using data analytics to detect fraud · using data analytics to detect fraud ... fuzzy logic...
TRANSCRIPT
© 2018 Association of Certified Fraud Examiners, Inc.
Using Data Analytics to
Detect Fraud
Fundamental Data Analysis Techniques
© 2018 Association of Certified Fraud Examiners, Inc. 2 of 27
Introduction
▪ In determining types of tests to run, consider:
• The particular fraud risks that are present
• The data available to work with
• The type of predication that exists
▪ Often, techniques are most effective when
used in combination.
© 2018 Association of Certified Fraud Examiners, Inc. 3 of 27
Aging
▪ Analyzing data
based on date
▪ Useful in
examining:
• Accounts
receivable
• Customer
payments
• Accounts payable
• Vendor payments
© 2018 Association of Certified Fraud Examiners, Inc. 4 of 27
Application: Aging
Excel ACL IDEA
▪ Date-based
subtraction
▪ Function
• AGE()
▪ Command• AGE
▪ Functions
• @Age()
• @AgeDateTime()
• @AgeTime()
▪ Command
• Aging
© 2018 Association of Certified Fraud Examiners, Inc. 5 of 27
Applying Filters
▪ Identifies only those
records meeting user-
defined criteria
▪ Used to extract
transactions outside of
expected norm
▪ Can further filter or
analyze results using
additional analysis
techniques
© 2018 Association of Certified Fraud Examiners, Inc. 6 of 27
Application: Filters
Excel ACL IDEA
▪ Advanced filter
▪ Meta-tagging
▪ Filter bar
▪ IF statements
▪ Criteria
© 2018 Association of Certified Fraud Examiners, Inc. 7 of 27
Benchmarking
▪ Comparing a
company’s processes
or performance metrics
to:
• Competitors
• Industry standards
• Historical data
• Budgeted data
© 2018 Association of Certified Fraud Examiners, Inc. 8 of 27
Application: Benchmarking
Excel
▪ Conditional Formatting
▪ Charts
▪ PivotChart
© 2018 Association of Certified Fraud Examiners, Inc. 9 of 27
Compliance Verification
▪ Determines whether
employee
transactions comply
with company policies
▪ Useful in identifying
whether a company
policy needs to be
either revised or
reinforced
© 2018 Association of Certified Fraud Examiners, Inc. 10 of 27
Application: Compliance Verification
Excel ACL IDEA
▪ IF()
▪ IFError()
▪ Expression with
conditions
▪ @If()
▪ @CompIf()
© 2018 Association of Certified Fraud Examiners, Inc. 11 of 27
Duplicate Testing
▪ Identifies transactions with duplicate values
in specified fields:
• Check numbers
• Invoice numbers
• Government identification numbers (e.g., Social
Security numbers)
• Employee or vendor addresses
© 2018 Association of Certified Fraud Examiners, Inc. 12 of 27
Application: Duplicates
Excel ACL IDEA
▪ COUNTIF()
▪ COUNTIFS()
▪ DUPLICATES
command
▪ Duplicate Key
Detection
command
▪ Duplicate Key
Exclusion
command
© 2018 Association of Certified Fraud Examiners, Inc. 13 of 27
Expressions and Equations
▪ Build expressions or equations based on
knowledge and expectations of what should
be in the data:
• Recomputing net payroll amounts based on gross
pay, taxes, and other deductions
• Recalculating amounts charged on invoices based
on unit price and quantity ordered
© 2018 Association of Certified Fraud Examiners, Inc. 14 of 27
Frequently Used Values
▪ Identifying values
that occur with
unexpected
frequency
▪ Red flag of fictitious
transactions
© 2018 Association of Certified Fraud Examiners, Inc. 15 of 27
Application: Frequently Used Values
Excel ACL IDEA
▪ COUNTIF()
▪ COUNTIFS()
▪ Benford’s Law
command
▪ Summarize
command
▪ Benford’s Law
command
▪ Summarization
command
© 2018 Association of Certified Fraud Examiners, Inc. 16 of 27
Fuzzy Logic Matching
▪ Identifies records with similar or potentially
duplicate—though not identical—values:
• First Street, First St., 1st Street, 1st St.
▪ Helps detect fraud in “gray areas” by
reviewing various iterations of data
▪ Can produce an increased number of false
positives
© 2018 Association of Certified Fraud Examiners, Inc. 17 of 27
Application: Fuzzy Logic
Excel ACL IDEA
▪ Normalize,
then compare
▪ Fuzzy Duplicates
command
▪ Normalize, then
compare
▪ Duplicate Key
Fuzzy command
▪ Normalize, then
compare
© 2018 Association of Certified Fraud Examiners, Inc. 18 of 27
Gap Tests
▪ Search for missing items in a series or
sequence of consecutive numbers:
• Check numbers
• Invoice numbers
• Purchase order numbers
• Inventory tags
▪ Search for sequences where none are
expected:
• Social Security numbers
© 2018 Association of Certified Fraud Examiners, Inc. 19 of 27
Application: Gaps
Excel ACL IDEA
▪ Sort, then value
comparison
using IF()
▪ GAPS command ▪ Gap Detection
command
© 2018 Association of Certified Fraud Examiners, Inc. 20 of 27
Graphing
▪ Provides a visual
representation of the
data and can
highlight patterns or
anomalies that might
indicate areas for
further examination
© 2018 Association of Certified Fraud Examiners, Inc. 21 of 27
Identifying Amounts Below a Threshold
▪ Search for patterns
of transactions that
fall just below
approval or review
thresholds.
© 2018 Association of Certified Fraud Examiners, Inc. 22 of 27
Application: Thresholds
Excel ACL IDEA
▪ Value
comparison
using IF()
▪ BETWEEN()
function
▪ Value comparison
using <, >
▪ @Between()
function
▪ Value comparison
using <, >
© 2018 Association of Certified Fraud Examiners, Inc. 23 of 27
Identifying Unusual Dates and Times
▪ Identifies
transactions that
occur during non-
business hours or
employee
vacations
© 2018 Association of Certified Fraud Examiners, Inc. 24 of 27
Application: Unusual Dates and Times
Excel ACL IDEA
▪ Value
comparison
using IF()
▪ NOT BETWEEN()
function
▪ Value comparison
using <, >
▪ .NOT. @BetweenDate()
function
▪ .NOT. @BetweenTime()
function
▪ Value comparison using
<, >
© 2018 Association of Certified Fraud Examiners, Inc. 25 of 27
Join/Relate
▪ Combines specified fields from two different
files into a single file using key fields
▪ Looks for matches or discrepancies between
the files
© 2018 Association of Certified Fraud Examiners, Inc. 26 of 27
Application: Join/Relate
Excel ACL IDEA
▪ VLOOKUP()
▪ HLOOKUP()
▪ INDEX()
▪ JOIN command
▪ RELATE command
▪ Join command
▪ Visual Connector
command
© 2018 Association of Certified Fraud Examiners, Inc. 27 of 27
Pivot Tables
▪ Interactive data summarization tool used to
sort, count, total, or give the average of
specified data in a spreadsheet
▪ Can perform the filter and sort functions
within the pivot table
▪ Helpful way to see the “big picture” of the
data
© 2018 Association of Certified Fraud Examiners, Inc. 28 of 27
Application: Pivot Tables
Excel ACL IDEA
▪ PivotTable
▪ PowerPivot
▪ Cross-Tabulate
command
▪ Pivot Table
command
© 2018 Association of Certified Fraud Examiners, Inc. 29 of 27
Round-Dollar Payments
▪ Most real-world cash
transactions do not
occur in simple round
numbers.
▪ Unusual amounts or
regular occurrences
of round-dollar
payments can be red
flags of fraud.
© 2018 Association of Certified Fraud Examiners, Inc. 30 of 27
Application: Round-Dollar Payments
Excel ACL IDEA
▪ MOD() ▪ MOD() function
▪ FIND() function
▪ @IsInI function
© 2018 Association of Certified Fraud Examiners, Inc. 31 of 27
Sort/Index
▪ Arranges the data in
ascending or
descending order
based on one or
more specified key
field(s)
• Alphabetically
• Numerically
• Chronologically
© 2018 Association of Certified Fraud Examiners, Inc. 32 of 27
Stratification
Invoice amount Count Percent of total Total amount
Less than $1,000 87 10.5% $ 66,078.24
$1,001–$5,000 196 23.6% $ 782,089.00
$5,001–$10,000 359 43.2% $ 2,515,940.21
$10,001–$20,000 102 12.3% $ 1,427,527.74
$20,001–$50,000 68 8.2% $ 2,022,600.16
Over $50,000 19 2.3% $ 1,298,874.96
Total: 831 100% $ 8,113,110.31
© 2018 Association of Certified Fraud Examiners, Inc. 33 of 27
Application: Stratification
Excel ACL IDEA
▪ SUMIFS() and
COUNTIFS()
▪ STRATIFY
command
▪ Stratification
command
© 2018 Association of Certified Fraud Examiners, Inc. 34 of 27
Summarization
▪ Counting the number of records with
common values within a specified field
State Count
Texas 704
Florida 362
Georgia 12
New Hampshire 1
Virginia 7
Total: 1,086