Digital Analysis to Detect Fraud and Abuse

1
Digital Analysis to Detect Fraud and Abuse
Dave Coderre CAATS Ottawa, Canada ISACA June 30,
2004
2
Digital Analysis Outline

• Approach to Fraud/Abuse Detection
• Advanced Analytical Techniques

3
Proactive Fraud Detection

• Build a Profile of Potential Frauds which can
  then be Tested
• Develop a Fraud Monitoring/Detection Plan to Look
  for Possible Indicators or Symptoms of Fraud
• Investigate, Drill Down and Verify Patterns
  Emerging from the Analysis
• Use Continual Auditing to Automate the Fraud
  Prevention and Detection Process

4
Fraud Monitoring/Detection Plan

• WHY - the objective of the analysis
• WHAT - the expected results
• WHERE - source of the data
• HOW - analyses to be performed
• WHAT NEXT - follow-up
• Corroborating Proof
• Additional Analysis
• Development of ACL Batches

5
Fraud Schemes - Approach

• Fraudsters Perspective
• Who could Benefit?
• How could they be Involved?
• What can they Effect/Manipulate or Control to
  Allow Fraud to Happen?
• Can they Act Alone or is Collusion Required?

6
Fraud Schemes - Approach
Quantity Received

• Source - Receipt File
• Data - Quantity Received Field
• Analysis - Recd Qty gt Ordered Qty
• Follow up - Contracting Officer
• - Receipt Clerk
• - Vendor

7
Summary

• Identify Risks
• Search for Symptoms of Fraud in the Data
• Systematic Review (Ongoing)
• Verify to Source

8
Symptoms of Fraud in the Data

• Known Symptoms
• Unknown Symptoms

• Definite Criteria
• Standard Analysis
• Scheduled Jobs

• Anomalies
• Patterns in the Data

9
Symptom Detection

• Known Symptoms
• Basic Tests
• Filters
• Extraction Routines
• Unknown Symptoms
• Digital Analysis
• Trends in the Data
• Patterns
• Anomalies

10
Credit Card Purchases

• Known Symptoms / Concerns
• Personal purchases

?
11
Credit Card Purchases

• Examples of Known Symptoms / Concerns
• Top 10 Transactions
• Weekend Purchases
• Bypassing Financial Controls
• Vendor Theft of Card
• Hotel Charges on Acquisition Card
• Cash Advances
• Exceeding Transaction Limit

12
Credit Card - Data Analyses

• Basic
• Completeness, Integrity, Duplicates, Gaps
• Cross Tabs (by Card by Std Object)
• Advanced
• Data Profiling
• Statistics and Anomalies
• Variance Analysis
• Ratios - Max / Min and Max / Max2
• Benfords Law
• First, First Two, First Three and Second Digits

13
Data Profiling of a Numeric Field

• Selected Numeric Field
• Statistical Analysis
• Min, Max, Average, Highest and Lowest Values
• Even Amounts
• Multiples of 1,000
• Frequently Used Values
• Multiples of 1,000
• Least/Most Used Categories
• Vendor, Credit Card Number, Contract Officer

14
Data Profiling - Results

• Credit Card Transactions
• Maximum of 84K - much higher than allowed
• 22 Transactions for even multiples of 1,000
  totaling 41,000 - problems with 2 credit cards
• 20.67 was used 170 times
• One Supplier used 5,092 times
• Questionable Suppliers

15
Variance Analysis

• Unit Prices
• Item High_Price High_Price2 Ratio
• 198 101.46 98.91
  1.026
• 773 123.48 57.23
  2.158
• 861 51.23 50.84
  1.008
• 634 26.31 11.63
  2.262
• 992 124.78 124.03
  1.006

16
RATIO ANALYSIS
Variance in Payments - by Vendor
17
Benfords Law
18
Contract/Purchase Order Amounts

• Frequency

First Digit
Symptom of Possible Kickbacks Over Billing
Contract Splitting etc.
19
Bank Deposit Amounts
Frequency
First Digit
20
Advanced Analysis - Results

• Variance Analysis
• Possible problems with Vendor Transactions based
  on Max/Max2 and Max/Min ratios
• wrong vendors and wrong amounts
• Unit Price anomalies for certain Products
• Benfords Law
• Overall agreement with expected values
• Possible problems in details

21
Conclusions

• Benefits of Using Audit Software
• Many auditors already familiar with the software
• Pro-active review of transactions
• Standard tests can be performed across a range of
  industries, locations and systems
• Require little user knowledge of analysis
  techniques
• Review 100 of transactions and focus attention
  on Few questionable items requiring follow up
• Automatic logging capabilities

22
References

• SAS and ISA AICPA standards etc
• Occupation Fraud and Abuse, ACFE, Joseph Wells.
• Fraud Detection Using Data Analysis Techniques
  to Detect Fraud and
• The Fraud Toolkit
• The last two publications are available at
  www.ekaros.ca

23
The End (finally)

• Thank you
• Any Questions ?????
• Dave_Coderre_at_hotmail.com