21/12/2024

CHAPTER 01
An Overview of Data
Analytics in Accounting

1
Prepared by Nguyen Huu [email protected]

Objectives

• Define Data Analytics.

• Understand why Data Analytics matters to business.
• Explain why Data Analytics matters to accountants.
• Describe the Data Analytics Process using the IMPACT cycle.
• Describe the skills needed by accountants.
• Explain how to IMPACT model may be used to address a specific
business questions.

2
Prepared by Nguyen Huu [email protected]

Contents

• What is data analytics?

• How does data analytics affect business?
• How does data analytics affect accounting?
• The data analytics process using the IMPACT cycle
• Hands-on example of the IMPACT model
• What data analytic skills do accountants need?

3
Prepared by Nguyen Huu [email protected]

1
 21/12/2024

What is data analytics?

• Data Analytics is the process of evaluating data with the purpose of

drawing conclusions to address business questions. Effective Data
Analytics provides a way to search through large structured and
unstructured data to identify unknown patterns or relationships.

While sometimes Data Analytics and Big Data are terms used interchangeably,
we will use the term Data Analytics throughout and focus on the ability to turn
data into knowledge and knowledge into value.

• Big Data refers to datasets which are too large and complex to be
analyzed traditionally.

4
Prepared by Nguyen Huu [email protected]

What is data analytics?

• Data volume refers to the amount of data created and stored by

an organization.
• Data velocity refers to the pace at which data is created and
stored.
• Data variety refers to the different forms data can take.
• Data veracity refers to the quality or trustworthiness of data.

5
Prepared by Nguyen Huu [email protected]

How does data analytics affect business?

• 85% of CEOs put a high value on Data Analytics.

• 86% of CEOs place data mining and analysis as the second-most
important strategic technology.
• Business analytics tops CEO’s list of priorities.
• Data Analytics could generate up to $2 trillion in value per year.

6
Prepared by Nguyen Huu [email protected]

2
 21/12/2024

How does data analytics affect accounting?

7
Prepared by Nguyen Huu [email protected]

How does data analytics affect accounting?

8
Prepared by Nguyen Huu [email protected]

How does data analytics affect accounting?

Reviewing 100% of a Identifying trends and

dataset rather than just a irregularities in financial
sample data • Enhancing audit quality.
• Enabling enhanced audits,
expanded services, and
For
added value to clients.
Auditing
• External auditors stay
Spending more time on engaged beyond the audit.
Automating the
more complex tasks that
detection of anomalies
require judgment and
or potential fraud
expertise

9
Prepared by Nguyen Huu [email protected]

3
 21/12/2024

How does data analytics affect accounting?

For management
accounting • Enhancing cost analysis.
• Budgeting • Enabling better decision-making.
• Performance • Enabling better forecasting, budgeting,
evaluation production, and sales.
• Cost management
• Forecasting

10
Prepared by Nguyen Huu [email protected]

10

How does data analytics affect accounting?

Estimates
 Accountants make better estimates of
For collectability, write-downs, etc.
financial  Managers better understand the business
accounting environment through social media.
 Analysts identify risks and opportunities
through analysis of Internet searches.
Valuations

11
Prepared by Nguyen Huu [email protected]

11

How does data analytics affect accounting?

• For tax
 Companies develop sophisticated tax planning strategies.
 Managers understand tax consequences of international transactions,
investment, mergers and acquisitions.
 The organization understands tax tables and other tax data to aid
compliance.

12
Prepared by Nguyen Huu [email protected]

12

4
 21/12/2024

The data analytics process using the IMPACT cycle

• Identify the questions.

• Master the data.
• Perform the test plan.
• Address and refine results.
• Communicate insights.
• Track outcomes.

Exhibit 1-1 The IMPACT Cycle

13
Prepared by Nguyen Huu [email protected]

13

The data analytics process using the IMPACT cycle

Step 1: Identify the Questions

• Understand the business problems that need to be addressed.
• Are employees circumventing internal controls over payments?
• Are there any suspicious travel and entertainment expenses?
• How can we increase the amount of add-on sales of additional goods to
our customers?
• Are our customers paying us in a timely manner?
• How can we predict the allowance for loan losses for our bank loans?
• How can we find transactions that are risky in terms of accounting
issues?
• Who authorizes checks above $100,000?
• How can errors be identified?
14
Prepared by Nguyen Huu [email protected]

14

The data analytics process using the IMPACT cycle

Step 2: Master the Data

• Know what data are available and how they relate to the problem.
 Internal ERP systems.
 External networks and data warehouses.
• Data dictionaries.
• Extraction, transformation, and loading.
• Data validation and completeness.
• Data normalization.
• Data preparation and scrubbing

15
Prepared by Nguyen Huu [email protected]

15

5
 21/12/2024

The data analytics process using the IMPACT cycle

Step 2: Master the Data

Question: How are our sales

performing this quarter
compared to last quarter?

Data Needed: Sales

transaction records (ID,
dates, products, quantities,
prices, amount)

16
Prepared by Nguyen Huu [email protected]

16

The data analytics process using the IMPACT cycle

Step 2: Master the Data

Question: What are the
different segments of
customers based on
buying behavior?
Data Needed:
• Customer demographic information (age,
gender, location, etc.)
• Transaction history (purchase amounts,
product categories)
• Behavioral data (frequency of purchases,
average order value)
17
Prepared by Nguyen Huu [email protected]

17

The data analytics process using the IMPACT cycle

Step 2: Master the Data

Which data are likely to be
predictive of loan rejection,
and which are not?

1. Amount Requested
2. Zip Code
3. Loan Title
4. Debt-To-Income Ratio
5. Application Date
6. Risk_Score
7. Employment Length
18
Prepared by Nguyen Huu [email protected]

18

6
 21/12/2024

The data analytics process using the IMPACT cycle

Step 3: Perform the Test Plan

• Select an appropriate model to find a target variable.
 Classification.
 Regression.
 Similarity matching.
 Clustering.
 Co-occurrence grouping.
 Profiling.

19
Prepared by Nguyen Huu [email protected]

19

The data analytics process using the IMPACT cycle

Step 4: Address and Refine Results

• Identify issues with the analyses, possible issues, and refine the
model
• Ask further questions.
• Explore the data.
• Rerun analyses.

20
Prepared by Nguyen Huu [email protected]

20

The data analytics process using the IMPACT cycle

Step 5: Communicate Insights

• Communicate effectively using clear language and visualizations:
 Dashboards.
 Static reports.
 Summaries.
• Data visualization is the use of a graphical representation of data to
convey meaning. A common way to display data visualization is
with a data dashboard, or dashboard for short.
• A data dashboard is a display of important data points, metrics, and
key performance indicators in easily understood data visualizations.

21
Prepared by Nguyen Huu [email protected]

21

7
 21/12/2024

The data analytics process using the IMPACT cycle

Step 5: Communicate Insights

22
Prepared by Nguyen Huu [email protected]

22

The data analytics process using the IMPACT cycle

Step 6: Track Outcomes

• Follow up on the results of the analysis.
• How frequently should the analysis be performed?
• Have the analytics changed?
• What are the trends?

23
Prepared by Nguyen Huu [email protected]

23

Hands-on Example of the IMPACT Model

• Step 1: Identify the Question

 Assume you want to get a loan to pay off some credit card debt.
 LendingClub is a peer-to-peer lender that connects individual lenders
with borrowers.
 Use the IMPACT model to determine whether you’re likely to get a loan.
 “Given my borrower profile, can I expect LendingClub to extend a loan
to me?”
 What other questions might you ask?

24
Prepared by Nguyen Huu [email protected]

24

8
 21/12/2024

Hands-on Example of the IMPACT Model

• Step 2: Master the Data – LendingClub

LendingClub is a U.S.-based, peer-to-peer Data available:

lending company, headquartered in San • Approved loans (LoanStats).
Francisco, California. LendingClub • Rejected loan stats
facilitates both borrowing and lending by (RejectStats).
providing a platform for unsecured
personal loans between $1,000 and
$35,000. The loan period is for either 3 or
5 years.

25
Prepared by Nguyen Huu [email protected]

25

Hands-on Example of the IMPACT Model

• Step 2: Master the Data – Loan Summary
Personal loans have grown
since 2011. The majority
are for refinancing.

Exhibit 1-7 LendingClub Statistics Exhibit 1-8 LendingClub Statistics

by purpose by Reported Loan Purpose
Source: Accessed December 2020, Source: Accessed December 2020,
https://www.lendingclub.com/info/statistics.action https://www.lendingclub.com/info/statistics.action
26
Prepared by Nguyen Huu [email protected]

26

Hands-on Example of the IMPACT Model

• Step 2: Master the Data – Rejected Stats Data Dictionary
RejectStats File Description
Amount Requested The total amount requested by the borrower
Application Date The date which the borrower applied
Loan Title The loan title provided by the borrower
Risk_Score For applications prior to November 5, 2013 the risk score is the borrower's FICO score. For applications after
November 5, 2013 the risk score is the borrower's Vantage score.
Debt-To-Income A ratio calculated using the borrower’s total monthly debt payments on the total debt obligations, excluding
Ratio mortgage and the requested LC loan, divided by the borrower’s self-reported monthly income.
Zip Code The first 3 numbers of the zip code provided by the borrower in the loan application.
State The state provided by the borrower in the loan application
Employment Length Employment length in years. Possible values are between 0 and 10 where 0 means less than one year and 10
means ten or more years.
Policy Code publicly available: policy_code=1
new products not publicly available: policy_code=2

Exhibit 1-9 2007 to 2012 LendingClub Data Dictionary for Declined Loan Data
Source: LendingClub

27
Prepared by Nguyen Huu [email protected]

27

9
 21/12/2024

Hands-on Example of the IMPACT Model

• Step 2: Master the Data – Sample Data
Amount Application Loan Title Risk_Score Debt-To-Income Zip Code State Employment Length
Requested Date Ratio
1000 5/26/2007 Wedding Covered but No Honeymoon 693 10% 481xx NM 4 years
1000 5/26/2007 Consolidating Debt 703 10% 010xx MA < 1 year
11000 5/27/2007 Want to consolidate my debt 715 10% 212xx MD 1 year
6000 5/27/2007 waksman 698 38.64% 017xx MA < 1 year
1500 5/27/2007 mdrigo 509 9.43% 209xx MD < 1 year
15000 5/27/2007 Trinfiniti 645 0% 105xx NY 3 years
10000 5/27/2007 NOTIFYi Inc 693 10% 210xx MD < 1 year
3900 5/27/2007 For Justin. 700 10% 469xx IN 2 years
3000 5/28/2007 title? 694 10% 808xx CO 4 years
2500 5/28/2007 timgerst 573 11.76% 407xx KY 4 years
3900 5/28/2007 need to consolidate 710 10% 705xx LA 10+ years
1000 5/28/2007 sixstrings 680 10% 424xx KY 1 year
3000 5/28/2007 bmoore5110 688 10% 190xx PA < 1 year
1500 5/28/2007 MHarkins 704 10% 189xx PA 3 years

Exhibit 1-10 2007 to 2012 Declined Loan Applications (DAA Chapter 1-1 Data) Dataset
Source: LendingClub

28
Prepared by Nguyen Huu [email protected]

28

Hands-on Example of the IMPACT Model

• Step 3: Perform the Test Plan – Analyses

 Perform three analyses to predict whether we receive a loan:
o The debt-to-income (DTI) ratios and number of rejected loans.
o The length of employment and number of rejected loans.
o The credit (or risk) score and number of rejected loans.

29
Prepared by Nguyen Huu [email protected]

29

Hands-on Example of the IMPACT Model

• Step 3: Perform the Test Plan – Pivot

 For DTI, we set buckets:
o High = debt > 20% of income
o Mid = debt is 10 to 20% of income
o Low = debt < 10% of income
Here we see a PivotTable with results on
RejectStats

Exhibit 1-11 LendingClub Declined Loan

Applications by DTI (Debt-to-Income)
Microsoft Excel

30
Prepared by Nguyen Huu [email protected]

30

10
 21/12/2024

Hands-on Example of the IMPACT Model

• Step 3: Perform the Test Plan – Buckets

 For employment length, we set buckets on
number of years.

Here we see a PivotTable with results on RejectStats

Exhibit 1-12 LendingClub Declined Loan

Applications by Employment Length (Years
of Experience)
Microsoft Excel
31
Prepared by Nguyen Huu [email protected]

31

Hands-on Example of the IMPACT Model

• Step 3: Perform the Test Plan – Scores

 For credit score, we set buckets:
o Excellent: 800 to 850
o Very good: 750 to 799
o Good: 700 to 749
o Fair: 650 to 699
o Poor: 600 to 649
o Very bad: 300 to 599

Exhibit 1-13 Breakdown of Customer Credit

Scores (or Risk Scores)
32
Prepared by Nguyen Huu [email protected]

32

Hands-on Example of the IMPACT Model

• Step 3: Perform the Test Plan – Scores

 For credit score, we set buckets:
 Note in the rejected loans that nearly 82%
[(167,379 + 151,716 + 207,234)/645,414] of
the applicants have either very bad, poor,
or fair credit ratings, suggesting this might
be a good reason for a loan rejection. Also
note that only 0.3% (2,494/645,414) of
those rejected loan applications had
excellent credit.

Exhibit 1-14 The Count of LendingClub Rejected

Loan Applications by Credit or Risk Score
Classification Using PivotTable Analysis
Microsoft Excel

33
Prepared by Nguyen Huu [email protected]

33

11
 21/12/2024

Hands-on Example of the IMPACT Model

• Step 4: Address and Refine Results - Summary

 From the PivotTable analysis, we find that of the rejected loans:
 82% have either very bad, poor, or fair credit
 48% had a high DTI ratio
 76% had a work history of one year or less

34
Prepared by Nguyen Huu [email protected]

34

Hands-on Example of the IMPACT Model

• Step 4: Address and Refine Results - Interactions

• If we look at interactions of credit score &
DTI & employment length in a Pivot Table,
we see they are somewhat predictive.
• Only 365 of the 645,414 loans (0.057% of
the total).were rejected with the top buckets
from each.

Exhibit 1-15 The Count of LendingClub

Declined Loan Applications by Credit Score,
Debt-to-Income, and Employment Length
Using PivotTable Analysis 35
Microsoft Excel
Prepared by Nguyen Huu [email protected]

35

Hands-on Example of the IMPACT Model

• Step 4: Address and Refine Results - Predictions

• If we look at interactions of credit score &
DTI in a Pivot Table, we see they are
somewhat predictive.
• For example, if these are the applications
that were all rejected, the question is how
many of these that might apply for a loan
not only had excellent credit, but also had
worked more than 10 years and had asked
for a loan that was less than 10 percent of
their income (in the low DTI bucket)?

Exhibit 1-13 The Average Debt-to-Income Ratio (shown as a

percentage) by Credit (Risk) Score for LendingClub Declined Loan
Applications Using PivotTable Analysis
Microsoft Excel
36
Prepared by Nguyen Huu [email protected]

36

12
 21/12/2024

Hands-on Example of the IMPACT Model

• Step 5: Communicate Insights

 The PivotTables provide a simple visual. Additional visualizations or tools
may provide quick analysis by those evaluating the loans.
 Another goal is to share the results in plain English: “If I have good credit,
low debt-to-income, and a long employment history, it is very likely that my
loan will be accepted.”

37
Prepared by Nguyen Huu [email protected]

37

Hands-on Example of the IMPACT Model

• Step 6: Track Outcomes

 Extending this analysis to future periods will help us determine whether
these factors hold true or if there is some shift in the future.
 We attempt to use past performance to predict future results, but that may
not always hold true. When factors change, we repeat the IMPACT cycle.

38
Prepared by Nguyen Huu [email protected]

38

What data analytic skills do accountants need?

• Accountants must know how to do the following:

 Articulate business problems.
 Communicate with data scientists.
 Draw appropriate conclusions.
 Present results in an accessible manner.

39
Prepared by Nguyen Huu [email protected]

39

13
 21/12/2024

What data analytic skills do accountants need?

• The skills that analytic-minded accountants should have:

 Develop an analytics mindset
 Data scrubbing and data preparation
 Data quality
 Descriptive data analysis
 Data analysis through data manipulation
 Define and address problems through statistical data analysis
 Data visualization and data reporting

40
Prepared by Nguyen Huu [email protected]

40

What data analytic skills do accountants need?

• Choose the right Data Analytics Tools

 Based on Gartner’s magic quadrant, Microsoft and Tableau provide
innovative solutions.
o The Microsoft Track includes Excel, Power Query, Power BI, and Power Automate
o The Tableau Track includes Tableau Prep Builder, Tableau Desktop, and Tableau
Public

41
Prepared by Nguyen Huu [email protected]

41

What data analytic skills do accountants need?

• Microsoft Data Analytics Tools

Tool Excel Power Query Power BI Power Automate
Large datasets
Large datasets Collect data from
Small datasets Advanced
Data joining multiple sources
Data tables visualization
Good for Data cleaning Robotics process
PivotTables Dashboards
Data automation
Basic visualization Presentation
transformation

Windows/Mac/Onl
Platform Windows Windows/Online Online
ine

EXHIBIT 1-5 Microsoft Data Analytics Tools

42
Prepared by Nguyen Huu [email protected]

42

14
 21/12/2024

What data analytic skills do accountants need?

• Tableau Data Analytics Tools

Tool Tableau Prep Builder Tableau Desktop Tableau Public

Large datasets
Large datasets
Data summarization Analyze and share
Advanced visualization
Good for Data joining public datasets
Dashboards
Data cleaning
Presentation
Data transformation
Platform Windows/Mac/Online Windows/Mac/Online Windows/Mac/Online

EXHIBIT 1-6 Tableau Data Analytics Tools

43
Prepared by Nguyen Huu [email protected]

43

Chapter 1 Summary

With data all around us, businesses and Eight data approaches address different ways of
accountants are looking to Data Analytics to testing the data: classification, regression,
extract the value that the data might possess. similarity matching, clustering, co-occurrence
Data Analytics is changing the audit and the way grouping, profiling, link prediction, and data
that accountants look for risk. Now, auditors can reduction. These are explained in more detail in
consider 100 percent of the transactions in their chapter 3.
audit testing. It is also helpful in finding the
anomalous or unusual transactions. Data Analytics Data analytic skills needed by analytic-minded
is also changing the way financial accounting, accountants are specified and are consistent with
managerial accounting, and taxes are done at a
company. the IMPACT cycle, including the following:
Develop an analytics mindset.
The IMPACT cycle is a means of doing Data Data scrubbing and data preparation.
Analytics that goes all the way from identifying the Data quality.
question, to mastering the data, to performing Descriptive data analysis.
data analyses and communicating results. It is Data analysis through data manipulation.
recursive in nature, suggesting that as questions Define and address problems through statistical data
are addressed, new important questions may analysis.
emerge that can be addressed in a similar way. Data visualization and data reporting.

44
Prepared by Nguyen Huu [email protected]

44

Key words
Big Data (4) Dữ liệu lớn
Classification (11) Phân loại
Clustering (11) Phân cụm
Co-occurrence grouping (11) Gom nhóm xảy ra đồng thời
Data Analytics (4) Phân tích dữ liệu
Data dictionary (19) Từ điển dữ liệu
Data reduction (12) Rút gọn dữ liệu
Link prediction (12) Dự đoán mối quan hệ
Predictor (independent or explanatory) variable (11) Biến dự đoán (biến độc lập hoặc biến giải thích)
Profiling (11) Lập hồ sơ dữ liệu
Regression (11) Hồi quy
Response (or dependent) variable (10) Biến phản hồi (hoặc biến phụ thuộc)
Similarity matching (11) Kết hợp tương đồng
Structured data (4) Dữ liệu có cấu trúc
Unstructured data (4) Dữ liệu phi cấu trúc

45
Prepared by Nguyen Huu [email protected]

45

15