; Australian Banking Industry Reports
Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>

Australian Banking Industry Reports

VIEWS: 53 PAGES: 98

Australian Banking Industry Reports document sample

More Info
  • pg 1
									Risk Solutions User Forum

Jeff Bottari, VP Risk Solutions Group
CheckFree




October 24, 2007
Welcome!!




            2
User Conference Objectives

 Very few CheckFree commercials
 Shared experiences using CheckFree products
 Shared industry concerns
 A time to talk with other banks about issues
 Advise CheckFree on what you like, don‘t like, and would like
  to see
 A feeling of community
 An opportunity to advance best practices
 It should be fun!



                                                                  3
          Jeff Bottari




                                                 Compton Harry
                          Debb
                         Gordon




                                  Don Crosby



Michael
                                               Robert McCann
Bunyard
                            Pete
                         Sullivan




Karen Taylor

                         Dee Millard




           Jeff
           S-S-Sargent

                                       Mark Steeber
         Roger
          Snell




                  Angela Bardowell




Rich
Rosner




                     Dan Barta
Whatever happened to Carreker?




                                 7
CheckFree / Carreker

 Acquisition completed on April 1st, 2007
 CheckFree is a $1 billion company with 4,400 employees world-
  wide, located in more than 20 different cities
 Carreker‘s current solutions are being integrated into
  CheckFree‘s current product structure
 The combination of the two predecessor companies makes
  CheckFree an industry leader in software applications that cross
  the traditional check-based & ACH payments arena
 We are uniquely qualified to help banks balance customer needs
  with needs for greater efficiency and profitability, as an already
  diverse payments environment continues to evolve and change



                                                                       8
CheckFree Snapshot

 Premier provider of financial electronic
  commerce services and software products
 Founded by current Chairman & CEO
  Pete Kight in 1981
 Became a publicly traded company in 1995
 26 years in operation
 2006 in Review:
   Revenue of $972.6 million
   Nearly 1.3 billion transactions processed
   Nearly 226 million electronic bills delivered
   Nearly 2.7 million portfolios under management at year end




                                                                 9
Whatever happened to CheckFree?




                                  10
Carreker/CheckFree‘s
Risk Management Solutions


 We are the premier supplier of Enterprise Risk and Fraud
  Mitigation Solutions.
 Our Pragmatic Convergence approach provides financial
  institutions with maximum protection via multi-channel
  transaction monitoring and customer behavior modeling.




                                                             11
Pragmatic Approach Defined

                  prag·mat·ic [prag-mat-ik] -adjective:
                  Concerned with practical matters; “a
                  matter-of-fact approach to the problem”
                                              — Webster



 The destiny: An enterprise risk mitigation platform which
  correlates fraud across access points and channels by customer

 Allows you to leverage your existing investments to create an
  achievable strategic plan

 Stay ahead of the fraudsters while gradually adding functionality

 Each step provides a return on investment in months not years

                                                                   12
Weaknesses of Current Risk Management Models

 Largely a Day 2 Process… Limited Day 1 and Day 0 Analytics
   Day 0: Real time instantaneous transaction monitoring at Customer
    Access Point — Proactive
   Day 1: Same-day analysis of transactions before posting (near real
    time or multiple batch runs) — Reactive
   Day 2: Batch analysis after Posting — Reactive
 Different capabilities in different silos
 No ability to correlate transactions in multiple customer access
  points in multiple timeframes
 Multiple analysts working same accounts in different channels


                                                                         13
Example of Fraud Detection in Individual Silos

     On-Us /                                                 Internet /
                                   ACH                                             Wires
     Deposit                                                    ATM
   Scoring Engines,          Scoring Engines,                Scoring Engines,   Scoring Engines,
    Models, Rules             Models, Rules                   Models, Rules      Models, Rules


     Shared Data                 Shared Data                   Shared Data        Shared Data


     Alert Mgmt                  Alert Mgmt                    Alert Mgmt         Alert Mgmt
       System                      System                        System             System



                                               Results in:
                             •   Duplication of solution investments
                             •   High/unnecessary IT overhead
                             •   Duplication of data and resource expenses
                             •   No leverage of cross-silo alerts

                   Sophisticated Fraudsters Will Find The Weakest Link                             14
The Growing Complexity of Fraud
                                        Customer Access Points




  Branch       Lockbox         ATM           Wires          Merchants        Telephone           Web          Call Center
                                                                              Banking




                                           Bank’s Challenges
               Increased                                                                  Balancing
                                        Maintaining   Achieving Risk
                           High False                                   Constrained       Customer
  Compliance     Fraud                   Silo Fraud    Management                                       Employee Fraud
                            Positives                                     Budgets        Satisfaction
                  Loss                    Systems     Best Practice
                                                                                          With Risk


 Transaction                                                                                               Customer
 Monitoring                                                                                                Behavior
                            Enterprise Risk Management
               Proactively identify fraud in and across channels to
                     mitigate financial and reputational loss

                                                                                                                            15
Industry Best Practice:
Enterprise Risk Management

 Holistic View of transactions, accounts, and relationships
 Monitor all transactions for suspicious behavior
 Analyze monetary and non-monetary data
 Enable creating rules containing cross-channel variables
 Manage potential fraud cases effectively, from detection through
  law enforcement reporting
 Move to Proactive vs. Reactive




                                                                 16
Carreker/CheckFree
Enterprise Risk Management
                                      Alerts                           Fraud

   Detection                                           Alert                       Case
  Management                                        Management                  Management
  Liability Accounts
 On-Us
         Deposit
                   Wires
                           ACH   Fraud Manager     Workflow Manager            Syfact Investigator

                                                        Acquire                   Investigation
                                 Other Detection

   Credit Accounts                                     Research
                                                                                  Link Analysis
                                 Other Detection       Decision
                                                                                    Reporting
 Investment Accounts
                                                        Analyze
                                 Other Detection


                                        Day 0, Day 1 or Day 2 Capabilities
                                                                                                     17
Enterprise Risk Management
     Data                                                            Data                                Workflow
  Acquisition                               Detection               Staging                              Manager               Disposition



 External                                   On-Us                                                              Alert
                                           Real Time   Modeling
  Data                                                                                                      Management
 Sources
                                            Deposit
                                           Real Time
               Data Acquisition Engine




                                                       Segments                                               Research
                                            On-Us
 All Trans-                                Day 1 & 2
actions File




                                                                                        Alert Packager
                                                                     Suspect Database
                                            Deposit
                                           Day 1 & 2                                                     Decisioning/ Fraud
                                                       Profiles
                                                                                                         Analyst Workstation
 Internal
                                              ACH
                                                                                                                                  Case
   Data                                                                                                                        Management
 Sources                                                User
                                                       Defined                                               Reporting
                                              ATM       Rules

FraudLink
                                            Internet
  On-Us
                                            Banking                                                          Queries /
Mainframe                                               Filter
                                                                                                             Dashboard
                                             Wires

FraudLink                                                                                                      Work
 Deposit                                     Other       Lists                                              Distribution
Mainframe




                                         Credit            LRM    ATM/Cards             Treasury Mgmt
                                                                                                                                            18
Dashboard Example

                            Enterprise           Region               Customer

  Frauds   On-Us           Deposit        ACH             Wires           Loans           Online        Internal



            Total Customers Alerted          Total Fraud Volume YTD               Number of Alerts
            (000)‘s                                                               Process per FTE
                                                                                     per Hour




                      Top 5 Customer Alerted                          Top 5 Alerted Amounts
                        Regions in 4Q 2005                             Regions in 4Q 2005
                 Regions      # Alerted % Increase              Regions       $ Alerted % Increase
                                        from last Qtr                                   from last Qtr
             Region 12            7,245            15       Region 12         3,796,380            15
             Region 14            5,895            10       Region 14         2,528,955            10
             Region 5             5,432             7       Region 8          2,371,908             7
             Region 8             5,236             3       Region 5          2,297,736             3
             Region 11            2,529             0       Region 11         1,044,477             0


                                                                                                                   19
Benefits of Enterprise Risk Management
 Efficiency
   Automated processes
   Review fraud-rich pool of suspects with no addition to staff
   Single platform for all fraud mitigation
 Effectiveness
   Improved fraud detection
      Lower false positives, reduce false negatives
   Improved analyst job satisfaction
 Flexibility
   Dynamic creation of rules
   Image-based workflow
   Champion vs. Challenger
                                                                   20
Agenda Day One

 Welcome and Introductions                Jeff Bottari, VP Risk Solutions Group,
                                            CheckFree

 Enterprise Alert Management:             Silvia Sarra, AVP Loss Prevention &
  Managing Alerts More Effectively          Security, Sovereign Bank
                                           Dan Barta, Service Delivery Manager,
                                            CheckFree

 Citibank and CheckFree Fraud Manager     Debb Gordon, Director Business
  Deposit Case Study                        Architecture and Analysis, CheckFree




 Reducing False Positives: Effectively    Lisa Zarzycki, Vice President and Risk
  using Account Types and Period            Manager Fraud Services, Comerica
  Parameters                                Bank

                                                                                     21
Agenda Day One, continued

 Internet Banking Fraud Trends         Carly Boardman, Manager Cheque
                                         Compliance & Fraud Detection, ANZ
                                        Peter Casey, Manager Fraud
                                         Detection, ANZ

 Understanding Your Bank‘s ―Fraud      Mark Steeber, Risk Advisory
  Profile‖: A Risk-Based Approach to     Consultant, CheckFree
  Re-calibration

 Closing Remarks                       Jeff Bottari


 Cocktail Happy Hour                   Tapas Bar and Drinks Reception
                                         Terrace Bay Lobby, Lower Level


 Client Conference Event               San Diego Zoo



                                                                             22
Agenda Day Two

 Emerging Fraud Trends:          Angela Bardowell, Director Risk
  What trends are you seeing?      Consulting Group, CheckFree

 Product Roadmap                 Michael Bunyard, Director Product
                                   Management, CheckFreeGroup
                                   Discussion

 Product Roadmap Session with    Group Discussion
  Customer Input

 Meeting Wrap-Up                 Jeff Bottari, VP Risk Solutions Group,
                                   CheckFree




                                                                            23
Questions?
Enterprise Alert Management:
Managing Alerts More Effectively

Silvia Sarra, Sovereign Bank
Dan Barta, CheckFree




October 24, 2007
What is Enterprise Alert Management?

       en·ter·prise [en'-ter-prahyz] −noun:
       1) a project or undertaking that is especially difficult,
       complicated, or risky
       2) readiness to engage in daring or difficult action: initiative
       <showed great enterprise in dealing with the crisis>
       3) a unit of economic organization or activity; especially:
       a) a business organization b) a systematic purposeful
       activity <agriculture is the main economic enterprise among
       these people>
                                                             — Webster




                                                                          26
What ―Enterprise‖ will we be Discussing Today?
 Enterprise Definition and Scope
   Focus on transaction accounts (DDA & SAV)
   Focus on payment transactions and account opening
   Limited inclusion of money laundering analysis
   Focus on fraud and loss prevention activities


 Other areas that could be included
   Mortgage and other lending transactions
   Investment accounts (brokerage, mutual funds, etc.)
   Insurance
   Other Industries

                                                          27
Enterprise Alert Management

                               Payment Channels

      Check          ACH       Debit        Credit     Wires         ATM




                                 Detection Tools
     FraudLink     FraudLink                          Fraud Mgr
                               Falcon       Falcon                Other Tools
   On-Us/Deposit    ACHeCK                              Wires




      Suspect       Suspect    Suspect      Suspect    Suspect      Suspect
      Report        Report     Report       Report     Report       Report




                                                                                28
Enterprise Alert Management

                               Payment Channels

      Check          ACH       Debit        Credit     Wires         ATM




                                 Detection Tools
     FraudLink     FraudLink                          Fraud Mgr
                               Falcon       Falcon                Other Tools
   On-Us/Deposit    ACHeCK                              Wires




      Suspect       Suspect    Suspect      Suspect    Suspect      Suspect
      Report        Report     Report       Report     Report       Report




                                 Workflow Tool
                                                                                29
Workflow Management Functions

                                      Elimination of Paper Reports
FraudLink On-Us
                                      Aggregation of Suspects by
                                       Account or Relationship
FraudLink Deposit

                                      Suspect/Alert Priorization
Early Warning ANF & RNF
                           CORE
                          Workflow    Work Assignment
Earns                     Manager
                                      Record Resolution/Action
                                       Information
Bank Specific Suspect/
Alert Tools
                                      Statistical and Other Reporting

Kite
                                      Data Mining




                                                                         30
Workflow Management Functions

                                                Elimination of Paper Reports
FraudLink On-Us
                            Mainframe
                          Communication         Aggregation of Suspects by
                                                 Account or Relationship
FraudLink Deposit

                                                Suspect/Alert Priorization
Early Warning ANF & RNF
                             CORE
                            Workflow            Work Assignment
Earns                       Manager
                                                Record Resolution/Action
                                                 Information
Bank Specific Suspect/
Alert Tools
                                                Statistical and Other Reporting

Kite
                                                Data Mining


                                  Document
                                  Generation


                                                                                   31
Benefits of Enterprise Alerts Management
 Utilization of Database software
 More complete view of risk at the account/customer level
 Better Prioritization of Suspect/transaction Activity
 Elimination of Redundant Effort
 Smarter/Faster Decisions
 Historical Picture of Suspect/Alert Activity
 Research capability
 Elimination of Paper Reports




                                                             32
Sovereign Bank – Company Overview

 Sovereign‘s headquarters in Wyomissing, PA
   $82 billion financial institution
      Markets primarily in the Northeast United States
   750 Community Banking Offices (CBOs) & 2,250 ATMs
   18th largest banking institution in the United States
   Successfully completed two dozen acquisitions since the late 1980s




                                                                         33
Loss Prevention – Operational Overview

 Centralized Loss Prevention Unit
   Team of 44
 Check fraud prevention (Deposit & On-Us)
 Case Management case input
 Centralized check fraud claims
 Debit card (signature and pinned)
   Fraud claims
 Single point of contact for ID Theft
 CBOs, customers, and other Sovereign units‘ support via a toll
  free response line
 Elderly Abuse
                                                                   34
Business Drivers to Implement
Enterprise Alert Management
 Mergers and Acquisitions
 Standardize staff training
 Establish a suspect/victim model
   Inability to prioritize highest risk alerts
   Analysts working in silos i.e. same suspects in multiple reports
 Manual processes
   Customer notifications (Reg CC)
   Re-keying same info in several applications
   Unable to identify new trends
 Lack of audit trails
 Paper driven
                                                                       35
Staff Efficiency & Operational Gains

 Prioritization of highest risk accounts
 Elimination of manual processes
   Customer notifications
   Connection to host system eliminating re-keying of same date
   Audit trail (tracks every keystroke)
   On average it takes 5 minutes vs. 10 minutes to make a decision to
    pay/return/hold/freeze
   Holistic review of suspects
   At a glance history of suspect transactions
 Detection rate of alerts reviewed year to date averages 90%
 Return on investment (ROI) year to date averages 22:1
 4 FTE reduction
                                                                         36
Customer Service Impacts

 Standardize notification to customers
   Info populating by pulling from host systems hence less chance for
    typos
 Any Analyst can assist customer that calls inquiring about a
  notification they received, less time spent on the phone




                                                                         37
Questions?
Citibank and CheckFree
Fraud Manager Deposit: A Case Study


Gail O‘Brien, Citibank
David Fapohunda, Citibank
Debb Gordon, CheckFree




October 24, 2007
Citibank Business Background

 Successfully used FraudLink for both On-Us and Deposit Fraud
  Detection
   However, false positive and false negative rates were becoming a
    continuing burden to the operation
 Current priority: Improve the efficiency of Deposit fraud detection
   Deposit False positive alerts were 683 to 1 for the sample period
    (8/1/2005 to 9/29/2005) tested
   FraudLink Deposit (ASI-19) was missing on average 52% of the
    Fraudulent transactions (false negatives) and these missed
    transactions accounted for an average of 62% of the Actual Losses
 The Goal for Carreker/CheckFree‘s Risk Solutions Analytic Team:
   Reduce total alerts by 50% and capture at least 98% of the
    current fraud alerted
   Enable the current rules set to be relaxed to alert the missed fraud
    with the same volumes currently used
                                                                           40
Analytic Project Background

 Early 2006, Carreker/CheckFree approached Citibank to perform
  a validation of the statistical models created from pooled bank
  data
 Citibank initially provided FraudLink Deposit Transaction alert
  data from 8/1/2005 thru 9/29/2005
 The Risk Solutions Analytic Team scored the transactions on the
  Generic model and developed a Custom Model for Citibank
 Following the Development process, Citibank provided three
  months of blind data (11/1/2005 thru 1/31/2006) to be scored
  and returned to their analysts
 The model was successfully able to meet the project criteria of a
  total alert reduction of 50% while maintaining a fraud detection
  rate of at least 98%
 21 months later, the validation was repeated and replicated the
  results
                                                                      41
Advanced Analytics

 System Capabilities
   Modeling
      Statistical fraud models designed and tailored to fit behavior in
       each institution
   Rules
      Custom defined rules written and published by the operation
   Lists
      Can be imported from an outside source, or created by the
       operation
   Segmentation
      Create segments that can be serviced with different logic
   Filters
      Filters limit what you want to alert                                42
Advanced Analytics

 The Score
   Each transaction is scored based on good customer profiles
   Scores range from Zero to 1,000, the higher the score the more
    likely it‘s fraud
   Scores are presented in a distribution, you pick the cut-off score that
    best fits you
   Use the score to prioritize workflow, or use a combination of score
    and any other information you use today




                                                                          43
Analytic Study Results

 Deposit Model and Blind Testing
   False Positives were reduced by 51%
   Fraud Capture with existing FraudLink alerts was 98%
 21 Months later
   False Positives were still showing a reduction on average of 45%
   Fraud Dollar Capture with existing FraudLink alerts was 98.3%


 The reduction in total alerts allows for relaxing existing
  FraudLink rule sets to allow for more of the false negative frauds
  to be scored



                                                                       44
Citibank‘s Business Application

 Scored transactions
 Defined rules
 Prioritization in Workflow
 Combining different information for better decisioning




                                                           45
Conclusion

 Based on these Model Validation studies, Citi expects a
  significant reduction in alert volume
 Combining the use of the score with other user written rules can
  improve these results even more.
 Citi is looking forward to greater operational efficiency in Day 2
  Batch
 Future releases will bring the detection to Day 0 Real Time,
  allowing for automated holds and returns at the point of Deposit




                                                                       46
Questions?
Comerica‘s Experience with FraudLink Deposit
Reducing False Positives:
Effectively using Account Types and Period Parameters


Lisa M. Zarzycki, Comerica




October 24, 2007
Comerica Overview*

 $58.6 Billion in total assets
 401 Banking Centers in 5 States
    Michigan,
    California,
    Texas,
    Florida, and
    Arizona
 Select businesses operating in several other states, as well as
  Canada, Mexico, and China
 Among the 20 largest U.S. banking companies

*As of July 18, 2007

                                                                    49
FraudLink Deposit History

 Comerica installed FraudLink Deposit v2.0 in October 2003.
 With the exception of ―home grown‖ ATM deposit fraud reports,
  Comerica had no deposit fraud prevention tool.
 Comerica estimated $375,000 in loss avoidance in the first year.
   Actual Loss Avoidance: $1.8 M
   Total At Risk: $1.9 M
   230 Cases
   Average Prevention: $7,800




                                                                  50
FraudLink Deposit Rules at Inception
 5 of the 7 available rules (excluding 3 & 5)
 3 Account Periods
 3 Separate Markets
 13 account types
   Type A – Access & Value Ckg     *Free Retail
   Type C – Correspondent Banks
   Type E – Employee Accounts
   Type H – High End Retail Ckg
   Type I – Interest Retail
   Type L – Large Business
   Type M – Interest & MMIA Bus
   Type O – Other Business         *Professional (Drs., IOLTA, etc.)
   Type R – Regular Chg
   Type S – Small Business         *Free Business
   Type 1 – Retail Savings
   Type 2 – Business Savings
   DFLT – Default Accounts         *Deposit is made in market other than home market
                                                                                        51
FraudLink Deposit Inception

 Average suspects per day –
  2,137
 4 FTE reviewed 994 or 46%
  on average
 To manage volumes, analysts
  review ―high risk‖ account
  types and high risk markets.




                                 52
Upgrade to FraudLink v3.0

 August 2006, Comerica Upgrades to FraudLink v3.0
 Charge off analysis reveals that 80% of deposit fraud losses
  occur in the first 90 days and 63% of deposit fraud losses occur
  in the first 10 days of account opening
 Move to 6 Account Periods:
   0 to 10 days
   11 to 90 days
   91 to 180 days
   181 to 365 days
   366 to 1095 days
   Greater than 1095 days
 Update Parameters Based on Account Periods
 Enable Rule 3 (36% reduction in suspects)
                                                                     53
Additional Filters

 January 2007, filter added to work flow management system
 If the count of the number of times an account has suspected is
  greater than X times (Y or more) the alert will not be passed
  into the system to be worked by an analyst.
 The filter is not applied to the FLK system but rather to the
  output from the system.
 This allows the group to identify in a charge off analysis if the
  filter caused the account not to suspect and there was
  subsequent fraud.
 The filter reduced suspects by an additional 16%.




                                                                      54
FraudLink Today

 Average: 1533 suspects per day
 Staffing: 5 FTE
 Results: 1080 suspects reviewed or 73%
 Focus on ―high risk‖ account types as defined by loss analysis
 High Risk Account Types:
   70% of Suspects
   82% of Deposit Fraud at Fisk
   79% of Deposit Fraud New Losses
 Continuous charge off analysis to identify high risk account
  types and markets and manage false positives


                                                                   55
Questions?


Contact Lisa M. Zarzycki
248-371-6742
Australia and New Zealand
   Banking Group Limited
                     Carly Boardman
                            Manager
                   Fraud Detection &
                  Cheque Compliance



                         Peter Casey
                            Manager
                     Fraud Detection
ANZ Banking Group Limited
• One of the 5 largest and most successful companies in Australia and
  the number one bank in New Zealand


• Represented in our primary markets of Australia and New Zealand, as
  well as Asia, the Pacific, the UK, Europe and USA


• 781 branches in Australia and 1,265 other worldwide points of
  representation


• 6 million customers worldwide – personal, private banking, small
  business, corporate, institutional & asset finance


• USD$298 billion in assets


• Employ more than 30,000 staff worldwide
Financial Intelligence Operations
                 Education of ANZ branch staff, detection of on us and deposit fraud by way of:
   Cheque         Fraudlink ASI16 & ASI19, Fraudlink Cheque Order Report, Fraudlink Kite
    Fraud         ANZ Visual Image Archive
  Detection       Data exchange with other Fraudlink enabled banks
                  Physical examination of large amount cheques

   Internet &   Detection of internet and phone banking fraud by way of:
Phone Banking  Fraudlink Billpay
Fraud Detection  Eunexus Internet Intelligence System


                 Assist with the design, approval and production of ‗special print‘ cheques
    Cheque       for ANZ business customers.
  Compliance     Represent ANZ on the Australian banking industry Printing Standards
                 Committee.

                 Team currently ‗under construction‘.
  AML / CTF      ANZ AML Program looking to introduce new technology and processes to
                 meet revised Australian legislation that will ensure compliance with
                 international standards (FATF).


                 Filter inward and outward messages against lists, looking for Sanctioned
    Denied
                 Parties, Countries, Assets (Commodities), Currencies by way of
   Payments
                 Metavante‘s Prime Compliance Suite of products.
    Australian Banking Industry
    • 8 ‗Tier 1‘ and 35 ‗Tier 2‘ banks
    • All ‗Tier 1‘ banks are image processors
    • 99.9% of all cheque value is exchanged electronically
    • All dishonours/returns are exchanged electronically between banks
    • 3 day cheque clearance cycle (funds available on day 3)
    • FraudLink enabled banks work collaboratively to combat cheque fraud, i.e.
      daily exchange of ‗suspect‘ cheque transactions
Thousands

                              12 month period




                                                                 ANZ saves to other banks via
                                                                 ASI19

   $                                                             Other bank saves to ANZ via
                                                                 ASI19




            Bank A   Bank B       Bank C        Bank D   Other
          Australian Domestic Payment Streams
                                  Avg. transaction volume per month

Million

                  Cheque             Online Debit
  $120
                  Online Credit      ATM


  $100            EFTPOS             Credit Card



   $80


   $60



   $40



   $20




           1997      1998         1999     2000     2001        2002         2003        2004         2005         2006            2007

                                                    Figures obtained via the Australian Payments Clearing Association (APCA) Ltd
              Australian Cheque Fraud Experience
    In 2006                  $148M in attempts, $32M in losses
                             Losses represent 0.0005% of cheque transaction volume
                             Losses represent 0.0019% of cheque transaction value


   Breach of Mandate
                                                                                                       % of Value

                                                                                                       % of Value
Third Party Conversion



           Counterfeit


       Theft / Forgery



            Alteration


Valueless / Kite Flying


                                 10%           20%                 30%                 40%                 50%

                                         Figures obtained via the Australian Payments Clearing Association (APCA) Ltd
ANZ Internet Banking
Functionality
   Balance and Transaction Enquiries
   Pay Bills to over 10,000 registered billers, e.g. utility companies
   Receive, view and pay bills online
   Transfer between connected accounts
   Transfer funds to accounts held with other Australian banks
   Transfer funds overseas
   Multi Payments, e.g. company payroll
   Purchase a Bank Cheque or International Draft
   Secure Mail
                           Security
                              128-bit SSL Encryption
                              Firewalls to prevent access to the ANZ network
                              Automatic time-outs
                              Fraud Detection (FraudLink)
                              Limited use of Two Factor Authentication
ANZ Internet Banking
What Have We Done?
   Implemented Fraudlink Billpay (2004)
   Integration of Eunexus Internet Intelligence System to enrich Fraudlink
    Billpay
   Real-time sharing of IP intelligence with other Eunexus enabled banks
    in Australia
   Consumer Education & Awareness Campaigns
   ANZ won Financial Insights Innovation Award in the category of
    security & fraud management
   Decreased losses by 40% increased detection by 80%


What We Plan To Do?
   Migration to CheckFree‘s Fraud Manager Platform (Business Case in progress)
   Move to real time detection (as opposed to intra day batch)
   Login Session Monitoring – Stop the fraud before it happens
   Continue Consumer Education & Awareness Campaigns
 ANZ Internet Banking … the journey so far.
Aug 04   Jun 05   Sep 05     Dec 05    Mar 06     Jun 06    Sep 06   Dec 06   Feb 07   May 07   Aug 07   Sept 07



  Implemented FraudLink Billpay
                           Multiple intra day FraudLink Billpay suspect alerts

                           Delay introduced to ANZ Credit Cards via BPAY

                           IP Address Range Introduced to Carreker

                                      Float applied on intra ANZ transfers

                                            ANZ to OFI transfers delayed until EOD transmission

                                                 Increased Data Flow to Carreker (Tele-Code -
                                                 IP Address) Password Changes

                                                           Integrated Eunexus IP data into Fraudlink Billpay
                                                           suspect alerts

                                                                Multi Payment facility exploited

                                                                Temporary loss of IP address data from
                                                                FraudLink Billpay
                                                                suspect alerts
                                                                                         IP address data
                                                                                         returned to
                                                                                         FraudLink Billpay
                                                                                         suspect alert
ANZ Internet Banking
How we are tracking against increased transaction volume …

                                                             Billion




                                                                   Transaction Value
Some Fraud Alert Triggers
   Payee or Recipient is not in previous account history
   Dollar Value is ―Greater Than Average‖
   IP Address originates from an overseas destination

   IP address has been marked as ‗fraud‘ by another Eunexus
    enabled bank

   IP address identified as either ‗malicious or ‗proxy‘ by the
    Eunexus Internet Intelligence System
   IP Address has never been used previously by customer

   Payment message or reference entered at the time of
    transaction, considered suspicious

   Payments to ‗high risk‘ billers (gambling institutions or money
    transfer agents)
   Telecode/Password resets (for telephone banking channel)
   Weight of a suspect alert (10 – 20 – 30 – 40)
IP Sharing & Reporting
 IP Sharing

     75% of all Australian banks, are using IP data provided by
      Eunexus
     ‗Eunexus‘ enabled banks are actively sharing IP intelligence,
      thereby effecting an industry approach to internet fraud, e.g.
      blacklisting IP addresses

 Reporting to Government

     Australian banks report all cases of online fraud to the
      Australian High Tech Crime Centre (AHTCC)
     The AHTCC is a collaboration between the government and
      private industry to enable a national and coordinated
      approach to combating serious, multi-jurisdictional technology
      enabled crimes.
Questions?
Understanding Your Banks ―Fraud Profile‖:
A Risk-Based Approach to FraudLink Re-Calibration



Mark Steeber, CheckFree




October 24, 2007
Agenda

 Overview: Check and Deposit Fraud – How Has It Changed? How
  Does It Remain the Same?
 Determining Your ―False Positive Rate‖ & ―Fraud Detection Rate‖
 FraudLink System Reports: Data Collection & Analysis
 Fraud Detection/Fraud Losses: Data Collection & Analysis
 ―Fraud Profile‖: Identifying Current & Emerging Fraud Activity
 Risk-Based Re-Calibration: Targeting Your ―Fraud Profile‖
Overview: Check and Deposit Fraud –
How has it changed? How does it remain the same?
 Back in the ―old days,‖ remember when…
   A $1000 fraud loss was catastrophic
   No automated way to detect in-clearing check fraud
   Two types of deposit fraud; new account fraud and kiting
   The fraudster had to come into the bank to commit fraud
   Depended on new account reps. and tellers to detect fraud
 Fraud today…
   A $10,000 fraud loss might be catastrophic?????
   FraudLink On-Us in-clearing check fraud detection
   FraudLink Deposit & Kite detecting deposit fraud schemes
   Fraudster doesn‘t have to enter bank to commit fraud
   Depend on new account reps and tellers to sell, sell, sell…
Overview: Check and Deposit Fraud –
How has it changed? How does it remain the same?

 Fundamentally check fraud     The playing field has just
  has not changed                gotten bigger
   Checks are still…             Professionals
      Lost or stolen             Amateurs
      Forged                     Victim Customers
      Counterfeited              Electronic Transaction – ACH
                                   & Debit Card
      Purchases
      Deposit fraud schemes
      Teller cashed
      Paid in-clearing
Overview: Check and Deposit Fraud –
How has it changed? How does it remains the same?
 Challenges
   Check Losses Highest Among All Payments Channels
   < Check Volume = > Check Fraud?????
     Federal Reserve Payments Study -
        Check Volume: 2000 – 41.9B & 2003 – 36.7B = ↓12.4%
        Electronic Payments: 2000 – 30.6B & 2003 – 44.5B = ↑44.5%
     ABA Fraud Survey
        Fraud Cases: 1999 – 447G; 2001 – 600G & 2003 – 616G
        Attempted Check Fraud: 1999 - $2.2B; 2001 - $4.3B & 2003 - $5.5B
        Losses: 1999 - $679M; 2001 - $698M & 2003 - $677M
Overview: Check and Deposit Fraud –
How has it changed? How does it remains the same?
 Challenges
   FinCEN SAR Reporting - (Check Fraud, Kiting & Counterfeit Checks
    Only)
     1999 – 27,682; 2001 – 43,501; 2003 – 61,611; 2006 – 124,905
   Association of Financial Professionals (AFP) 2007 Payments Fraud
    Survey
     Check fraud is increasing despite check volume decline
         Check Fraud                  93%
         ACH Debit Fraud              35%
         Consumer Credit Card Fraud   17%
         Corporate Purchasing Card    14%
         Consumer Debit Card Fraud    5%
         ACH Credit Fraud             4%
         Wire Transfer Fraud          3%
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


 FraudLink On-Us ―False Positive Rate‖
     FraudLink On-Us Suspects Deemed Good ÷ Total FraudLink On-
      Us Suspects = False Positive Rate
        1,068,000 Suspect Items - 1,360 Fraud Items = 1,066,640
         Good Suspects
        1,066,640 Goods ÷ 1,068,000 Total Suspects = 99.8% False
         Positive Rate
     Loss Avoidance Total: $4.4M – Charge Off Total : $400K
     Daily Average Suspect Volume: 4,240/5 FTE = 850 Items/FTE
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


   FraudLink On-Us ―False Positive Ratio‖
     FraudLink On-Us Suspect Items ÷ FraudLink On-Us Items
      Detected = False Positive Ratio
        1,068,000 Suspect Items ÷ 1,360 Fraud Items = 785:1 Ratio
     One Fraud Item for Every 785 Suspects
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


 FraudLink Deposit ―False Positive Rate‖
      FraudLink Deposit Suspects Deemed Good ÷ Total FraudLink
       Deposit Frauds = False Positive Rate
         156,200 Suspect Accounts - 560 Deposit Frauds = 155,640
          Good Accounts
         155,640 Goods ÷ 156,200 Total Suspects = 99.7% False
          Positive Rate
      Loss Avoidance Total: $19M – Charge Off Total: $2.2M
      Daily Average Suspect Volume:620/5 FTE = 125 Accounts/FTE
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


   FraudLink Deposit ―False Positive Ratio‖
     FraudLink Deposit Suspect Accounts ÷ FraudLink Deposit Fraud
      Accounts = False Positive Ratio
        156,200 Suspect Accounts ÷ 560 Deposit Fraud Accounts =
         278:1 Ratio
     One Deposit Fraud for Every 278 Suspect Accounts
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


 FraudLink On-Us ―Fraud Detection Rate‖ Dollars
     On-Us Fraud Dollars Charged Off + FraudLink On-Us Fraud
      Dollars Detected = Total On-Us Check Fraud Dollars Exposure
        $400K Charged Off + $4.4 Detected = $4.8M Total Fraud
         Exposure
     FraudLink On-Us Fraud Dollars Detected ÷ Total Dollars Exposure
      = Fraud Dollars Detection Rate
        $4.4M Detected ÷ $4.8 Exposure = 92% Fraud Dollars
         Detection Rate
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


   FraudLink On-Us ―Fraud Detection Rate‖ Items
     On-Us Fraud Items Charged Off + FraudLink On-Us Fraud items
      Detected = Total On-Us Check Fraud Items Exposure
        670 items Charged Off + 1,360 items Detected = 2,030 Total
        Items Exposure
     FraudLink On-Us Items Detected ÷ Total Items Exposure = Fraud
      Items Detection Rate
        1,360 Detected ÷ 2,030 Exposure = 70% Fraud Items
         Detection Rate
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


 FraudLink Deposit ―Fraud Detection Rate‖ Dollars
     Deposit Fraud Dollars Charged Off + FraudLink Deposit Dollars
      Detected = Total Deposit Fraud Dollars Exposure
        $2.2M Charged Off + $19M Detected = $21.2M Total Fraud
         Exposure
     FraudLink Deposit Dollars Detected ÷ Total Dollars Exposure =
      Fraud Detection Rate
        $19M Detected ÷ $21.2 Exposure = 90% Fraud Detection
         Rate
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖


 FraudLink Deposit ―Fraud Detection Rate‖ Accounts
     Deposit Fraud Accounts Charged Off + FraudLink Deposit
      Accounts Detected = Total Deposit Fraud Accounts Exposure
        250 Accounts Charged Off + 560 Accounts Detected = 810
         Total Accounts Exposure
     FraudLink Deposit Accounts Detected ÷ Total Accounts Exposure
      = Fraud Detection Rate
        560 Detected ÷ 810 Exposure = 69% Fraud Account
         Detection Rate
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖
 ―Justifiable False Positive Rate‖
   How many suspects are you willing to review to catch fraud?
   Do you know?
   Do you care?
 Choose a Strategy
   Operational Objective?
      Reduce cost/staff
      Maintain current budget and improve detection
      Improve budget and improve detection
      Improve detection and increase cost
      Reduce cost per suspect
      Control daily volume
      Status Quo
Determining Your ―False Positive Rate‖ &
―Fraud Detection Rate‖
   Baseline measurements?
      Average number of suspects per day
      Average number of false positives
      Average number of frauds observed per period
      Average number of ―false negatives‖ observed per period
         ―False negative‖ losses that failed to appear on Suspect Report
      Detection rate observed per period
   No ―One Size Fits All‖ Solution
   Decision up to each individual bank
FraudLink System Reports:
Data Collection & Analysis
 Understanding FraudLink Suspect Distribution
   FraudLink On-Us Back Office Summary Report
     Produces Daily Reports
        Bank
        Group
        Account Type
        Reason/Rule
        Number Checks/Accounts
        Grand Total
     Average Suspect Activity
        Observe the distribution of Suspects across all Account Types and
         Reason/Rule
FraudLink System Reports:
Data Collection & Analysis

     DATE: 01/16/2002 08:37
    DATE: 01/16/2002 08:37                 CARREKER FRAUDLINK ON-US FRAUD DETECTION SYSTEM
                                          CARREKER FRAUDLINK ON-US FRAUD DETECTION SYSTEM                             A16RPT04
                                                                                                                     A16RPT04
     POSTED DATE    01/15/2002
    POSTED DATE : :01/15/2002           BACK OFFICE SUMMARY REPORT CONTAINING ITEMS FROM ALL SOURCES
                                       BACK OFFICE SUMMARY REPORT CONTAINING ITEMS FROM ALL SOURCES                   PAGE 12
                                                                                                                     PAGE 12

                                                                    GRAND TOTALS
                                                                   GRAND TOTALS

     REASON
    REASON                                           CHECKS
                                                    CHECKS        ACCOUNTS
                                                                 ACCOUNTS             AMOUNT
                                                                                     AMOUNT


    DUPLICATE SERIAL NUMBER
   DUPLICATE SERIAL NUMBER                               14
                                                        14              88        $18,302.02
                                                                                 $18,302.02
    SERIAL NUMBER OUT OF RANGE
   SERIAL NUMBER OUT OF RANGE                           102
                                                       102              65
                                                                       65        $371,305.26
                                                                                $371,305.26
    AMOUNT GREATER THAN AVERAGE
   AMOUNT GREATER THAN AVERAGE                           34
                                                        34              29
                                                                       29      $1,976,687.96
                                                                              $1,976,687.96
    AMOUNT EXCEEDS LARGEST ON FILE
   AMOUNT EXCEEDS LARGEST ON FILE                        21
                                                        21              19
                                                                       19        $412,938.72
                                                                                $412,938.72
    NO HISTORY FOR ACCOUNT
   NO HISTORY FOR ACCOUNT                                44             44         $3,528.50
                                                                                  $3,528.50
    NO HISTORY FOR NEW ACCOUNT
   NO HISTORY FOR NEW ACCOUNT                            22             11         $4,326.42
                                                                                  $4,326.42
    MISSING SERIAL NUMBER
   MISSING SERIAL NUMBER                                 17
                                                        17              99        $31,177.61
                                                                                 $31,177.61
    LOW DOLLAR CHECK PULL
   LOW DOLLAR CHECK PULL                                148
                                                       148             148
                                                                      148        $384,295.54
                                                                                $384,295.54
    LOWEST CHECK NUMBER ON FILE
   LOWEST CHECK NUMBER ON FILE                           99             88        $12,200.17
                                                                                 $12,200.17
    VELOCITY BACK OFFICE
   VELOCITY BACK OFFICE                                  44             22         $1,540.80
                                                                                  $1,540.80
    BRANCH VELOCITY
   BRANCH VELOCITY                                       00             00             $0.00
                                                                                      $0.00
    BRANCH DUPLICATE SERIAL
   BRANCH DUPLICATE SERIAL                               10
                                                        10              33         $8,339.44
                                                                                  $8,339.44
    DUPLICATE AMOUNT
   DUPLICATE AMOUNT                                      24
                                                        24              88        $14,765.21
                                                                                 $14,765.21
    SERIAL NUMBER IN NEW CHECK RANGE
   SERIAL NUMBER IN NEW CHECK RANGE                      00             00             $0.00
                                                                                      $0.00
    HIGH DOLLAR
   HIGH DOLLAR                                           98
                                                        98              69
                                                                       69      $3,463,688.35
                                                                              $3,463,688.35
    DUPLICATE SERIAL AND AMOUNT
   DUPLICATE SERIAL AND AMOUNT                           00             00             $0.00
                                                                                      $0.00
    EXCEEDED DOLLAR THRESHOLD
   EXCEEDED DOLLAR THRESHOLD                             00             00             $0.00
                                                                                      $0.00
    PAYEE VELOCITY
   PAYEE VELOCITY                                        00             00             $0.00
                                                                                      $0.00
    SUSPECT ONLY ITEMS
   SUSPECT ONLY ITEMS                                   204
                                                       204             117
                                                                      117      $3,315,609.02
                                                                              $3,315,609.02
    COMPANION ONLY ITEMS
   COMPANION ONLY ITEMS                                 108
                                                       108             108
                                                                      108         $29,477.99
                                                                                 $29,477.99
    SUSPECT COMPANION ITEMS
   SUSPECT COMPANION ITEMS                               40
                                                        40              40
                                                                       40        $354,817.55
                                                                                $354,817.55


    FRAUD ANALYSIS HAS FLAGGED
   FRAUD ANALYSIS HAS FLAGGED     352 CHECKS FOR
                                 352 CHECKS FOR       153 ACCOUNTS WITH    TOTAL VALUE OF
                                                     153 ACCOUNTS WITH A ATOTAL VALUE OF     3,699,904.56 FOR THIS DAYS WORK
                                                                                            3,699,904.56 FOR THIS DAYS WORK
FraudLink System Reports:
Data Collection & Analysis
 Understanding FraudLink Suspect Distribution
   FraudLink Deposit Daily ReCap Report
     Produces Daily Reports:
         Bank
         Application
         Account Type
         Rule/Reason
         Account Period
         Grand Total
     Average Suspect Activity
         Observe the distribution of Suspects across all Account Types,
          Rule/Reason and Account Period
FraudLink System Reports:
Data Collection & Analysis
Fraud Detection/Fraud Losses:
Data Collection & Analysis
 Collect, Sort and Stratify Your On-Us Detection and Loss Data
   On-Us Fraud Analysis
      Geographical Risk
      Product Risk
      FraudLink Suspect Rule
      Loss Avoidance Amount
      Loss Amount – FraudLink Suspect Y/N
      Return Reason

 Understand Your Entire Risk Exposure
   What‘s working
   What‘s not working
   Where changes are needed
Fraud Detection/Fraud Losses:
Data Collection & Analysis
 Collect, Sort and Stratify Your Deposit Detection and Loss Data
   Deposit Fraud Analysis
      Geographical Risk
      Product Risk
      Age of Account Risk – FraudLink Account Periods
      FraudLink Suspect Rule
      Deposit Amount
      Loss Avoidance Amount
      Loss Amount – FraudLink Suspect Y/N
      RDI Reason
 Understand Your Entire Risk Exposure
   What‘s working
   What‘s not working
   Where changes are needed
―Fraud Profile‖: Identifying Current
& Emerging Fraud Activity
 ―Fraud Profile‖
   Current Trends – Commonality
      Common Fraud Amounts
      Common Bank Products
      Common Geographic's
      Common Account Age
      Common Detection & Failures
      In-clearing vs. Teller Cashed
   Emerging Fraud – Un-Commonality
      New Fraud Amounts
      Fraud Below Current FraudLink System Settings
      New Bank Products
      New Geographic‘s
      New Account Ages
      Fraud Moving From Paper to other Delivery Channels
Risk-Based Re-Calibration:
Targeting Your ―Fraud Profile‖
 Where is the Fraud Risk?
   FraudLink On-Us
      Product – Commercial DDA
      Amounts - $375.00 - $998.00 & $4,800 - $9,900
      Detection – Rule 2 Serial Number Out of Range-83% & Rule 1 Duplicate
       Serial Number-12%
      Determine ―False Negatives‖
 Where isn‘t the Fraud Risk?
   FraudLink On-Us
      Product – Senior 50+ DDA & MMDDA
      Amounts < $374.00
      Detection – Rule 3 Amount Greater Than Average & Rule 13 Duplicate
       Check Amounts
      Determine ―False Negatives‖
Risk-Based Re-Calibration:
Targeting Your ―Fraud Profile‖
 Where is the Fraud Risk?
   FraudLink Deposit
      Product – Free Personal DDA & Internet Free Personal DDA
      Amounts – Account Period 1: $2,000 - $5,500 Account Period 3: $35,000
       - $425,000
      Detection – Rule 1 Daily Total Above Average-88% & Rule 5 Invalid
       Routing & Transit Number-9%
      Region – 1 & 2
 Where isn‘t the Fraud Risk?
   FraudLink Deposit
      Product – Commercial DDA & Public Funds Accounts
      Amounts < $1,000
      Detection – Rule 8 Deposit Velocity Exceeds Average & Rule 6 Duplicate
       Items Amounts
      Region – 6 & 8
Risk-Based Re-Calibration:
Targeting Your ―Fraud Profile‖
 What Have We Learned?
   False Positive Rate
   False Negative Rate
   Fraud Detection Rate
   ―Justifiable False Positive‖
   FraudLink Suspect Volume Distribution
   Fraud Exposure Distribution
   Fraud Profile
      Common Fraud Trends
      Emerging Fraud Trends
   Where Your Fraud Is
Risk-Based Re-Calibration:
Targeting Your ―Fraud Profile‖
 Re-Calibration
   Set and balance Rules and Parameters
      Target highest Fraud Risk activity
      Generate More Suspects that Provide the Greatest Value
      Generate Less Suspects Where Fraud is Least Likely
   Results
      More Fraud Detection
      Less False Negatives
      Less False Positives
      Happier Employees
   What You Might Find…
      A need to generate more Suspects than current staff can handle
      Business case for added staff with significant payback
      Need for full-time Business Analyst to collect and analyze data and
       conduct re-calibration testing
Questions?

 Q: I don’t have the staff to do all of this, is there an automated
  way to collect this data?
 A: Yes, CORE Workflow Manager & Syfact Case Management
  System


 Q: Can Carreker/CheckFree help?
 A: Yes, we provide consulting and re-calibration services, check
  with your Account Representative.


                     Now Your Questions!
Closing Remarks



Jeff Botari, CheckFree

								
To top