Isaac Corporation. Confidential.
Link Analysis/Visualization
13
© 2014 Fair Isaac Corporation. Confidential.
FICO® Claims Fraud Solution Architecture
User Interaction
Alert Manager
Link Visualization
Fraud Rules/
Strategy
Reporting
128 Bit SSL Encryption
DMP
Analytic
Services
Components
Claims Fraud
Service
Onboarding
Services
Other Fraud
Service
Routing
Service
Rules
Services
Data
Services
Development/Testing and Deployment
Scoring
Governance
Fraud Rules
Change
Management
Audit
Transparency
Compliance
Routing Strategies
Link Analytics
Deployment
Manager
Input Data
Data Server
Third
Party
14
© 2014 Fair Isaac Corporation. Confidential.
Data Warehouse
Operational Data Store
Fraud Rule and Models
FICO® Claims Fraud Solution—Demo
Participant Claim
Rick Meadows
► Policy
holder for 18 months
► Involved in a 2 car collision
► Multiple passengers in
each vehicle
► Multiple injuries and claims
15
© 2014 Fair Isaac Corporation. Confidential.
FICO® Claims Fraud Solution
FICO’s Key Benefits and Value Proposition
Detect more Fraud
Prioritize Claims
Reduce False
Positives
Stop Fraud Rings
Grow Customer
Satisfaction
Gain Financial
Benefits
16
© 2014 Fair Isaac Corporation. Confidential.
Find more fraud with predictive analytic methods that cast a wide net as well as learn from
one another—producing significant impact to the bottom line
Prioritize SIU investigations—tackle the fraud with the highest ROI based on fraud score
rankings
Via the use of analytics to create more operational efficiencies; optimize models to balance
false positive rates with risk appetite
Detect fraud linked by as little as a single related item—exposing fraud rings and greater
volumes of claims fraud
By streamlining the process and paying legitimate claims faster
Realize substantial ROI from day one, impacting the bottom line and capital position
Questions
17
© 2014 Fair Isaac Corporation. Confidential.
Thank You!
Michael Betron
Product Management
FICO
Scott Horwitz
Insurance Consulting
FICO
© 2014 Fair Isaac Corporation. Confidential.
This presentation is provided for the recipient only and cannot be reproduced or shared without Fair Isaac Corporation’s express consent.
Learn More at FICO World
Related Sessions
►Sul America Session
►Karen Pauli Session
Products in Solution Center
►FICO® Identity Resolution Engine
►FICO® Claims Fraud Solution
Experts at FICO World
►Liz Lasher
►Scott Horwitz
►Mike Betron
White Papers Online
►Gartner Newsletter
Blogs
►www.fico.com/blog
19
© 2014 Fair Isaac Corporation. Confidential.
Please rate this session online!
Michael Betron
Product Management
FICO
20
© 2014 Fair Isaac Corporation. Confidential.
Scott Horwitz
Insurance Consulting
FICO
Business Processing and Orchestration
Flash Fraud Rules
22
© 2014 Fair Isaac Corporation. Confidential.
Routing Strategies/Threshold
Download