Data Analytics and Internal Audits_Final

Report
Data Analytics & Internal Audits
IIA, Boise Chapter
March 2014
How Did I Get Here?
• Student Auditor at University Internal Audit Department
• Accounting Degree
• Protiviti Internal Controls Testing
• Protiviti Data Analytics Team
1
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Objectives
2
1)
Encourage the use of data analytics in existing internal audit reviews
2)
Share fundamental knowledge to successfully incorporate data analytics into audit work programs
3)
Introduce Continuous Monitoring concepts
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Establishing a Common Language
Data Analytics
Techniques used by auditors or management to
manipulate large volumes of data to provide meaningful
insight into activities occurring throughout the business.
Continuous Auditing
Method used by auditors to perform audit-related
activities on a continuous basis. Activities range from
continuous control assessment to continuous risk
assessment.
Continuous Monitoring
Process that management puts in place to ensure that
its policies and procedures are adhered to, and that
business processes are operating effectively.
Continuous monitoring typically involves automated
continuous testing of ALL transactions within a given
business process area against a suite of controls.
3
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Why Data Analytics?
Data Analytics Opportunities and Value Proposition
1.
Optimize the return on your existing data investments
2.
Provide insights to help pinpoint new opportunities and improve operational efficiencies and visibility across the
organization
3.
Enable faster problem-solving and decision-making at the strategic, operational and tactical levels
4.
Find hidden meaning – patterns, trends, relationships – in your data
5.
Deliver intelligence to the field in real-time
6.
Mitigate the risk of fraud
7.
Improve your company's competitive advantage
8.
Achieve or validate compliance with government and regulatory guidelines
9.
Confirm existing controls are working properly
10. Reconcile data across disparate systems
5
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Example Internal Audit Case Studies
Case Study #1
6
Organization
Need
Our client, a grocery chain, systematically transmitted multiple price change files per day originating from three
different corporate systems to each of their 90 store locations. Store systems relied on manual processes for price
files to be imported and applied at the register. Compliance with corporate supplied prices had never been tested.
Solution
To test compliance with corporate supplied pricing, we sought to analyze 100% of the transactions in each store
during the audit period to determine if the corporate supplied price had been appropriately applied. This POS data
extract totaled 300 million records for the one year audit period.
After interviewing corporate and store-level stakeholders, Protiviti identified the relevant price system data sources
(Ad, DSD, Pricing Dept.) and the rules governing price hierarchy. Prices were systematically applied based on the
key characteristics of SKU, store, date, and price type.
Result: Underpricing on sales of $14M. Results of this analysis were validated and reviewed with management.
Root cause was identified as store department managers manually overriding corporate suggested prices to sell
excess inventory purchases.
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Advantages of Data Analytics
1.
Testing of full populations

No need to extrapolate sample results

Drill down to individual transactions
2.
Efficient & Repeatable

3.
Undisputed results

7
New understanding can be incorporated and the calculations rerun
With agreement on inputs and model
4.
Enhanced risk assessment for audit area selection
5.
Targeted samples for testing
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Tools for Internal Audit Data Analytics
Comparative Tool Features and Efficiencies
8
MS Excel
MS Access
ACL
SQL/Oracle
Complexity
Low
Low
Medium
Very High
Capacity
Up to 1.04 Million
Records only
Up to 2 GB
Performance dips as the data
volume increases. Good
performance up to 1 GB of data
No Limit
Skill Sets
Limited availability
but easy to train
Limited availability
but easy to train
Limited availability
but easy to train
Limited availability
Analysis Time
Quick
Quick
Quick
High
Cost
Very Low
Very Low
Medium
Very High
Database
Security
Low
Medium
Medium
Very High
Calculation
Integrity
Low
High
High
High
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Phases of a Successful Internal Audit
Using Data Analytics
These phases are not unique to data analytics focused internal audits. They can easily be integrated into
the existing framework of internal audit work programs.
Scoping
9
Data Request
Data Integration
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Data Analysis
Findings
Validation &
Reporting
Methodology Phases
Scoping
Data Request
Data Integration
Data Analysis
Objectives
•
•
•
Perform existing scoping activities to prepare for the review
Understand areas where Data Analytics would be most useful
Identify relevant systems and data sets
Key Questions
•
•
•
•
•
•
10
What data exists?
Can the subject of our audit be readily observed in existing data sets?
Does the auditee agree that Data Analytics can lead to an accurate answer?
Are multiple data sets required? Can these be tied together?
What thresholds, characteristics, etc. constitute an exception?
What business processes support the generation of the data?
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Findings
Validation &
Reporting
Example Internal Audit Case Studies
Case Study #2 – Importance of Scoping
11
Organization
Need
A large telecom client maintains a large contingent workforce supplied by a third party who themselves subcontract
to 135 staffing firms. Our client desired to audit contingent workforce invoice details to determine if contractual terms
including appropriate hourly rates were being observed.
Result
Client insisted that the contract party not be engaged during the scoping phase. Relevant data sets existed in a
Vendor Management System to which our client had access. After over 100 hours of analysis, data analysis results
showed millions of dollars of overbilling due to excessive hourly rates being applied. Upon review with the staffing
firm, it was learned that data extracts used did not reflect approved waivers to hourly rates and that these waivers, in
fact, did not exist in a system of record. All findings were cleared.
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Methodology Phases
Scoping
Data Request
Data Integration
Objectives
•
•
Obtain data necessary to complete the audit
Minimize need to re-request data
Key Questions
•
•
•
•
12
Who can provide the data?
‒ Business Unit: May not be able to change filters, output
fields, or file format
‒ IT: May take longer, may not understand “business” meaning
of data fields, will give you exactly what you ask for
Do we understand what data is available?
‒ More vs. Less – Fields, filters, etc.
How do we want the data?
How can we receive the data?
‒ Size & Security?
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Data Analysis
Findings
Validation &
Reporting
Methodology Phases
Scoping
Data Request
Data Integration
Objectives
•
•
Prepare data received for analysis
Validate completeness of data received and imported
Key Tasks
•
•
•
•
•
13
Select the technology best suited for the analytics selected.
Things to consider:
‒ Volume and type of data
‒ Complexity of the modeling
‒ Accessibility of certain technologies
‒ In-house expertise and skills
Importing data into analysis tool – data types are important!
Validating completeness using record counts, system report,
GL, etc.
Determine how disparate data will be integrated (“joined”)
Prepare any transformations or mappings
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Data Analysis
Findings
Validation &
Reporting
Example Internal Audit Case Studies
Case Study #3 – Validating Completeness
Organization
Need
Many of our clients have licensing agreements requiring the contract party to make periodic royalty payments based
on sales activity. Our clients exercise the audit clause of these agreements to validate the completeness and
accuracy of these royalty payments.
Solution
Protiviti employs a top-down approach to these audits where a complete population of the licensee’s sales detail is
requested. Completeness of the sales data provided is validated by agreeing to audited financial statements. Once
completeness has been established, we can isolate those sales which are subject to the agreement and perform a
full recreation of payment obligations.
Audited Financial Statements
Business Unit / Segment sales
Licensee
Audited Fin.
Stmts.
All Licensee Invoices at
Line Level
Business Unit
/ Segment
Sales
Sales with IP
Content
IP
Components
14
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Sales with IP
Content
Transformations and Mappings
Some “housekeeping” may be required to make data usable for analysis.
Transformations:
Mappings
15
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Methodology Phases
Scoping
Data Request
Data Integration
Data Analysis
Objectives
•
Perform desired analysis
Key Tasks
•
•
•
16
Be Logical
Be Creative
Be Evolving
‒ Build test scripts
‒ Validate the accuracy of the scripts and other applicable KPIs or metrics
‒ Confirm test scripts are identifying the intended results
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Findings
Validation &
Reporting
Be Logical
Build analysis step-by-step. Royalty example:
17
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Example Internal Audit Case Studies
Case Study #4 – Time & Expense Analysis
Organization
Need
Solution
18
A technology client terminated a field rep for submitting fraudulent and abusive expense reports. Our client desired
to identify if any other field reps within his department were utilizing similar schemes in travel expense
reimbursement.
Utilized the expense categories of submitted expense reports in performing analysis to identify fraud or abuse:
• Incompatible expense reimbursements (fuel without a rental car, etc.)
• Excessive reimbursements (More than 3 meals a day)
• Ratio analysis (Lodging cost per travel night)
• Simple descriptive spend summaries (Total yearly reimbursements for hotel, car, meal, etc.)
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Example Internal Audit Case Studies
Case Study #5 – Pattern Analysis
19
Organization
Need
A hospital system terminated two employees who had gotten access to patient records with the intention of possibly
using the information to file false tax returns. Our client wanted to identify if any other employees were using the
same scheme to steal client information.
Solution
The hospital system identified that the terminated employees had been using their appropriate access to admittance
systems to steal patient information. Our client identified the objects (“screens”) within that system which contained
sensitive client information. Using the log records from that system, we identified a pattern of use consistent with a
user stealing PII from the admittance system.
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Historical and Descriptive Analysis
Although organizations will generally solve customized problems, recognized data and statistical
analysis techniques are the basis for solving those problems.
Business Performance identifies key ratios and metrics that track how business operations and processes are
functioning:
• Financial, operational metrics
• Scorecards and KPIs
Clustering data classifies data variables into similar data types for easy visualization and identification of problem areas:
• Grouping
• Deciles, quartiles, percentiles or other rank order measurements
• Stratifications
• Geographical, product, business unit or other segmentation criteria
Trending creates visual displays of the data over time showing information such as:
• Cyclicality (e.g., time series analysis, by day of the week, month, season, etc.)
• Event driven results
• Abnormalities
Descriptive statistical analysis brings the science of statistics into
data analysis:
• Distributions
• Outliers and standard deviation measurements (z-scores, etc.)
• Correlations and regression
• Volatility
20
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Methodology Phases
Scoping
Data Request
Data Integration
Objectives
•
Deliver quality and actionable audit results
Key Tasks
•
•
•
•
•
21
Validate approach with auditee
Validate any exceptions with auditee
Update data analysis and incorporate lessons learned as appropriate
Investigate variances, offending transactions, etc.
‒ Categorize variances if possible
Finalize audit report
‒ 100% of the population tested
‒ Undisputed results
‒ Targeted follow-up
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Data Analysis
Findings
Validation &
Reporting
Transitioning Data Analytics to
Continuous Monitoring
Continuous Monitoring is an outgrowth of the Data Analytics phases discussed. Data Analytics are
formalized, productionalized, and scheduled to allow for repeatable auditing and monitoring. Key
characteristics of these phases as they apply to Continuous Monitoring are summarized below.
Scoping
• Inventory potential
analytics areas at
a macro level
• Select analytics
area and identify
detailed
tests/metrics
• Define
Requirements for
the individual tests
or metrics
22
Data Request
• Formalize data
request
• Schedule periodic
supply of data
• Design transfer
protocols for data
extracts
Data Integration
• Automate periodic
refresh or load of
data request
• Automate data
validation
procedures
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Data Analysis
• Design analysis to
be performed
• Make analysis
repeatable
Findings
Validation &
Reporting
• Make results of
analysis available
to end users in
near real time
Example Internal Audit Case Studies
Case Study #6 – Continuous Monitoring
Organization
Need
Solution
23
A large retailer wanted to monitor its point-of-sale (POS) transactions trying to identify fraud and abuse by
associates at the register.
The following was performed as the solution to the need:
1. Established daily ETL procedures to obtain data from the POS systems for approximately 600 stores.
2. Created a data warehouse and supporting data models to maintain and store data into perpetuity and drive
dashboard performance.
3. Created approximately 20 red-flag algorithms that monitored transaction activity on a daily basis. Transactions
that are flagged by the algorithms are systematically placed in an Excel file and emailed directly to the divisional
personnel responsible for loss prevention.
4. Created a web-based dashboard and score-card solution that identifies outliers and gives end-users the ability to
perform research (see trends, investigate transactions, and extract data to Excel for ease-of-use).
Below are sample screenshots of the email alerts and scorecard/dashboard.
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Questions?
Contact Information
Reed Belliston
Manager – Internal Audit and Financial Advisory
[email protected]
P: 801.401.8166
Visit us the web at:
www.protiviti.com
Access KnowledgeLeader ™
Protiviti’s Subscription Audit Resources Site
Join our Community on LinkedIn
www.linkedin.com/company/protiviti
Follow us on Twitter
@Protiviti (www.twitter.com/protiviti)
Become a Fan on Facebook
www.facebook.com/home.php?#/Protiviti
Visit the Protiviti YouTube Channel
www.youtube.com/protivitiinc
Listen to the Protiviti Powerful Insights Podcast
Available at Protiviti.com or Subscribe on iTunes
Visit our Pinterest board
http://pinterest.com/protiviti/
25
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.
Confidentiality Statement and Restriction for Use
This document contains confidential material proprietary to Protiviti Inc. ("Protiviti"), a wholly-owned subsidiary of Robert Half International Inc. ("RHI"). RHI is a publicly-traded
company and as such, the materials, information, ideas, and concepts contained herein are non-public, should be used solely and exclusively to evaluate the capabilities of Protiviti
to provide assistance to your Company, and should not be used in any inappropriate manner or in violation of applicable securities laws. The contents are intended for the use of
your Company and may not be distributed to third parties.
26
© 2014 Protiviti Inc.
CONFIDENTIAL: This document is for presentation purposes only and may not be copied nor distributed to another third party.

similar documents