Smarter Analytics Leadership Summit Content Review
Agenda
Fraud Point of View
IBM Claims Fraud Solution Overview
Infinity Insurance: Combating Fraud with IBM Claims Fraud Solution
Building the Business Case
© 2013 IBM Corporation2
Cross Industry Problem – focused Industries include
Anti Money Laundering• Sanctions screening• KYC scoring• Suspicious Activity
Online Fraud
Transactional
Credit Card
Internal/Employee
Loan or Application
First Party Fraud
Tax/Revenue • IRS, • SSA
Healthcare • Medicaid• Medicare• DoD• others
Food & Nutrition Programs• Food Stamps• WIC
Compliance and Audit• Departments• Agencies
Social Services
Claims Fraud • Cash for Crash• Staged accidents / thefts• Exaggerated claims• Provider fraud
Policy fraud• Fake policies / documents• Rate jumping• Online fraud
Internal / Employee fraud• Fake policies / documents• Pocketing premiums• Workers Compensation
Provider/Vendor Fraud
Pharmaceutical Fraud
Physician / Employee Fraud
Banking Insurance
Public Sector Health Care
© 2013 IBM Corporation3
We integrate technology, people, and experience
Consulting Expertise Intellectual Capital (Asset)
Software Assets and Capabilities
Thousands of cross industry risk indicators
Many years of experience and over a hundred global fraud, waste and abuse
implementations
Best-in-breed software capabilities enable organizationsto harness the IP and Expertise in a complete Fraud, Waste
and Abuse solution
© 2013 IBM Corporation4
What our insurance clients telling us about fraud…
Fraud is pervasive, growing and has become a board room issue
Fraudsters are much more sophisticated, technology enabled and transnational
Fraudulent activity continues to rise, methods more sophisticated
Organized crime rings are more prevalent
Continued and growing concern about e-commerce attacks
Insiders (employees) are a prominent threat
Enormous amounts of data to uncover fraudulent activity is a daunting task (Structured & Unstructured)
There is simply not enough resources to attack the problems
© 2013 IBM Corporation5
IBM’s Claim Fraud Point of View
For Smart Insurers, claim fraud is not treated as a point solution or as a step in the process. It is more than a “score”. It starts before Intake. It is seen as preventable,
predictable, provable, and is managed pervasively across the claim life cycle
© 2013 IBM Corporation6
Attacking fraud is one aspect of transforming to Smart Claims
Business Analytics & Optimization to Improve Claim OutcomesBusiness Analytics & Optimization to Improve Claim Outcomes
Smart Claims initiatives use Business Analytics & Optimization go beyond automation and core transformation to optimize outcomes across all aspects
of the claim life cycle
© 2013 IBM Corporation7
One potential source for improving Combined Ratios is reducing Claim FraudClaim Fraud may account for as much as 10% of a carriers Incurred loss expense…which is especially challenging in today’s market with Combined Ratios rising well above 100
Fraud is a hidden cost – not fully known
• Historically accepted as a cost of doing business
• Very expensive to find and reduce as was very labor intensive
Market conditions are pressuring the bottom line
• Soft market pressuring revenue and profit
• Low interest rate limiting investment returns
• Operational costs already cut to the bone
Crime rings are increasingly turning to fraud
• Organized crime is seeing a low risk, high reward potential
• Banking is more difficult due to AML, so seeing a shift to Insurance
• Lower risk…courts are clogged with violent crimes
Economic downturn historically leads to greater fraud and abuse
• Opportunistic fraud rises during bad economic eras
• Advances in analytics make finding fraud possible & economical
• Higher volume and velocity of data available to identify fraud
• Cheaper, more sophisticated, and faster running models
Carriers are increasing their investments in finding and preventing fraud
FraudDetection
$15+ Billion
savings opportunity
1 to 3%
10%
0%
All Claims
Fraud
100%
Industry estimates suggest 10% of claims are fraudulent, but the actual figures are likely higher.
Historic best practice only find 1-3% of fraudulent claims
Technology is poised to change this and allow the economic
recovery of more fraud
Industry estimates suggest 10% of claims are fraudulent, but the actual figures are likely higher.
Historic best practice only find 1-3% of fraudulent claims
Technology is poised to change this and allow the economic
recovery of more fraud
© 2013 IBM Corporation8
Carriers find the obvious through
normal adjudication process. Room for improvement to find less obvious using
better tools.
Common focus for most P&C (and some Life) carriers. Room for improvement as this is being fought
with “traditional” tools
Carriers are expanding their focus to detect and prevent more types of fraud
Fraud originates from various sources using creative schemes. The combination of new technology and challenging business conditions makes uncovering Opportunistic fraud economically viable
Opportunistic
Organized
Mature for high volume or regulated
Not as strong for low volume/non-
Regulated providers
Currently viewed as too costly to uncover
anything but the obvious. Viewed as a
“Cost of Doing Business”
© 2013 IBM Corporation9
Carriers can be much smarter when it comes to finding and stopping fraud
Lack of Insight
Inability to Predict
Inefficient Access
Variety
VolumeVelocity
Predict and Act(Predict at Intake)
Real-time, Fact-driven(Data Driven)
Smart Claims Approach
Sense and respond(After the Act)
Instinct and intuition(Adjuster Driven)
Traditional Approach
Automated(Claims Admin)
Skilled Adjusters(Instinct)
Back office(SIU)
Optimized(Case Management)
Everyone(Insight)
Point of Impact(Agents, Adjusters, Investigators)
…Results in rapid, informed and confident actions
Today’s insurance companies must transform to reduce costs
© 2013 IBM Corporation10
© 2011 IBM Corporation11
IBM Claim Fraud Appliance
© 2013 IBM Corporation11
Rules Library
Uniquely identifies claim entities and relationships
Manage, Automate, and Collaborate Cases
Predict Suspicious and Anomalous Behavior
Identity Resolution / Social Network Analysis
Resolves identities across claims
Identifies relationships across entities
Business Rules
Industry specific rules• Ex: More patients per day than is possible
Business expertise rules
Predictive Models and Anomaly Detection
Combines structured and unstructured big data
Identifies anomalous behavior – claim, provider, employee, etc.
Case Management
Centralized referral management
Collaboration for improved efficiencies and results
Integrated text mining of social media and other big data forms
Closed Loop Analysis for Fraud Metrics
Fraud Insights
Monitor fraud metrics ensures performance
Provides insights to front-line managers and executives
Robust Approach To Fraud Prevention, Detection and Investigation
© 2013 IBM Corporation12
13
EAS Entity #9453
Attribute Value Source
Name Marc R Smith A-70001Name Randal M Smith B-9103Name Mark Randy Smith C-6251Address 123 Main St A-70001Address 456 First St C-6251Phone (713) 730 5769 A-70001Phone (713) 731 5577 B-9103Phone (713) 731 5577 C-6251Tax ID 537-27-6402 A-70001License 0001133107 A-70001License 1133107 C-6251DOB 06/17/1934 B-9103
EAS Entity #9453
Attribute Value Source
Name Marc R Smith A-70001Name Mark Randy Smith C-6251Address 123 Main St A-70001Address 456 First St C-6251Phone (713) 730 5769 A-70001Phone (713) 731 5577 C-6251Tax ID 537-27-6402 A-70001License 0001133107 A-70001License 1133107 C-6251
EAS Entity #9452
Attribute Value Source
Name Randal M Smith B-9103DOB 06/17/1934 B-9103Phone (713) 731 5577 B-9103
Observations
Identity Resolution – “Who is who?”
Record #70001Marc R Smith123 Main St
(713) 730 5769537-27-6402
DL: 0001133107
Record #9103Randal M Smith
DOB: 06/17/1934(713) 731 5577
Record #6251Mark Randy Smith
456 First Street(713) 731 5577
DL:1133107
EAS Entity #9451
Attribute Value Source
Name Marc R Smith A-70001Address 123 Main St A-70001Phone (713) 730 5769 A-70001Tax ID 537-27-6402 A-70001License 0001133107 A-70001
OCC
Gang
Arrest
Interactions Entity Context
© 2013 IBM Corporation13
Social Network/Link Analysis
Visual and mathematical analyses of complex human and computer networks
Discover sources and distribution of power: Who knows whom; who does business with whom; and who wields greatest power
See how communications flow in networks and sub-networks
Monitor patterns between communications nodes, performance and key goals
Combines data extraction, manipulation, analysis and visualization
© 2013 IBM Corporation14
Decision Management Rules Support
Rules Library and Execution
Incorporates domain and business expertise to identify suspicious claims
Configure business rules easily
• No programming
• No IT required
Rule to value based assignment
• Different rules have different risk
Create and manage thresholds
Graphical, thin client interface
Simulate and test scenarios
© 2013 IBM Corporation15
Predictive Analytics and Anomaly Detection Uncover Fraud Patterns
Combines structured claims, policy and billing data with unstructured data like claim notes
Identifies behavior that is uncharacteristic of similar claims, providers, attorneys, etc
© 2013 IBM Corporation16
Adjuster and Investigator notes can be used in fraud detection models
Combining this vital data with customer, policy, and claim data provides the most robust application for fraud detection
© 2013 IBM Corporation17
Detection Engine: Optimize Fraud Decisions for Better Results
Flexible and adaptive environment for fraud identification
Combines Rules, Predictive Analytics, and Social Network Analysis
Real-time claims monitoring
Business led application
• Easily modify rules and decision outcomes
• Little reliance upon IT
• Less reliance upon Quant department
• Simulation and What/If Analysis
Sophisticated visualization for identifying non-obvious patterns and relationships
© 2013 IBM Corporation18
Case Application Design
Case Lifecycle Management
Case Infrastructure Case ContextCase Runtime
Framework
CaseAnalyticsCase Templates Case
Tasks
Workflow Monitoring & AnalyticsCollaborationRulesContent Social
SoftwareEvents
360o View of Case
Advanced Case Management
people o process o information
Leverages Web 2.0 standard with user configurable widgets, Event processing, built in text analytics, instant messaging
and collaboration tools
Case Management
© 2013 IBM Corporation19
Claims Management Dashboards Provide Immediate Insights
© 2013 IBM Corporation20
The IBM Claim Fraud Appliance was invoked in real time and it
was determined to be a high risk
The IBM Claim Fraud Appliance was invoked in real time and it
was determined to be a high risk
The Fraud Rules presented a low probability of fraud
The Fraud Rules presented a low probability of fraud
However, the analytical model – using claim description notes – indicated a
high probability of fraud
However, the analytical model – using claim description notes – indicated a
high probability of fraud
© 2013 IBM Corporation21