308869907
1
Chapter 8
ANALYTICAL PROCEDURES
8.1 Learning Objectives
After studying this chapter you should be able to:
1
Understand the general nature of analytical procedures.
2
Describe four general analytical procedures.
3
Explain how expectations are developed and what sources are used.
4
Clarify how the effectiveness of an analytical procedure is a function of the nature of the
account and the reliability and other characteristics of the data.
5
Calculate the customary ratios that are used during the planning phase to determine
accounts that may represent significant risks to the entity of liquidity, solvency, profitability,
and activity.
6
Understand some indications that the going concern assumption may be questioned.
7
Comprehend why and how analytical procedures may be used at each audit phase.
8
Grasp how analytical procedures are used in substantive procedures.
9
Describe some of the analytical procedures carried out by computer-aided audit techniques.
10
Illustrate data mining methods, techniques, and algorithms used to analyze client data.
11
Portray the auditor’s procedures when analytical procedures identify significant
fluctuations that deviate from predicted amounts.
308869907
308869907
2
8.2 Introduction
Analytical procedures are evaluations of financial information through analysis of plausible
relationships among both financial and non-financial data. Analytical procedures also encompass
such investigation as is necessary of identified fluctuations or relationships that are inconsistent
with other relevant information or that differ from expected values by a significant amount.1 Put
another way, analytical procedures entail the use of comparisons and relationships to determine
whether account balances or other data appear reasonable. Such procedures allow the auditor to
look at things in overview and answer the question: Do the numbers make sense?
Relationships Among Data
A basic premise of using analytical procedures is that there exist plausible relationships among
data and these relationships can reasonably be expected to continue. Analytical procedures include
the comparison of the entity’s financial statements with prior period information, anticipated results
such as budgets, and similar industry information. For instance, an entity’s accounts receivable
turnover ratio may be compared to that ratio for the industry or for a similar company.
Analytical procedures are used to determine relationships among financial information that
would be expected to conform to predictable patterns based on the entity’s experience, such as
gross margin percentages, and between financial and non-financial information such as the
relationship between payroll costs and the number of employees.
Types of Analytical Procedures
General analytical procedures include trend analysis, ratio analysis, statistical and data mining
analysis, and reasonableness tests. Trend analysis is the analysis of changes in an account balance
over time. Ratio analysis is the comparison of relationships between financial statement accounts,
308869907
308869907
3
the comparison of an account with non-financial data, or the comparison of relationships between
firms in an industry. Reasonableness testing is the analysis of account balances or changes in
account balances within an accounting period in terms of their “reasonableness” in light of
expected relationships between accounts. Data mining is a set of computer-assisted techniques that
use sophisticated statistical analysis, including artificial intelligence techniques, to examine large
volumes of data with the objective of indicating hidden or unexpected information or patterns. For
these tests auditors generally use computer-aided audit software (CAATs).
When to Use Analytical Procedures
The auditor shall design and perform analytical procedures near the end of the audit that assist the
auditor when forming an overall conclusion as to whether the financial statements are consistent
with the auditor’s understanding of the entity according to ISA 520. 2. In the overall review stage
(at the end of the audit) the objective of analytical procedures is to assess the conclusions reached
and evaluate the overall financial statement presentation. It may be used to detect material
misstatements that other tests can overlook, such as fraud or understatement errors.
Analytical procedures may also be used in planning and applied to substantive testing. In
the planning stage analytical procedures may be used to highlight risk areas to narrow the focus of
planning the nature, timing, and extent of auditing procedures. In the substantive testing stage of
the audit, analytical procedures are used to see “the big picture”, i.e. obtain evidence to identify
misstatements in account balances and thus to reduce the risk of misstatements.
Computer Assisted Audit Techniques (CAATs)
The use of computer assisted audit techniques (CAATs) may enable more extensive testing
of electronic transactions and account files. CAATs can be used to select sample transactions from
308869907
308869907
4
key electronic files, to sort transactions with specific characteristics, or to test an entire population
instead of a sample. CAATs generally include data manipulation, calculation, data selection, data
analysis, identification of exceptions and unusual transactions, regression analysis, and statistical
analysis. CAATs are available on generalized audit software (GAS) which provides the auditors
with the ability to access, manipulate, manage, analyze and report data in a variety of formats. Data
mining software may also be used for analytical procedures.
When analytical procedures identify significant fluctuations or relationships that are
inconsistent with other relevant information or that deviate from predicted amounts, the auditor
should investigate and obtain adequate explanations and appropriate corroborative evidence. The
investigation of unusual fluctuations and relationships ordinarily begins with inquiries of
management followed by corroboration of management’s responses and other audit procedures
based on the results of these inquiries.
8.3 The Analytical Review Process
The process of planning, executing, and drawing conclusions from analytical procedures is
called analytical review. There are several views, theoretical and practical, of the sub-processes
involved in analytical review.
The theoretical view3 is that the review process consists of four diagnostic processes:
1
mental representation,
2
hypothesis generation,
3
information search,
4
hypothesis evaluation.
In particular, auditors hypothesize causes and related probabilities, gather evidence to test the
hypotheses, and ultimately select which hypotheses are most likely to cause the fluctuation.4
308869907
308869907
5
Concept and a Company 8.1
“Crazy Eddie – His prices are insane!”
Concept
Analytical procedures to test inventory.
Story
In 1969, Eddie Antar, a 21-year-old high school dropout from Brooklyn, opened a consumer
electronics store with 15 square meters of floor space in New York City. By 1987, Antar’s firm.
Crazy Eddie, Inc., had 43 retail outlets, sales exceeding $350 million, and outstanding common
shares with a collective market value of $600 million.
Shortly after a hostile takeover of the company in November 1987, the firm’s new owners
discovered that Crazy Eddie’s inventory was overstated by more than $65 million. Subsequent
investigations by regulatory authorities would demonstrate that Crazy Eddie’s profits had been
intentionally overstated by Eddie Antar and several subordinates (Belsky and Furman 1989).
Crazy Like a Fox
Antar acquired the nickname “Crazy Eddie” because of his unique sales tactic. Whenever a
customer would attempt to leave his store without purchasing something, Eddie would block the
store’s exit, sometimes locking the door until the individual agreed to buy something – anything.
To entice a reluctant customer to make a purchase, Antar would lower the price until the
customer finally gave in. From 1972, Doctor Jerry was the spokesperson for Crazy Eddie. He
made a series of ear-piercing television commercials that featured him screaming “Crazy Eddie
– His prices are insane!” The company promised to refund the difference between the selling
price of a product and any lower price for that same item that a customer found within 30 days of
the purchase date. (Knapp 2001)
Inventory Overstated
Trouble was that in late 1986 the boom days had ended for the consumer electronics industry.
To continue the growth of the company and keep the stock price up, Antar had to do something.
Within the first six months after the company went public, Antar ordered a subordinate to
overstate inventory by $2 million, resulting in the firm’s gross profit being overstated by the same
amount. The following year Antar ordered year-end inventory to be overstated by $9 million and
308869907
308869907
6
accounts payable to be understated by $3 million. (Belsky and Furman 1989) Crazy Eddie
employees overstated year-end inventory by preparing inventory count sheets for items that did
not exist. To overstate accounts payable, bogus debit memos were prepared and entered in the
company’s accounting records.
The Audits
Crazy Eddie’s auditor was Main Hurdman (later merged with Peat Marwick – now KPMG). Their
audits were generally made difficult by management and employee collusion. There were
several reported instances in which the auditors requested client documents, only to be told that
those documents had been lost or inadvertently destroyed. Upon discovering which sites the
auditors would be visiting to perform year-end inventory procedures, Antar would ship sufficient
inventory to those stores or warehouses to conceal any shortages. Furthermore, personnel
systematically destroyed incriminating documents to conceal inventory shortages from the
auditors. (Weiss 1993)
Main Hurdman has been criticized for charging only $85,000 for a complete SEC audit, but
millions to install a computerized inventory system. This is even more interesting because Antar
ordered his employees to stop using the sophisticated, computer-based inventory system
designed by Main Hurdman. Instead, the accounting personnel were required to return to a
manual inventory system previously used by the company. The absence of a computer-based
inventory system made it much more difficult for the auditors to determine exactly how much
inventory the firm had at any point in time (Weiss 1993).
Crazy Eddie Comparative Income Statements 1984–87
31 March 87
31 March 86
31 March 85
31 March 84
$352,523
$262,268
$136,319
$137,285
–0272,255
–194,371
–103,421
–106,934
Gross profit
80,268
67,897
32,898
30,351
Selling, G&A
–61,341
–42,975
–20,508
–22,560
7,403
3,210
1,211
706
Net sales
Cost of goods sold
expense
Interest and other
income
308869907
308869907
7
–5,233
–820
–438
–522
Income before taxes
$21,097
$27,312
$13,163
$7,975
Pension contribution
–500
–800
–600
–4,202
Income taxes
–10,001
–13,268
–6,734
Net income
$10,596
$13,244
$5,829
$3,773
$0.34
$0.48
$0.24
$0.18
Interest expense
Net income per
share
Crazy Eddie Comparative Balance Sheets 1984–87
31 March 87
31 March 86
31 March 85
31 March 84
$9,347
$13,296
$22,273
$1,375
121,957
26,840
Receivables
10,846
2,246
2,740
2,604
Merchandise
109,072
59,864
26,543
23,343
10,639
2,363
645
514
261,861
104,609
52,201
27,836
3,356
7,058
Current assets
Cash
Short-term
investments
inventories
Prepaid expenses
Total current assets
Restricted cash
Due from affiliates
Property, plant and
5,739
26,401
7,172
3,696
6,253
1,154
1,845
equipment
Construction in
process
Other assets
6,596
5,560
1,419
1,149
Total assets
$294,858
$126,950
$65,528
$36,569
$50,022
$51,723
$23,078
$20,106
Current liabilities
Accounts payable
Notes payable
Short-term debt
2,900
49,571
308869907
2,254
423
124
308869907
8
Unearned revenue
3,641
3,696
1,173
764
Accrued expenses
5,593
17,126
8,733
6,078
108,827
74,799
33,407
29,972
8,459
7,701
7,625
46
3,337
1,829
635
327
313
280
134
50
57,678
17,668
12,298
574
Retained earnings
35,269
24,673
11,429
5,600
Total stockholders’
93,260
42,621
23,861
6,224
$294,858
$126,950
$65,528
$36,569
Total current
liabilities
Long-term debt
Convertible
subordinated
Debentures
80,975
Unearned revenue
Stockholders’ equity
Common stock
Additional paid-in
capital
equity
Total liabilities and
stockholders’equity
Discussion Questions
Review the Crazy Eddie comparative financial statements and discuss:
What analytical procedures can be performed to determine if inventory is misstated?
What indicators other than financial hinted that there might be problems?
References Belsky, G. and Furman, P., 1989, “Calculated Madness: The Rise and Fall of Crazy Eddie
Antar,” Cram’s New York Business, June 5, pp.21–33.
Knapp, M., 2001, “Crazy Eddie, Inc,” Contemporary Auditing Real Issues & Cases, South
Western College Publishing, Cincinnati, Ohio, pp.71–82.
Weiss, M.I., 1993, “Auditors: Be Watchdogs, Not Just Bean Counters,” Accounting Today,
November 15, p.41.
308869907
308869907
9
Four-Phase Process
Here we use a practitioner approach, the four-phase process most common in professional
literature:5

phase one – formulate expectations (expectations);

phase two – compare the expected value to the recorded amount (identification);

phase three – investigate possible explanations for a difference between expected and recorded
values (investigation);

phase four – evaluate the impact of the differences between expectation and recorded amounts
on the audit and the financial statements (evaluation).
Illustration 8.1 shows the four-phase process and its inputs and outputs.
Phase One
According to ISA 5206, the auditor should “evaluate the reliability of data from which the
auditor’s expectation of recorded amounts or ratios is developed, taking account of source,
comparability, and nature and relevance of information available, and controls over preparation.”
In phase one of the analytical review process, the auditor develops expectations of what amounts
should appear in financial statement account balances based on prior year financial statements,
budgets, industry information and non-financial information. Expectations are the auditor’s
estimations of recorded accounts or ratios. The auditor develops his expectation in such a way that
a significant difference between it and the recorded amount will indicate a misstatement.
ILLUSTRATION 8.1
THE ANALYTICAL REVIEW PROCESS
[Figure 8.1 near here]
308869907
308869907
10
Forming an expectation is the most important phase of the analytical procedure process. The
closer the auditor’s expectation is to the correct balance or relationship, the more effective the
procedure will be at identifying potential misstatements.
Expectations are formed from a variety of sources. Research7 suggests that the use of
industrial, economic, or environmental data can improve the predictive ability of analytical
procedures. Other resources include industry data, data about similar businesses, and auditor
experience. Expectations are also based on the entities prior financial statements, same store sales,
non-financial data, budgets and public reports.
308869907
308869907
11
Before determining what analytical procedures to use the audor must “determine the suitability
of particular substantive analytical procedures for given assertions, taking account of the assessed
risks of material misstatement and tests of details, if any, for these assertions.” 8
Phase Two
Phase two of the analytical review process (identification) is when the auditor compares his
expected value with the recorded amount. Audit efficiency and effectiveness depend on
competency in recognizing error patterns in financial data and in hypothesizing likely causes of
those patterns to serve as a guide for further testing.
The auditor should determine the amount of any difference of recorded amounts from
expected values that is acceptable without further investigation The auditor must consider how
large a difference between expected value and recorded amount he will accept. The threshold
amount may very well relate to or derived from materiality. If the difference is less than the
acceptable threshold, the auditor accepts the book value without further investigation. If the
difference is greater, the next step is to investigate the difference.
Phase Three
In phase three of the analytical review process (investigation), the auditor undertakes an
investigation of possible explanations for the expected/recorded amount difference. The difference
between an auditor’s expectation and the recorded book value of an account not subject to auditing
procedures can be due to misstatements, inherent factors that affect the account being audited, and
factors related to the reliability of data used to develop the expectation.
The greater the precision of the expectation, the more likely the difference between the
auditor’s expectation and the recorded value will be due to misstatements. Conversely, the less
308869907
308869907
12
precise the expectation, the more likely the difference is due to factors related to inherent factors,
and the reliability of data used to develop the expectation.
Inquiries and Corroboration of Differences
Where differences between expectation and recorded amounts are found, the first step is
usually to ask management for an explanation. However, it is important that the auditor maintains
his professional skepticism (see Chapter 4) when considering these answers and it is suggested
that the auditor conduct other audit procedures to corroborate them. When analytical procedures
are used in the planning phase, corroboration is not immediately required because the purpose of
analytical procedures in planning is to chart the work to follow.
Phase Four
The final phase (phase four – evaluation) of the analytical review process involves evaluating
the impact on the financial statements of the difference between the auditor’s expected value and
the recorded amount. It is usually not practical to identify factors that explain the exact amount of a
difference investigated. The auditor attempts to quantify that portion of the difference for which
plausible explanations can be obtained and, where appropriate, corroborated. If the amount that
cannot be explained is sufficiently small, the auditor may conclude there is no material
misstatement. However, if the amount that cannot be explained exceeds the threshold (phase two),
additional substantive procedures are required.
308869907
308869907
13
8.4 Formulating Expectations
Expectations are developed by identifying plausible relationships that are reasonably expected
to exist based on the auditor’s understanding of the client and of his industry. These relationships
may be determined by comparisons with the following sources9:

comparable information for prior periods;

anticipated results (such as budgets and forecasts, or auditor expectations);

elements of financial information within the period;

similar industry information;

non-financial information.
The auditor can identify account balances that have changed significantly simply by
comparing the current client data with prior period data. He may compare the current year’s
account balances with that of the preceding year; the current trial balance with similar detail for the
preceding year; and ratios and percentage relationships between years. He compares current
recorded account balances of the entity with results expected by the entity. For example, company
budgets may be compared with actual results for indications of potential misstatements or the
auditor calculates the expected balance for interest expense (notes payable monthly balance times
average monthly interest rate) and compares this to recorded interest rates.
One of the standard bases of comparison is with like companies in the same industry. For
example, the auditor compares the gross margin of the industry to the client’s gross margin.
Analytical procedures may include consideration of relationships between financial
information and relevant non-financial information, such as payroll costs to number of employees.
Another example is the revenue of a hotel may be estimated by multiplying the average room rate
times the number of rooms times the average occupancy percentage (e.g. $100 average rate”20
rooms”60 percent average occupancy). Similarly revenues may be calculated for school tuition
308869907
308869907
14
(average number of students enrolled times the average tuition cost), payroll, and cost of materials
sold using non-financial factors.
Sources of Information and Precision of Expectations
The source of information on which the expectations are based (e.g. prior period statements,
forecasts, industry information) determines, in part, the precision with which the auditor predicts an
account balance. For example, information from other, similar stores in the same retail chain is
more precise than general industry information. Recent years’ financial statements are more precise
a predictor of this year’s balance than older financial statements. The desired precision of the
expectation varies according to the purpose of the analytical procedure. Expectations developed at
a detailed level generally have a greater chance of detecting misstatement of a given amount than
do broad comparisons. Monthly amounts will generally be more effective than annual amounts and
comparisons by location or line of business usually will be more effective than company-wide
comparisons. Precision is more important for analytical procedures used as substantive tests than
for those used in planning.
Nature of Account and Characteristics of Data
The effectiveness of an analytical procedure is a function of the nature of the account and the
reliability and other characteristics of the data. In determining the nature of the account we
consider whether the balance is based on estimates or accumulations of transactions, the number of
transactions represented by the balance, and the control environment. Subjectively determined
balances are more easily manipulated than accumulations of transactions. If the characteristic of the
account is that it comprises millions of transactions (e.g. retail revenue), it should be more
308869907
308869907
15
predictable than those comprising a few transactions (e.g. obsolete inventory). Fixed expenses (e.g.
leases) are more predictable than variable expenses (e.g. shipping).
Other data characteristics such as the level of detail (aggregation) on which the auditor is able
to base his expectation and the reliability of the data are key characteristics. In general, the more
disaggregated the data, the more precise the expectation. For example, the use of monthly instead
of annual data tends to improve the precision of the expectation. Preparing an expectation by
division is also more precise than an expectation based on consolidated data. Accounting
researchers9 conclude that disaggregated monthly, segment, or product line balances are required to
implement reliable attention-directing analytical procedures.
The more reliable the source of the data, the more precise the expectation will be. Reliability
of data is determined based on the strength of the company’s internal control, if the data source is
objective or independent, and if data has been subject to auditing procedures or not. Stronger
internal control over financial reporting and accounting systems produces more reliable data on the
financial statements. The use of reliable non-financial data (e.g. store square footage or occupancy
rates) and the use of data that has been subjected to auditing procedures improve the precision of
the expectation based on that data.
8.5 General Analytical Procedures
The general analytical procedures are trend analysis, ratio analysis, statistical and data mining
analysis, and reasonableness tests. Determining which type of analytical procedure is appropriate is
a matter of professional judgment. A review of audit practice indicates that simple judgmental
approaches (such as comparison and ratio analysis) are used more frequently than complex
statistical approaches (such as time series modeling or regression analysis). 10 These tests are
generally carried out using computer software (i.e. CAATs). Trend analysis, ratio analysis, and
308869907
308869907
16
reasonableness tests are discussed in this section. Statistical and data mining analysis is discussed
in the following section on CAATs.
Trend Analysis
Trend analysis is the analysis of changes in an account balance or ratio over time. Trend
analysis could compare last year’s account balance to the current unaudited balance or balances in
many time periods. Trend analysis works best when the account or relationship is fairly predictable
(e.g. rent expense in a stable environment). It is less effective when the audited entity has
experienced significant operating or accounting changes. The number of years used in the trend
analysis is a function of the stability of operations. The more stable the operations over time, the
more predictable the relations and the more appropriate the use of multiple time periods. Trend
analysis at an aggregate level (e.g. on a consolidated basis) is relatively imprecise because a
material misstatement is often small relative to the aggregate account balance. The most precise
trend analysis would be on disaggregated data (e.g. by segment, product, or location, and monthly
or quarterly rather than on an annual basis).
Ratio Analysis
Ratio analysis is the comparison of relationships between financial statement accounts, the
comparison of an account with non-financial data, or the comparison of relationships between firms
in an industry. Another example of ratio analysis (which is sometimes referred to as common size
analysis) is to set all the account balances as either a percentage of total assets or revenue.
Ratio analysis is most appropriate when the relationship between accounts is fairly predictable
and stable (e.g. the relationship between sales and accounts receivable). Ratio analysis can be more
effective than trend analysis because comparisons between the balance sheet and income statement
308869907
308869907
17
can often reveal unusual fluctuations that an analysis of the individual accounts would not. Like
trend analysis, ratio analysis at an aggregate level is relatively imprecise because a material
misstatement is often small relative to the natural variations in the ratios.
Types of Ratio Analysis
There are five types of ratio analysis used in analytical procedures (See Illustration 8.2):
1
ratios that compare client and industry data;
2
ratios that compare client data with similar prior period data;
3
ratios that compare client data with client-determined expected results;
4
ratios that compare client data with auditor-determined expected results;
5
ratios that compare client data with expected results using non-financial data.
ILLUSTRATION 8.2
FIVE TYPES OF RATIO ANALYSIS
Procedures
Examples
Ratios that compare client data with industry.
Standard ratios published by the industry by the
Risk Management Association, Standard &
Poors, Dun & Bradstreet and others.
Ratios that compare client data with similar
Auditor compares the current year’s account
prior period data.
balances with that for the preceding year; current
trial balances with similar detail for the preceding
year; and ratios and percentage relationships
between years.
Ratios that compare client data with client-
Client budgets may be compared with actual
determined expected results.
results for indications of potential misstatements.
308869907
308869907
Ratios that compare client data with auditor-
Auditor calculates the expected balance for
determined expected results.
interest expense and compares to recorded
18
interest.
Ratios that compare client data with expected
Non-financial data may serve as a basis for
results using non-financial data.
expected results and comparison, the revenue of
a hotel may be estimated by multiplying average
room rate times the number of rooms times the
average occupancy percentage.
The Risk Management Association (RMA), Standard and Poors, Dun & Bradstreet, Ibis
World. and others publish standard ratios by industry.11 Similar ratios for both industry and entity
may be compared to indicate differences that might affect the auditor’s judgment of the nature and
extent of audit procedures. The ratios indicate entity liquidity, solvency, profitability, and activity.
See Illustration8.3 for some standard ratios often used and compared.
ILLUSTRATION 8.3
STANDARD CLIENT AND INDUSTRY RATIOS
Client and industry standard ratios
Calculation
Liquidity:
(1) Current ratio
(1) Current assets/Current liabilities
(2) Quick ratio
(2) (Cash!Short-term securities!Accounts
receivable)/Current liabilities
Solvency:
(1) Debt to equity
(1) Long-term debt/Stockholders’ equity
(2) Times interest earned
(2) (Net income before interest and taxes)/Interest
308869907
308869907
19
expense
(3) Debt service coverage
(3) (Net income before interest and
depreciation)/Principal and interest payments
Profitability:
(1) Net profit margin
(1) Net profit/Revenue
(2) Gross margin
(2) (Revenue less cost of goods sold)/Revenue
(3) Return on investment
(3) Net income/Stockholders’ equity
(4) Times interest earned
(4) (Net income before interest and taxes)/Interest
expense
Activity:
(1) Receivable turnover
(1) Revenue/Average accounts receivable
(2) Inventory turnover
(2) Cost of goods sold/Average inventory
(3) Asset turnover
(3) Revenue/Total assets
Reasonableness Testing
Reasonableness testing is the analysis of account balances or changes in account balances
within an accounting period in terms of their “reasonableness” in light of expected relationships
between accounts. This involves the development of an expectation based on financial data, nonfinancial data, or both. For example, using the number of employees hired and terminated, the
timing of pay changes, and the effect of vacation and sick days, the model could predict the change
in payroll expense from the previous year to the current balance within a fairly narrow dollar range.
In contrast to both trend and ratio analyses (which implicitly assume stable relationships),
reasonableness tests use information to develop an explicit prediction of the account balance. The
auditor develops assumptions for each of the key factors (e.g. industry and economic factors) to
estimate the account balance. Considering the number of units sold, the unit price by product line,
different pricing structures, and an understanding of industry trends during the period could
308869907
308869907
20
explicitly form a reasonableness test for sales. This is in contrast to an implicit trend expectation
for sales based on last year’s sales. The latter expectation is appropriate only if there were no other
factors affecting sales during the current year, which is not the usual situation.
Trend Analysis, Ratio Analysis, and Reasonableness Tests Compared
Trend analysis, ratio analysis, and reasonableness tests differ as to the number of independent
predictive variables considered, use of external data, and statistical precision. Trend analysis is
limited to a single predictor, that is, the prior periods’ data for that account. Trend analysis, by
relying on a single predictor, does not allow the use of potentially relevant operating data, as do the
other types of procedures. Because ratio analysis employs two or more related financial or nonfinancial sources of information, the result is a more precise expectation. Reasonableness tests and
regression analysis further improve the precision of the expectation by allowing potentially as
many variables (financial and non-financial) as are relevant for forming the expectation.
Reasonableness tests and regression analysis are able to use external data (e.g. general economic
and industry data) directly in forming the expectation. The most statistically precise expectations
are formed using statistical and data mining analysis.
Standard Client and Industry Ratios
At the planning stage of an audit, there are certain customary ratios that are always calculated
to determine accounts that may represent significant risks to the entity of liquidity, solvency,
profitability, and activity. These ratios help to answer some key questions:

Is there a possible going concern problem (liquidity ratios)?

Is the entity’s capital structure sustainable (solvency ratios)?

Is gross margin reasonable (profitability)?
308869907
308869907

21
Could inventory be overstated (activity)?
Illustration 8.3 (earlier) gives a list of these ratios.
Liquidity and Going Concern
Auditors must determine the possibility that the company is having liquidity problems – that is, is
there a possibility that the company may no longer be a going concern? ISA 57012 states that under
the going concern assumption, an entity is viewed as continuing in business for the foreseeable
future. General purpose financial statements are prepared on a going concern basis, unless
management either intends to liquidate the entity or to cease operations, or has no realistic
alternative but to do so. When the use of the going concern assumption is appropriate, assets and
liabilities are recorded on the basis that the entity will be able to realize its assets and discharge its
liabilities in the normal course of business.
Analytical procedures may point to indications of risk that the going concern assumption
needs to be questioned. Illustration 8.4 shows the indications of risk that the going concern
assumption may be questioned. The significance of the indications in Illustration 8.4 can often be
mitigated by other factors. For example, the effect of an entity being unable to make its normal debt
repayments may be counterbalanced by management’s plans to maintain adequate cash flows by
alternative means, such as disposal of assets.
ILLUSTRATION 8.4
INDICATIONS THAT THE GOING CONCERN ASSUMPTION MIGHT BE QUESTIONED13
308869907
308869907
The following are examples of events or conditions that, individually or collectively, may cast
significant doubt about the going concern assumption. This listing is not all-inclusive nor does the
existence of one or more of the items always signify that a material uncertainty exists.
Financial

Net liability or net current liability position.

Fixed-term borrowings approaching maturity without realistic prospects of renewal or
repayment, or excessive reliance on short-term borrowing to finance long-term assets.

Indications of withdrawal of financial support by creditors.

Negative operating cash flows indicated by historical or prospective financial statements.

Adverse key financial ratios.

Substantial operating losses or significant deterioration in the value of assets used to
generate cash flows.

Arrears or discontinuance of dividends.

Inability to pay creditors on due dates.

Difficulty in complying with the terms of loan agreements.

Change from credit to cash-on-delivery transactions with suppliers.

Inability to obtain financing for essential new product development or other essential
investments.
Operating

Management intentions to liquidate the entity or to cease operations.

Loss of key management without replacement.

Loss of a major market, key customer(s), franchise, license, or principal supplier(s). .

Labor difficulties.

Shortages of important supplies. Labor difficulties or shortages of important supplies.

Emergence of a highly successful competitor.
Other

Non-compliance with capital or other statutory requirements.
308869907
22
308869907

23
Pending legal proceedings against the entity that may, if successful, result in claims
that are likely to be satisfied.

Changes in law or regulation or government policy expected to

adversely affect the entity. Changes in legislation or government policy.

Uninsured or underinsured catastrophes when they occur.
Concept and a Company 8.2
Peregrine Systems – The 37th of December
Concept
Analytical procedures for revenue tests.
Story
The accounting irregularities that brought down Peregrine Systems, a San Diego, California,
maker of software that large companies use to manage their technology resources, included
inflated revenues of more than 60 percent or some $509 million.
Peregrine had reported revenues of $1.34 billion for 2000 and 2001. Of that, $225 million was
based on “non-substantiated transactions” as the company booked revenues when it transferred
goods to a software reseller, even when there were no firm commitments in place. Another $70
million of the fictitious revenue was from swap transactions and $100 million came from the
premature booking of revenues from long-term instalment contracts. Peregrine had reported
another $80 million of revenues that, for a variety of reasons, should not have been recorded
until future years and that $34 million of revenues had been based on “erroneous calculations or
unsupported transactions.” (Waters 2003)
An SEC complaint, settled by Peregrine, alleged that the company improperly booked millions of
dollars of revenue for purported software license sales to resellers. These transactions were
non-binding sales of Peregrine software with the understanding – reflected in secret side
agreements – that the resellers were not obligated to pay Peregrine. Those involved in the
scheme called this “parking” the transaction. Peregrine personnel parked transactions when
Peregrine was unable to complete direct sales it was negotiating (or hoping to negotiate) with
end-users, but needed revenue to achieve its forecasts. (SEC 2003a)
308869907
308869907
24
Peregrine engaged in other deceptive practices to inflate the company’s revenue, including
entering into reciprocal transactions in which Peregrine essentially paid for its customers’
purchases of Peregrine software. Peregrine routinely kept its books open after fiscal quarters
ended, and improperly recorded as revenue, for the prior quarter, software transactions that
were not consummated until after quarter end. Certain Peregrine officers characterized these
transactions as having been completed on “the 37th of December”. (SEC 2003a)
To conceal the revenue recognition scheme, Peregrine abused the receivable financing process.
When Peregrine booked the non-binding contracts, and the customers predictably did not pay,
the receivables ballooned on Peregrine’s balance sheet. To make it appear that Peregrine was
collecting its receivables more quickly than it was, Peregrine “sold” receivables to banks and
then removed them from the company’s balance sheet. There were several problems with this.
First, Peregrine had given the banks recourse and frequently paid or repurchased unpaid
receivables from them. Peregrine should have accounted for the bank transactions as loans and
left the receivables on its balance sheet. Secondly, the sold receivables were not valid because
the customers were not obligated to pay. Thirdly, several of the sold invoices were fake,
including one that purported to reflect a $19.58 million sale. (SEC 2003b)
The SEC complaint also alleged that, as part of the cover up, Peregrine personnel wrote off
millions of dollars in uncollectible – primarily sham – receivables, to acquisition-related accounts
in Peregrine’s financial statements and books and records. These write-offs were improper
because they had nothing to do with acquisitions. (SEC 2003a)
Chief Executive Officer, Mathew C. Gless, alleged in a SEC civil suit, signed false SEC filings
and false management representation letters to Peregrine’s outside auditors, and responded
falsely to an SEC Division of Corporation Finance Comment Letter that inquired about
Peregrine’s revenue recognition practice. According to the complaint, while Gless was aware of
the ongoing fraud, he illegally sold 68,625 shares of Peregrine stock for approximately $4 million,
based on material non-public information he possessed about Peregrine’s true financial
condition. (SEC 2003b)
Peregrine Comparative Balance Sheet 2000–2001 (in 000)
31 March 01
ASSETS
308869907
31 March 00
308869907
Cash and cash equivalents
25
$286,658
$33,511
180,372
69,940
Other current assets
62,811
22,826
Total current assets
529,841
126,277
82,717
29,537
1,192,855
233,504
198,353
134,112
$2,003,766
$523,430
Accounts payable
$36,024
$19,850
Accrued expenses
200,886
49,064
86,653
36,779
1,731
74
325,294
105,767
8,299
4,556
Accounts receivable, net of allowance for doubtful
accounts of $11,511 and $2,179, respectively
Property and equipment, net
Goodwill, net of accumulated amortization of $334,178 and
$54,406, respectively
Other intangible assets, investments and other, net of
accumulated amortization of $24,015 and $1,398,
Respectively
LIABILITIES AND STOCKHOLDERS’ EQUITY
Current liabilities:
Current portion of deferred revenue
Current portion of long-term debt
Total current liabilities
Deferred revenue, net of current portion
Other long-term liabilities
17,197
Long-term debt, net of current portion
884
1,257
Convertible subordinated notes
262,327
Total liabilities
614,001
111,580
160
110
2,342,235
480,957
Stockholders’ equity:
Preferred stock, $0.001 par value, 5,000 shares authorized,
no shares issued or outstanding
Common stock, $0.001 par value, 500,000 shares authorized,
160,359 and 109,501 shares issued and outstanding,respectively
Additional paid-in capital
308869907
308869907
26
–917,104
–64,863
–22,151
–678
Cumulative translation adjustment
–3,950
–666
Treasury stock, at cost
–9,425
–3,010
1,389,765
411,850
$2,003,766
$523,430
Accumulated deficit
Unearned portion of deferred compensation
Total stockholders’ equity
Peregrine Systems Comparative Income Statement 1999–2001 (in 000)
Revenues:
31 March 01
31 March 00
31 March 99
Licenses
$354,610
$168,467
$87,362
Services
210,073
84,833
50,701
Total revenues
564,683
253,300
138,063
Cost of licenses
2,582
1,426
1,020
Cost of services
111,165
51,441
31,561
11,844
1,338
50
223,966
101,443
50,803
Research and development
61,957
28,517
13,919
General and administrative
48,420
19,871
10,482
Acquisition costs and other
918,156
57,920
43,967
1,378,090
261,956
151,802
–813,407
–8,656
–13,739
–538
38
664
–813,945
–8,618
–13,075
–38,296
–16,452
–10,295
–$852,241
–$25,070
–$23,370
–$6.16
–$0.24
–$0.27
138,447
102,332
87,166
Costs and expenses:
Amortization of purchased technology
Sales and marketing
Total costs and expenses
Loss from operations before interest (net)
and income tax expense
Interest income (expense), net
Loss from operations before income tax expense
Income tax expense
Net loss
Net loss per share basic and diluted:
Net loss per share
Shares used in computation
Source: US Securities and Exchange Commission, www.sec.gov.
308869907
308869907
Discussion Questions
27
Review the Peregrine financial statements and determine what analytical
procedures could be used to predict the revenue recognition fraud. Are there other indicators
that a revenue recognition fraud was under way?
References SEC, 2003a, Litigation Release No. 18205A, Accounting and Auditing Enforcement Release No.
1808A, “SEC Charges Peregrine Systems, Inc. with Financial Fraud and Agrees to Partial
Settlement,” US Securities and Exchange Commission, June 30.
SEC, 2003b, Litigation Release No. 18093, Accounting and Auditing Enforcement Release No.
1759, “SEC Charges Former Peregrine CFO with Financial Fraud,” US Securities and Exchange
Commission, April 16.
Waters, R., 2003, “Irregularities at Peregrine lifted revenues 60 percent,” Financial Times, March
3, p.19.
8.6 Analytical Procedures During Different Phases In The Audit
Process
Analytical procedures are used: (a) to assist the auditor in planning the nature, timing, and
extent of audit procedures; (b) as substantive procedures; and (c) as an overall review of the
financial statements in the final stage of the audit. The auditor is required to apply analytical
procedures at overall review stages of the audit.
Planning
Analytical procedures performed in the planning stage (Stage II in the Audit Process Model,
Illustration5.1) are used to identify unusual changes in the financial statements, or the absence of
expected changes, and specific risks. During the planning stage, analytical procedures are usually
focused on account balances aggregated at the financial statement level and relationships.
Application of analytical procedures at the planning stage indicates aspects of the business of
which the auditor was unaware and will assist in determining the nature, timing, and extent of other
308869907
308869907
28
audit procedures. Surveys of auditors show that the most extensive use of analytical procedures has
been in the planning and completion stages.14
Substantive Testing
During the substantive testing stage (Phase III of the Audit Process Model, Illustration 5.1),
analytical procedures are performed to obtain assurance that financial statement account balances
do not contain material misstatements. In substantive testing, analytical procedures focus on
underlying factors that affect those account balances through the development of an expectation of
how the recorded balance should look.
Overall Review
Analytical procedures performed during the overall review stage (Phase IV of the Audit
Process Model, Illustration5.1) are designed to assist the auditor in assessing that all significant
fluctuations and other unusual items have been adequately explained and that the overall financial
statement presentation makes sense based on the audit results and an understanding of the business.
According to ISA 520 “The auditor shall design and perform analytical procedures near the
end of the audit that assist the auditor when forming an overall conclusion as to whether the
financial statements are consistent with the auditor’s understanding of the entity.”15 Analytical
procedures at the review stage are intended to corroborate conclusions formed during the audit of
individual components of the financial statements. Moreover, they assist in determining the
reasonableness of the financial statements. They may also identify areas requiring further
procedures.
308869907
308869907
29
Tests of Controls Over Information Used for Analytics
An important consideration in applying analytical procedures is tests of controls over the
preparation of information used for analytics. When those controls are effective, the auditor will
have more confidence in the reliability of the information and, therefore, in the results of analytical
procedures.
The controls over non-financial information can often be tested in conjunction with tests of
accounting-related controls. For example, a company’s controls over the processing of sales
invoices may include controls over the recording of unit sales; therefore an auditor could test the
controls over the recording of unit sales in conjunction with tests of the controls over the
processing of sales invoices.
8.7 Analytical Procedures As Substantive Tests
Substantive procedures in the audit are designed to reduce detection risk relating to specific
financial statement assertions. Substantive tests include tests of details (either of balances or of
transactions) and analytical procedures. Auditors use analytical procedures to identify situations
that require increased use of other procedures (i.e. tests of control, substantive audit procedures),
but seldom to reduce audit effort.16
Analytical Procedures Instead of Tests of Details
There are a number of advantages of performing substantive analytical procedures instead of
tests of details. One advantage is that the auditor may use his understanding of the client’s business
obtained during planning procedures. The key factors affecting business may be expected to reflect
underlying financial data. Substantive analytical procedures often enable auditors to focus on a few
key factors that affect the account balance. Substantive analytical procedures may be more efficient
308869907
308869907
30
in performing understatement tests. For example, in a test for unrecorded sales, it may be easier to
develop an expectation of sales and investigate any significant differences between the expectation
and the recorded amount, than to sample statistically a reciprocal population and then perform tests
of details.
In planning an audit that uses analytical procedures as substantive procedures, the auditor
should17:

Determine the suitability of particular substantive analytical procedures for given assertions,
considering the assessed risks of material misstatement;

Evaluate the reliability of data from which the auditor’s expectation is developed, taking
account of quality of controls, as well as the source, comparability, and relevance of the
information

Evaluate whether the expectation is sufficiently precise to identify a misstatement that may
cause the financial statements to be materially misstated; and

Determine the amount of any difference of recorded amounts from expected values that is
acceptable without further investigation
Disadvantages of Analytical Procedures
Substantive analytical procedures have some disadvantages. They may be more timeconsuming initially to design and might be less effective than performing tests of details of
balances. Obtaining data used to develop an expectation and ensuring the reliability of that data at
a disaggregated level can take a substantial amount of the time otherwise spent performing tests of
details. Analytical procedures may be less effective when applied to the financial statements as a
whole than when applied to financial information on individual sections of an operation or to
financial statements of components of a the company.
308869907
308869907
31
Substantive analytical procedures will not necessarily deliver the desired results every year. In
periods of instability and rapid change, it may be difficult to develop a sufficiently precise
expectation of the recorded amount, and it may be more appropriate to apply tests of details. For
example, if an economy reaches hyperinflation, it is unlikely that we will be able to develop
meaningful expectations efficiently, except in limited circumstances.
Corroboration
When analytical procedures serve as substantive tests, the auditor should corroborate
explanations for significant differences by obtaining sufficient audit evidence. For example, recalculation of invoice extensions (quantity multiplied by price) may be corroborated by
interviewing a salesperson about how invoices are filled out. This evidence needs to be of the same
quality, as the evidence the auditor would expect to obtain to support tests of details.
To corroborate an explanation, one or more of the following techniques may be used:

inquiries of persons outside the client’s organization including bankers, suppliers, customers,
etc.;

inquiries of independent persons inside the client’s organization (e.g. an explanation received
from the Chief Financial Officer for an increase in advertising expenditures might be
corroborated with the marketing director: it is normally inappropriate to corroborate
explanations only by discussion with other accounting department personnel);

evidence obtained from other auditing procedures;

examination of supporting evidence. The auditor may examine supporting documentary
evidence of transactions to corroborate explanations. For example, if an increase in cost of
sales in one month was attributed to an unusually large sales contract, the auditor might
examine supporting documentation, such as the sales contract and delivery dockets.
308869907
308869907
32
Substantive Analytical Procedures Examples
Fraudulent payments are often in large amounts. An analytical procedure to detect this is the
use of a CAATs program (discussed in the next section) to stratify the payments by size and then
extract all large payments. The auditor may sort the records by type of purchase, since the size of
an expenditure is related to the typical cost of the product or service.
By analyzing revenue over at least three years, the auditor can detect unexpected trends in
revenues. Sales can also be analyzed by type, activity, salesperson, month, or customer. The sales
data can be stratified to determine if sales in a certain area or by a certain salesperson are made up
of a few large or unusual transactions.
Benford’s Law calculations may be done by a CAATs. Benford’s Law determines the
expected frequency for each digit in any position in a set of random numbers. This means that the
chances of any number appearing in a given database are mathematically predictable. Since the
expected frequency for each number in the set is known, every number that appears in the database
in excess of the expected frequency requires further investigation. For instance, payment amounts
authorized by a manager may be consistently just below the maximum allowed for that manager.
Payroll
If the auditor suspects fraud in the payroll area, he may do a number of substantive tests to
detect a “ghost employee,” an employee who still “works for the company” even though his
employment has been terminated, or excessive overtime charges. Three types of analytical tests can
be performed to help detect these kinds of irregularities: duplicate and validity tests, exception
testing, and recalculations.
Duplicate and validity tests are used to detect a ghost or a terminated employee. A CAATs
program can help the auditor check for duplicate social security numbers, names, or, if direct
308869907
308869907
33
deposit is used, bank account numbers. To find ghost employees, an auditor can also identify
employees who take no sick or annual leave, or those who do not have insurance or other
deductions taken out of their pay. Additionally, a CAATs program can verify that each employee’s
salary or wage is within the ranges for his job description and that tax withholding amounts are
reasonable.
8.8 Computer Assisted Audit Techniques (CAATs) and Generalized
Audit Software (GAS)
[Note to Philip: Contact PwC re: new research.]
The use of computer assisted audit techniques (CAATs) may enable more extensive testing of
electronic transactions and account files18. CAATs can be used to select sample transactions from
key electronic files, to sort transactions with specific characteristics, or to test an entire population
instead of a sample. CAATs generally include regression and statistical analysis as well as the more
widely used file interrogation techniques using generalized audit software (GAS) such as data
manipulation, calculation, data selection, data analysis, identification of exceptions and unusual
transactions.
Regression Analysis
Complicated analytical procedures may use regression analysis. Regression analysis is the
use of statistical models to quantify the auditor’s expectation in financial (euro, dollar) terms, with
measurable risk and precision levels. For example, an expectation for sales may be developed based
on management’s sales forecast, commission expense, and changes in advertising expenditures.
Regression analysis provides a very high level of precision because an explicit expectation is
formed in which the relevant data can be incorporated in a model to predict current year sales.
308869907
308869907
34
Regression analysis potentially can take into account all of the relevant operating data (sales
volume by product), changes in operations (changes in advertising levels, changes in product lines
or product mix), and changes in economic conditions. Regression analysis provides the benefits of
statistical precision. The statistical model provides not only a “best” expectation given the data at
hand, but also provides quantitative measures of the “fit” of the model.
Generalized Audit Software (GAS)
Generalized audit software (GAS) packages contain numerous computer-assisted audit
techniques for both doing analytical procedures and statistical sampling bundled into one piece of
software. There are widely used GAS packages such as ACL19 and Idea, and the Big Four audit
firms have their own software such as Deloitte and Touche’s STAR and MINI MAX. GAS
packages provide the auditors with the ability to access, manipulate, manage, analyze, and report
data in a variety of formats. This software allows the auditor to move from analytical procedures to
statistical sampling for analytical procedures fairly easily.
File Interrogation Procedures Using GAS
Using GAS in an audit requires converting client data into a common format and then
analyzing the data. This is generally referred to as file interrogation. File interrogation is a CAAT
that allows the auditor to perform automated audit routines on client computer data. It is a method
of using a computer to capture accounting data and reports and test the information contained
therein. Because the nature of audit evidence changes, audit techniques that take advantage of
technology are often more appropriate than traditional auditing techniques. In sophisticated
environments, file interrogation techniques can often create efficiency and improve the quality of
audit work.
308869907
308869907
35
Extracting information that meets specific criteria, selecting information, and testing the
accuracy of calculations can be done with file interrogation. CAATs used in GAS allows auditors
to analyze and test every item on a report to determine whether it meets predefined criteria, identify
significant differences, create an independent report of exceptions, select an audit sample, and
export the results into audit software.
Audit Tasks
In general terms, file interrogation can accomplish the following six types of audit tasks:
1
convert client data into common format;
2
analyze data;
3
compare different sets of data;
4
confirm the accuracy of calculations and make computations;
5
sample statistically;
6
test for gaps or duplicates in a sequence.
We will discuss the first three audit tasks in this section. For a discussion of the last three
items (confirm accuracy of the calculations, statistical sampling, and tests for gaps) see Appendix
A to Chapter 8 Audit Sampling and Other Selective Testing Procedures.
Convert Data Into a Common Format
The auditor can use GAS (e.g. ACL)) to convert client data into a common format (e.g. files
ending in the extension *.fil) that can be manipulated by the software. Audit work is more efficient
since data does not have to be manually entered and the conversion is usually performed with 100
percent accuracy. The data can then be used by the GAS or exported as several formats such as
plain text, comma delimited, XML, Microsoft Word, Excel, or Access.
308869907
308869907
36
Analyze Data
GAS can handle large volumes of data quickly. Often, file interrogation can be used to analyze
an entire population in less time than it would take to test a sample of items manually. The auditor
can identify all records in a data file that meet specified criteria or reformat and aggregate data in a
variety of ways.
Audit tests that can be performed with GAS include the following:

Identify all inventory items relating to products no longer sold.

Select all inventory items with no recorded location.

Summarize inventory items by location to facilitate physical observation.

Review account receivable balances for amounts over credit limits or older than a specified
period.

Summarize accounts receivable by age for comparison to the client’s schedules.

Review inventory quantities and unit costs for negative or unusually large amounts.

Isolate all inventory items that have not moved since a specified date.

Review assets for negative net book values.

Summarize inventory by age to assess the reasonableness of obsolescence provisions.
Compare Different Sets of Data
If records on separate files contain comparable data, GAS can be used to compare the different
sets of data. For example, the auditor could compare the following:

changes in accounts receivable balances between two dates with the details of sales and cash
receipts on transaction files;

payroll details with personnel records;
308869907
308869907

37
current inventory files with prior period files to identify potentially obsolete or slow-moving
items;

portfolio positions recorded in the accounting records of an investment company with the
records maintained by the custodian.
Structured Approach for GAS-Based Analytical Procedures
To use analytical procedures in testing an account balance with GAS, the auditor is likely to
follow the basic four-phase audit review model already discussed in section 8.3. The four phases
using GAS are shown in Illustration 8.5. When using GAS to conduct the four phase audit review
process, the auditor must first format the data so that it might be read with the software.
ILLUSTRATION 8.5
THE FOUR-PHASE ANALYTICAL REVIEW PROCESS USING GAS
Phase one in performing analytical procedures – expectations

Determine appropriate base data and an appropriate level of disaggregation.

Use regression analysis techniques to develop from the base data a plausible relationship
(a regression model) between the amounts to be tested (the test variable such as accounts
receivable balance) and one or more independent sets of data (predicting variables such as
revenue, volume of shipments, collection history, selling square footage, number of customers,
etc.) that are expected to relate to the test variable.

Based on this relationship, use GAS software to calculate the expectations (regression
estimates) for the test variable based on the current-period values of the predicting variables.
Phase two in performing analytical procedures – identification

Use GAS’s statistical techniques to assist in identifying significant differences for
investigation (i.e. differences exceeding the materiality thresholds) based on the regression
308869907
308869907
38
model, audit judgments as to monetary precision (MP), required audit assurance (R factor), and
the direction of the test.
Phase three in performing analytical procedures – investigation

Investigate and corroborate explanations for significant differences between the
expectations and the recorded amounts.
Phase four in performing analytical procedures – evaluation

Evaluate findings and determine the level of assurance, if any, to be drawn from the
analytical procedures.
Examples
To illustrate, assume that the test variable is sales and the predicting variable is cost of sales.
In most cases, it is plausible that a relationship exists between these two components of the income
statement. The GAS uses sales and cost of sales data from prior periods to determine the precise
nature of the relationship and to develop a regression model. Then, based on audited current-period
amounts for cost of sales, GAS will project the current-period expectations for sales. GAS will then
compare the expectations for sales with the actual recorded amounts and calculate the differences.
GAS identifies for investigation any statistically significant differences between the current-period
projected amounts and the actual recorded amounts for sales.
A regression model might be constructed for a chain of retail outlets based on prior-year
performance, in which annual sales per outlet are related to floor area and number of sales
personnel (relationships that are expected to be reasonably constant over time). That model might
then be used to develop sales expectations based on the corresponding current-year floor area and
sales personnel data. The application would identify those outlets where sales results require
further investigation, and may assist in selecting locations to visit.
308869907
308869907
39
8.9 Analytical Procedures Using Data Mining Techniques
Data mining is a set of computer-assisted techniques that use sophisticated statistical analysis,
including artificial intelligence techniques, to examine large volumes of data with the objective of
indicating hidden or unexpected information or patterns. In database terms, data mining is referred
to as knowledge discovery in databases (KDD). Data mining can be used in all types of databases
or other information repositories. Data to be mined can be numerical data, textual data or even
graphics and audio.
Used most extensively in customer relationship management (CRM) and fraud detection,
data mining is for both verification and discovery objectives. Data mining is used in a top-down
approach to verify auditors’ expectations or explain events or conditions observed. For example,
merchandise order and delivery dates are examined to see if the delivery date falls after the order
date. Discovery is a bottom-up approach that uses automated exploration of hitherto unknown
patterns. For example, the auditor uses a neural network to sift through financial and nonfinancial revenue and accounts receivable data to discover unusual patterns.
GAS and Data Mining
GAS’s capability to assist in the overall audit process while requiring little technical skill is a
major reason for its success. However, GAS has been criticized because it makes some tasks easier
but it cannot complete any data analysis by itself. Data mining, on the other hand, analyzes data
automatically but is more difficult to employ.
Data mining tools remain promising in a variety of application areas. With the development of
appropriate data mining tools for the auditing profession, it may be expected to replace some
professional expertise required in certain auditing processes. For now, although data mining
308869907
308869907
40
procedures are useful in almost all steps of the audit process, its most practical and useful
applications are in analytical procedures.
Data mining may use many methods and techniques and algorithms to analyze client data.
Data mining methods include data description, dependency analysis, classification and prediction,
cluster analysis, outlier analysis and evolution analysis The most frequently used algorithms are
decision trees, apriori algorithms, and neural networks.
The purpose of dependency analysis is to search for the most significant relationship across
large number of variables or attributes. Classification is the process of finding models, also known
as classifiers, or functions that map records into one of several discrete prescribed classes.
The objective of data description is to provide an overall description of data, either in itself
or in each class or concept. There are two main approaches in obtaining data description – data
characterization and data discrimination. Data characterization is summarizing general
characteristics of data and data discrimination, also called data comparison, by comparing
characters of data between contrasting groups or classes.
The objective of evolution analysis is to determine the most significant changes in data sets
over time. In other words, it is other types of algorithm methods (i.e. data description, dependency
analysis, classification or clustering) plus time-related and sequence-related characteristics.
The objective of cluster analysis is to separate data with similar characteristics from the
dissimilar ones. The difference between clustering and classification is that while clustering does
not require pre-identified class labels, classification does. Outliers are data items that are distinctly
dissimilar to others and can be viewed as noises or errors. However, such noises can be useful in
some cases, such as fraud detection, where unusual items or exceptions are major concerns.
An example of the use of the clustering method is when the auditor clusters accounting
transactions in such categories as assets, liabilities, revenue, expenses, etc. This might reveal those
308869907
308869907
41
small transactions that occur repeatedly in a certain period of the month or the same transactions
recorded in different account numbers. The auditor might find that sales in some months or
divisions are excessively higher or lower than the normal. Expenses that are highly variable during
the year might be found. Clustering might show the repeated purchase of the same fixed asset.
Loans that are transacted between related companies and subsidiaries may be uncovered.
Algorithms – Decision Tree, Apriori, Neural Network
Data mining most frequently uses three algorithms.

A decision tree is a predictive model that classifies data with a hierarchical structure. It
consists of nodes, which contain classification questions, and branches that are the result of the
questions. (For example the question, “Does this item increase at the same rate as revenue?”
may be answered yes – which leads to one branch “growth similar to revenue” or may be
answered no – which leads to another branch.)

The apriori algorithm attempts to discover frequent item sets using rules to find associations
between the presence or absence of items (Boolean association rules). The group of item sets
that most frequently come together is identified.

A neural network is a computer model based on the architecture of the brain. It first detects a
pattern from data sets then predicts the best classifiers of that pattern, and finally learns from
the mistakes.
8.10 Follow-Up in Case of Unexpected Deviations
When analytical procedures identify significant fluctuations or relationships that are
inconsistent with other relevant information or that deviate from predicted amounts, the auditor
should investigate and obtain adequate explanations and appropriate corroborative evidence. 20 A
308869907
308869907
42
comparison of actual results with expected should include a consideration of why there is a
difference.
There are primarily two reasons for a significant fluctuation or inconsistency. One is that there
is a genuine business reason that was not obvious during planning procedures. The second reason is
that there is a misstatement. Work must be done to determine which reason.
Investigation
The investigation of unusual fluctuations and relationships ordinarily begins with inquiries of
management, followed by corroboration of management responses and determination if additional
audit procedures are needed. Management’s responses may be corroborated by comparing them
with the auditor’s knowledge of the business and other evidence obtained during the course of the
audit.
If a reasonable explanation cannot be obtained, the auditor aggregates misstatements that the
entity has not corrected. The auditor would then consider whether, in relation to individual
amounts, subtotals, or totals in the financial statements, they materially misstate the financial
statements taken as a whole. If management cannot provide a satisfactory explanation and there is a
possibility of material misstatement, other audit procedures should be determined.
Upon finding unexpected deviations that exceed the threshold, there may be a need to do
some root cause analysis. In that root cause analyses, a reassessment of the effectiveness of controls
may be necessary. If the auditor detects deviations from controls upon which he intends to rely, he
must make specific inquiries to understand these matters and their potential consequences.
Furthermore, he, must determine whether:21

The tests of controls show an appropriate basis for reliance on the controls;

Additional tests of controls are necessary; or
308869907
308869907

43
The potential risks of misstatement need to be addressed using substantive procedures.
8.11 Summary
Analytical procedures are evaluations of financial information through analysis of plausible
relationships among both financial and non-financial data. Analytical procedures also encompass
such investigation as is necessary of identified fluctuations or relationships that are inconsistent
with other relevant information or that differ from expected values by a significant amount. That is,
analytical procedures entail the use of comparisons and relationships to determine whether account
balances or other data appear reasonable. A basic premise of using analytical procedures is that
there exist plausible relationships among data and these relationships can reasonably be expected to
continue.
General analytical procedures include trend analysis, ratio analysis, statistical and data mining
analysis, and reasonableness tests. Trend analysis is the analysis of changes in an account balance
over time. Ratio analysis is the comparison of relationships between financial statement accounts,
the comparison of an account with non-financial data, or the comparison of relationships between
firms in an industry. Reasonableness testing is the analysis of account balances or changes in
account balances within an accounting period in terms of their “reasonableness” in light of
expected relationships between accounts. Data mining is a set of computer-assisted techniques that
use sophisticated statistical analysis, including artificial intelligence techniques, to examine large
volumes of data with the objective of indicating hidden or unexpected information or patterns. For
these tests auditors generally use computer-aided audit software (CAATs).
The process of planning, executing and drawing conclusions from analytical procedures is
called analytical review. The four-phase process consists of the following:
1. phase one is to formulate expectations (expectations);
308869907
308869907
44
2. phase two is the comparison of the expected value to the recorded amount (identification);
3. phase three requires investigation of possible explanations for a difference between expected
and recorded values (investigation);
4. phase four involves evaluation of the impact on the audit and the financial statements of the
differences between expectation and recorded amounts (evaluation).
Expectations are developed by identifying plausible relationships that are reasonably expected
to exist based on the auditor’s understanding of the client and of his industry. These relationships
may be determined by comparisons with the following sources:

comparable information for prior periods;

anticipated results (such as budgets and forecasts, or auditor expectations);

elements of financial information within the period;

similar industry information;

non-financial information.
Determining which type of analytical procedure is appropriate is a matter of professional
judgment. A review of audit practice indicates that simple judgmental approaches (such as
comparison and ratio analysis) are used more frequently than complex statistical approaches (such
as time series modeling or regression analysis. These tests are generally carried out using computer
software (i.e. CAATs).
At the planning stage of an audit, there are certain customary ratios that are always calculated
to determine accounts that may represent significant risks to the entity of liquidity, solvency,
profitability, and activity. Illustration 8.3 (earlier) gives a list of these ratios.
Analytical procedures are used: (a) to assist the auditor in planning the nature, timing, and
extent of audit procedures; (b) as substantive procedures; and (c) as an overall review of the
financial statements in the final stage of the audit. The auditor is required to apply analytical
308869907
308869907
45
procedures at the overall review stages of the audit. In the substantive testing stage of the audit,
analytical procedures are used to obtain evidence to identify misstatements in account balances and
thus to reduce the risk of misstatements. In the overall review stage, the objective of analytical
procedures is to assess the conclusions reached and evaluate the overall financial statement
presentation.
Substantive procedures in the audit are designed to reduce detection risk relating to specific
financial statement assertions. Substantive tests include tests of details and analytical procedures.
Auditors use analytical procedures to identify situations that require increased use of other
procedures (i.e. tests of control, substantive audit procedures), but seldom to reduce audit effort.
There are a number of advantages of performing substantive analytical procedures instead of tests
of details. The key factors affecting business may be expected to reflect the underlying financial
data. Substantive analytical procedures often enable auditors to focus on a few key factors that
affect the account balance.
The use of computer-assisted audit techniques (CAATs) may enable more extensive testing of
electronic transactions and account files. Computer-assisted audit techniques can be used to select
sample transactions from key electronic files, to sort transactions with specific characteristics, or to
test an entire population instead of a sample. CAATs generally include regression and statistical
analysis as well as the more widely used file interrogation techniques using generalized audit
software (GAS) such as data manipulation, calculation, data selection, data analysis, identification
of exceptions, and unusual transactions.
Data mining is a set of computer-assisted techniques that use sophisticated statistical analysis,
including artificial intelligence techniques, to examine large volumes of data with the objective of
indicating hidden or unexpected information or patterns. Data mining is used for both verification
and discovery objectives. It is used in a top-down approach to verify auditors’ expectations or
308869907
308869907
46
explain events or conditions observed. Discovery is a bottom-up approach that uses automated
exploration of hitherto unknown patterns.
If the auditor detects deviations from controls upon which he intends to rely, he must make
specific inquiries to understand these matters and their potential consequences. Furthermore, he
must determine whether the tests of controls show an appropriate basis for reliance on the controls;
additional tests of controls are necessary; or the potential risks of misstatement need to be
addressed using substantive procedures. When analytical procedures identify significant
fluctuations or relationships that are inconsistent with other relevant information or that deviate
from predicted amounts, the auditor should investigate and obtain adequate explanations and
appropriate corroborative evidence. A comparison of actual results with expected should include a
consideration of why there is a difference. The investigation of unusual fluctuations and
relationships ordinarily begins with inquiries of management followed by corroboration of
management’s responses and other audit procedures based on the results of these inquiries.
8.12 Notes
IFAC, 2012, International Standards on Auditing 520 (ISA 520), “Analytical Procedures,” para 4.
Handbook of International Quality Control, Auditing, Review, Other Assurance, and Related Services
Pronouncements Part 1 2012 Edition. International Federation of Accountants, New York.
1
2
Ibid, ISA 520, paragraph 6.
See Blocher, E. and Cooper, J., 1988, “A study of auditors’ analytical review performance,” Auditing: A
Journal of Practice & Theory, Spring, pp.1–28; and Koonce, L., 1993, “A cognitive characterization of audit
analytical review,” Auditing: A Journal of Practice & Theory 12 (Supplement), pp.57–76.
3
Asare, S. and Wright, A., 1997, “Hypothesis revision strategies in conducting analytical procedures,”
Accounting, Organizations and Society 22, November, pp.737–55.
4
5
AICPA, 2012. AICPA Audit Guide Analytical Procedures. American Institute of Certified Public
Accountants. New York
6
Ibid ISA 520. Paragraph 5(b).
308869907
308869907
47
7
See: (1) Bell, T.B., Marrs, F.O., Solomon, I., and Thomas, H., 1997, Auditing Organizations Through A
Strategic-Systems Lens, New York, NY: KPMG Peat Marwick LLP, (2) Loebbecke, J.K. and Steinbart, P.J.,
1987, “An investigation of the use of preliminary analytical review to provide substantive audit evidence,”
Auditing: A Journal of Practice & Theory, Spring, pp.74–89; (3) Wild, J.J., 1987, “The prediction
performance of a structural model of accounting numbers,” Journal of Accounting Research, Vol.25, no.1,
pp.139–60.
8
Ibid. ISA 520. Paragraph 5a
AICPA, 2010, AU Section 329, “Substantive Analytical Procedures”, paragraph 5. American Institute of
Certified Public Accountants.
9
Biggs, S.F., Mock, T.J., and Simnett, R., 1999, “Analytical Procedures: Promise, Problems and Implications
for Practice,” Australian Accounting Review, Vol.9, No.1, pp.42–52.
10
11
See Standard & Poors. NetAdvantage. 2012. McGraw Hill; IBISWorld. 2012. http://www.ibisworld.com.
Los Angeles; RMA, 2012, Annual Statement Studies: Financial Ratio Benchmarks, Risk Management
Association, Philadelphia.
IFAC, 2012, International Standards on Auditing 570, “Going Concern”, para2. Handbook Of International
Quality Control, Auditing, Review, Other Assurance, And Related Services Pronouncements Part I 2012
Edition , . International Federation of Accountants, New York.
12
13
Ibid. ISA 570 paragraph A2.
See, for example, Booth, P. and Simnett, R., 1991, “Auditors’ perception of analytical review procedures,”
Accounting Research Journal, Spring, pp.5–10.
14
15
Ibid. ISA 520, paragraph 6.
See, for example, Bedard, J., 1989, “An archival investigation of audit program planning,” Auditing: A
Journal of Practice and Theory, Fall, pp.57–71.
16
17
Ibid. ISA 520. Paragraph 5.
IFAC, 2012, International Standards on Auditing 330, “The Auditor’s Responses To Assessed
Risks”. para A16. Handbook Of International Quality Control, Auditing, Review, Other Assurance, And
Related Services Pronouncements Part I 2012 Edition , . International Federation of Accountants, New York.
18
19
20
21
ACL software is developed by ACL Services, Ltd. (www.acl.com).
Ibid ISA 330. Para 17
Ibid. ISA 330. Paragraph 17.
308869907