Proceedings of the 2010 Conference on Belief Functions, Brest, France, April 1-2, 2010.

Audit Risk Formula with Mixed Evidence

Rajendra P. Srivastava1, and Theodore J. Mock2
1
School of Business, The University of Kansas, Lawrence, Kansas 66045
Email: rsrivastava@ku.edu
2
Anderson Graduate School of Management, University of California, Riverside, CA and
University of Maastricht, The Netherlands
Email: tmock@ucr.edu

Abstract- The financial statement audit is the process of balance, a class of transactions, or a disclosure, and
collecting, evaluating, and aggregating relevant items of issued an opinion that the financial statements are fairly
evidence pertaining to various management assertions presented.
related to the financial statement accounts to determine The AICPA audit risk model suggests that audit risk is
whether the company’s financial statements present fairly the product of four risks:
its financial position. The Dempster-Shafer theory [1] of • IR, the inherent risk that a material misstatement
belief functions has been argued to be an appropriate associated with an assertion is present due to
framework for representing uncertainties in the audit. inherent nature of the account, class of
This note extends Srivastava and Shafer [2] by deriving
transactions, or the disclosure.
an audit risk formula under the Dempster-Shafer theory
• CR, the control risk that internal accounting
for more realistic situations where the auditor has mixed
items of evidence. In their derivation, Srivastava and controls has failed to prevent or detect and correct
Shafer assume only positive evidence. In addition, their a material misstatement relevant to the assertion.
work is extended by considering interrelationships among • AP, the analytical procedures risk that analytical
the balance sheet accounts and the transaction streams procedures have failed to detect material
accounts. Such interrelationships which are prominent in misstatements relevant to the assertion and
practice, were not considered by Srivastava and Shafer. • TD, the test-of-details risk that the audit
procedures have failed to detect material
Key Words: Financial Statement Audit, Audit Risk, misstatements relevant to the assertion.
Audit Planning and Evaluation, Dempster-Shafer Theory, While the above audit risk model provides a way to
Belief Functions assess the risk of material misstatement pertaining to an
assertion, it does not provide an appropriate way to
I. INTRODUCTION aggregate items of evidence for various accounts and
The financial statement audit of a company is the process transaction streams constituting the financial statements.
of collecting, evaluating, and aggregating relevant items of The overall aggregation process is left on the auditor’s
evidence pertaining to various management assertions related professional judgment.
to the balance sheet accounts and the corresponding Srivastava and Shafer [2] present analytical formulas
transaction stream (income statement) accounts to determine under Dempster-Shafer (DS) theory [1] which combine
whether the company’s financial statements present fairly its items of evidence at three levels: the accounts in the
financial position and financial performance. The auditing balance sheet, their respective assertions, and at the
profession world-wide has developed auditing standards to overall financial statement level. They demonstrated that
help auditors achieve the above goal. In the process, the audit the AICPA audit risk model may be interpreted as a
profession in each country has not only developed standards plausibility model under DS theory where each risk term
as to what kinds of evidence should be gathered but also has in the model represents the plausibility that material
provided a model to aggregate various items of evidence misstatement is present in the assertion being considered
gathered in the audit process. For example, the American in the model. For example, CR may be interpreted as the
Institute of Certified Public Accountants (AICPA) in the plausibility that material misstatement is not prevented or
USA has provided the following audit risk model [3]: AR = detected and corrected by internal accounting controls
IR.CR.AP.DR, for assessing audit risk. AR is the risk that relevant to each assertion.
the auditor has failed to detect material misstatements Although the general evidential diagram for the audit
relevant to an assertion1 pertaining to either an account process is a network [5], Srivastava and Shafer used a
tree type evidential diagram to develop the analytical
1
Management assertions are the implicit assertions made by formulas for audit risk and considered only affirmative
the management when they publish company’s financial
information to imply that the financial information presented company. The AICPA [4] has published these assertions
in the report present fairly the financial position of the as listed Table 1.
1
Proceedings of the 2010 Conference on Belief Functions, Brest, France, April 1-2, 2010.
items of audit evidence, that is evidence that supports the variable on the left with the variables on the right. These
variable being evaluated. Also, they did not consider the variables represent the various assertions, as given in
interrelationships between the accounts in the balance sheet Table 1, for which the auditor collects evidence to
and the transaction streams although it is clear from the ascertain that they are true and ultimately to make the
nature of double-entry accounting systems that these are judgment as to whether the overall financial statements of
interrelated. Thus, consideration of such interrelationships is the company are free or are not free from material
important for developing a more comprehensive analytical misstatements.
formula for assessing the overall audit risk. Auditing
standards [3, 4] also suggest such interrelationships exist in Table 1
practice. Management assertions based on AICPA [3] with
In the present paper, we extend Srivastava and Shafer [2] modifications to make the assertions generic to an
by developing an audit risk formula, which incorporates account on the balance sheet.
mixed items of evidence and the interrelationships among the
accounts in the balance sheet and the transaction streams. Sub Assertions related to an Account or
Main Assertions
While the Srivastava and Shafer models are useful for Class of Transactions
planning purposes, our model is useful for both planning and Ai.1: Existence. The ith balance sheet
account (asset, liabilities, or equity) exists,
evaluation purposes. For the planning phase, Srivastava and
that is it is not fictitious.
Shafer have argued that the auditor assumes positive items of Ai.2: Rights and obligations. For the ith
evidence. However, given that both negative and mixed items account, the entity holds or controls the
of evidence may be encountered in an audit, a framework that The ith Account rights to an asset account, or has the
incorporates both types of evidence may be used for both (Ai) in the obligation for a liability account.
Balance Sheet is Ai.3: Completeness. All assets, liabilities,
planning and evaluation. Srivastava et al [5] have developed
free from material and equity interests relevant to ith account
a fraud risk assessment model that considers mixed evidence. misstatements. that should have been recorded have been
i∈{1, 2, 3, n} recorded.
II. AUDIT PROCESS AND EVIDENTIAL DIAGRAM Ai.4: Valuation and allocation. The ith
account is included in the financial
In a typical audit, the auditor has evidence at various levels statements at the appropriate amount and any
of the financial statements: resulting valuation or allocation adjustment
• Evidence at the overall financial statements level such as is appropriately recorded.
information concerning management integrity and the CAi.j.k.1: Occurrence. The kth class of
The kth Class of transactions and events related to assertion
economic environment, transactions and Ai.j that have been recorded have occurred
• Evidence at the individual balance sheet account level events (CAi.j.k) and pertain to the entity.
such as industry norms for the balances, for the period CAi.j.k.2: Completeness. All transactions
• Evidence at the transaction streams or class of under and events related to kth class of transactions
transactions level such as ratio analyses of sales and cash audit related to and events pertaining to assertion Ai.j that
assertion Ai.j is should have been recorded have been
receipts which are related to the accounts receivable free from material recorded.
balance in the balance sheet misstatement. CAi.j.k.3: Accuracy. Amounts and other
• Evidence at the assertion level such as physical counts of j∈{1, 2, 3, 4} data relating to recorded kth transactions and
inventory for the assertion that inventory exists and is and events pertaining to assertion Ai.j have been
valued correctly and k∈{1, 2, 3, qij} recorded accurately.
where qij is the CAi.j.k.4: Cutoff. The kth class of
• Evidence related to disclosure requirements at the overall transactions and events related to assertion
total number of
financial statement level or individual account and classes of Ai.j have been recorded in the correct
transaction stream level such as evidence about the sale transactions accounting period.
of an important subsidiary after the year-end but before associated with CAi.j.k.5: Classification. The kth class of
the completion of the audit. the assertion Ai.j. transactions and events relevant to Ai.j have
been recorded in the proper accounts.
A. Evidential Diagram and the Audit Process Logic DAi.1: Occurrence and rights and
An evidential diagram of an audit can be represented as a obligations. Disclosed events and
transactions relevant to account Ai have
network of variable nodes [6] where the variables include the occurred and pertain to the entity.
overall financial position of the company (i.e. the balance The presentation
DAi.2: Completeness. All disclosures that
sheet), the individual asset, liability and equity accounts and disclosures
should have been included related to account
which constitute the balance sheet, the management related account Ai Ai have been included.
(DAi) are
assertions associated with these accounts, and the DAi.3: Classification and understand-
appropriate.
corresponding classes of transactions and their management ability. Financial information related to
assertions. In Figure 1, the rectangular boxes with rounded account Ai is appropriately presented and
described and disclosures are clearly
corners represent the variable nodes and the circles with the
expressed.
symbol ‘&’ represent the ‘AND’ relationship between the
2
Proceedings of the 2010 Conference on Belief Functions, Brest, France, April 1-2, 2010.
All variables in Figure 1 are binary variables. We use the Step 1: Assessing the Belief Masses at Each Variable
upper case letters to represent the names of the variables, and Since we are considering multiple items of evidence for
the lower case letters for their values. For example, the each binary variable in the evidential network, we use
variable B represents the assertion that the overall balance Dempster rule as simplified by Srivastava [7] to
sheet of the company is free from material misstatements. B determine the overall belief masses directly defined at
has two possible values, ‘b’ and ‘~b’ where ‘b’ represents the each variable by combining all evidence, say qX of them,
state where the assertion B is true and ‘~b’ represents the
directly bearing on the variable. We obtain the following
state where the assertion is not true. Consistent with + −
Srivastava and Shafer [2], we assert that the balance sheet is belief masses in support of X, m X , and against X, m X :
fairly stated if and only if all the accounts in the balance sheet qX
are free from material misstatements. In addition, we assert
+
mX = 1 − ∏ (1 − mi (x))/K X
, (1)
that all the disclosures at the overall level and at the account i =1
level must also be appropriate, that is consistent with the qX
standards of disclosure, which have recently been added to
−
mX = 1 − ∏ (1 − mi (~x))/K X
, (2)
the auditing standards [4]. The relationship between an i =1
qX
account balance and the corresponding classes of transactions
is represented by the ‘AND’ relationship as depicted in
Θ
mX = ∏ mi ({x, ~x) / K X . (3)
i=1
Figure 1.
where KX is the renormalization constant and is given by
In summary, the Figure 1 evidential diagram asserts that a
qX qX qX
balance sheet account, say Ai, is free from material
misstatements if and only if all its assertions (Ai.1, Ai.2,
KX = ∏(1 − mi (x)) + ∏(1− mi (~ x)) − ∏mi ({x, ~x}) . (4)
i =1 i =1 i=1
Ai.3, and Ai.4) are true and the corresponding disclosures are
made properly. Also, the jth assertion of account Ai is true if Step 2: Propagation of Beliefs from Level 5 to Level 4
and only if all the related classes of transactions are free from We use Srivastava, Shenoy, and Shafer [8] Proposition
material misstatements. All these relationships are depicted 1 to propagate the beliefs defined at the variables in level
through the ‘AND’ tree depicted in Figure 1. Because of lack 5 to the variables in level 4 (see Figure 1). According to
of space, we have not depicted the various items of evidence Proposition 1, the belief masses propagated from the
pertaining to each variable in the network. However, we variables at Level 5 to the variables at Level 4 can be
discuss these items of evidence in brief in the next section. written as:
q X5 qX5
B. Items of Evidence and Corresponding Belief Masses (m-
values)
+
m X4←all X5 = ∏ m X5 , and mX4←all X5 = 1-∏ (1- mX5 ) . (5)
+ − −

i=1 i =1
In general, the auditor obtains multiple items of evidence Where X4 represents all the variables in Figure 1 at Level
pertaining to each variable in Figure 1. We depict an item of 4, and X5 represents all the variables at Level 5. For
evidence with letter E with a subscript that indicates the example, if X4 = CAi.2.k then X5∈{CAi.2.k.1,
corresponding variable. For simplicity, let us denote a CAi.2.k.2, CAi.2.k.3, CAi.2.k.4, CAi.2.k.5, DAi.2.k}.
variable node by X with values {x, ~x} and the Next, we combine the beliefs propagated from Level 5 to
corresponding items of evidence directly bearing on that Level 4 variables with the beliefs directly bearing on
variable by EpX, where the subscript pX represents pth item of these variables at Level 4 as determined in Step 1. We
evidence directly bearing on X. The corresponding belief use Dempster rule as simplified by Srivastava [7] to
+ − combine the two sets of beliefs and obtain the following
masses are represented by symbols, m pX = mp(x), m pX =
belief masses at Level 4:
Θ ' + + +
mp(~x) and m pX = mp({x, ~x}), respectively, representing the m X4 = 1 − (1 − m X4←all X5 )(1 − m X4 ) / K X4 , (6)
belief masses, that is the strength or weight of the evidence, ' − − −
m X4 = 1 − (1 − m X4←all X5 )(1 − m X4 ) / K X4 , (7)
in support of the assertion x, in support of the negation of the
'Θ Θ Θ
assertion ~x, and the residue level of ambiguity, such that m X4 = m X4←all X5 m X4 / K X4 , (8)
+ − Θ + + − −
m pX + m pX + m pX = 1. KX4 = (1 − mX4← all X5 )(1 − mX4 ) + (1 − mX4← all X5 )(1 − mX4 )
(9)
Θ Θ
− mX4 ← all X5mX4
III. AUDIT RISK FORMULA AT THE BALANCE
SHEET LEVEL Step 3: Propagation of Beliefs from Level 4 to Level 3
To derive the analytical formula for audit risk at the overall This step is similar to Step 2. We again propagate the
balance sheet level, we propagate all the belief masses from beliefs from the variables at Level 4 to the variables at
all the variable nodes in the evidential diagram given in Level 3 by using Srivastava et. al [8] and then combine
Figure 1 to the main variable B. For this purpose, we these beliefs with the beliefs defined at each variables at
complete the following steps.
3
Proceedings of the 2010 Conference on Belief Functions, Brest, France, April 1-2, 2010.
Level 3 using Srivastava [7]. Here are the resulting belief Step 5: Propagation of Beliefs from Level 2 to Level 1
masses at the variables at Level 3: This step is the final step in the derivation of the audit
' + + +
m X3 = 1 − (1 − m X3←all X4 )(1 − m X3 ) / K X3 , (10) risk formula at the balance sheet level. We again use [8]
' − − −
to propagate the beliefs from Level 2 to Level 1 and
m X3 = 1 − (1 − m X3←all X4 )(1 − m X3 ) / K X3 , (11) combine these beliefs with the beliefs obtained from the
'Θ Θ
m X3 = m X3←all X4 m X3 / K X3 ,
Θ
(12) evidence directly bearing on the variable B at Level 1
using [7]. The resulting belief masses are:
+ + − −
KX3 = (1 − mX3← all X4 )(1 − mX3 ) + (1 − mX3← all X4 )(1 − mX3 ) mB
' + + +
= 1 − (1 − m B←all X2 )(1 − m B ) / K B , (26)
(13)
Θ Θ
− mX3← all X4mX3 , ' − − −
mB = 1 − (1 − m B←all X2 )(1 − m B ) / K B , (27)
where X3∈{Ai.1, Ai.2, Ai.3, Ai.4}. Similarly, the belief 'Θ Θ Θ
masses propagated from DAi.1, DAi.2, and DAi.3, to DAi mB = m B←all X2 m B / KB , (28)
when combined with the belief masses at DAi yields: + + − −
KB = (1 − mB←all X2 )(1 − mB ) + (1 − mB←all X2 )(1 − mB )
'+ + + (29)
mDAi = 1 − (1 − mDAi←{DAi.1,DAi.2,DAi.3} )(1 − mDAi ) / KDAi , (14) Θ Θ
− mB←all X4mB ,
'− − −
mDAi =1 − (1 − mDAi←{DAi.1,DAi.2,DAi.3} )(1 − mDAi ) / KDAi , (15) where X2∈{Ai, DB; i =1, 2, 3, .. n}.
'Θ Θ Θ
m DAi = m DAi ←{DAi.1,DAi.2,DAi.3}m DAi / K DAi , (16) Audit Risk Formula at the Balance Sheet Level
+ + From Equations (26) – (28), we obtain the following
KDAi = (1 − mDAi←{DAi.1,DAi.2,DAi.3} )(1 − mDAi )
beliefs that overall the balance sheet is free from material
− − misstatements, b, and has material misstatements, ~b:
+ (1 − mDAi←{DAi.1,DAi.2,DAi.3} )(1 − mDAi ) (17)
+ +
Θ Θ Bel total (b) = 1 − (1 − m B←all X2 )(1 − m B ) / K B , (30)
− mDAi←{DAi.1,DAi.2,DAi.3}mDAi ,
− −
Bel total (~b)= 1 − (1 − m B←all X2 )(1 − m B ) / K B . (31)
Step 4: Propagation of Beliefs from Level 3 to Level 2
The plausibility that material misstatements are present at
In this step, we propagate belief from the variables at Level the overall level is given by
3 to the variables at Level 2 again using Srivastava et al [8] + +
Pl total (~b) = (1 − m B← all X2 )(1 − m B ) / K B (32)
and then combine these beliefs with the beliefs directly
defined at the variables at Level 2 using Dempster’s rule as Srivastava and Shafer [2] argued that the plausibility that
simplified by Srivastava [7]. One obtains the following belief material misstatements are present in the balance sheet
masses at the variables at Level 2: defines the audit risk. Thus, based on their definition, the
' + + + overall audit risk formula at the balance sheet level is
m X2 = 1 − (1 − m X2←all X3 )(1 − m X2 ) / K X2 , (18)
given by (32).
' − − −
m X2 = 1 − (1 − m X2←all X3 )(1 − m X2 ) / K X2 , (19) + +
Audit Risk = (1 − m B←all X2 )(1 − m B ) / K B (33)
'Θ Θ Θ
m X2 = m X2←all X3 m X2 / K X2 , (20) +
where m B←all X2 is given in Table 2 and represents the
+ + − −
KX2 = (1 − mX2←all X3 )(1 − mX2 ) + (1 − mX2←all X3 )(1 − mX2 ) belief masses propagated from the variables at Level 2 to
(21) variable B.
Θ Θ
− mX2←all X3mX2 , Equation (33) along with Equations (1)-(29) presents a
where X2∈{A1, A2, …. An}. Similarly, the belief masses at general audit risk model which can be used by the auditor
the variable DB at Level 2 can be obtained by propagating not only for planning purposes but also for the evaluation
the beliefs from DB.1, DB.2, and DB.3 to DB using [8] and purpose. As mentioned earlier, during the planning phase
combining these beliefs with the beliefs at DB using [7]: since the auditor has no knowledge about the actual
' + + + nature of the audit evidence, he/she assumes the nature to
m DB = 1 − (1 − m DB←{DB.1, DB.2, DB.3} )(1 − m DB ) / K DB , (22)
be positive as argued by Srivastava and Shafer. However,
' − − −
m DB = 1 − (1 − m DB←{DB.1, DB.2, DB.3} )(1 − m DB ) / K DB , (23) while conducting the audit, the auditor may encounter all
'Θ Θ Θ kinds of evidence, positive, negative, and mixed. Thus,
m DB = m DB←{DB.1, DB.2, DB.3}m DB / K DB , (24) our approach provides an appropriate way to combine all
+
KDB = (1 − mDB←{DB.1, DB.2, DB.3} )(1 − mDB )
+ the evidence encountered on the actual audit and hence
helps assess the audit risk even at the completion of the
− −
audit.
+ (1 − mDB←{DB.1, DB.2, DB.3} )(1 − mDB ) (25)
Θ Θ
If the assessed audit risk is more than what the auditor
− mDB←{DB.1, DB.2, DB.3}mDB , can accept, say PlTotal(~b)>0.05, the auditor has several
Equations (18) through (25) determine the belief masses at all options. For example, the auditor can gather more or
the variables at Level 2. different audit evidence. If the belief that overall financial
statement is materially misstatement, i.e., BelTotal(~b), is
4
Proceedings of the 2010 Conference on Belief Functions, Brest, France, April 1-2, 2010.
still significantly high, say 0.10 or higher, then the auditor we extend the evidential diagram from just three levels as
can negotiate with the client and adjust the account balances treated by Srivastava and Shafer to two additional levels:
in error and reassess the audit risk. If the reassessed audit risk Classes of Transactions and the associated assertions and
is acceptable, then the auditor could provide an opinion that disclosure requirements. As noted, our audit risk formula
the financials are fairly stated. However, if the auditee is not is based on a more realistic setting where both positive
willing to adjust the account balances in error then the auditor evidence which confirms the management assertions) and
may issue an opinion that is either qualified, adverse or negative which implies that such assertions are not valid.
incorporate a disclaimer of opinion (see [9] for alternative Thus, importantly, the new risk formulations can be used
types of audit opinions). not only for planning purposes but also for evaluation
If, given additional audit evidence, the total belief, purposes
BelTotal(~b), is so high that additional evidence is unlikely to
decrease the audit risk, PlTotal(~b), to an acceptable level, and REFERENCES
if the client is not willing to adjust the accounts, then again [1] Glenn Shafer, A Mathematical Theory of Evidence,
the auditor could provide a qualified or adverse opinion Princeton University Press, 1976.
depending on the severity of the indicated errors. [2] Rajendra Srivastava, and G. Shafer, Belief-Function
As expected, audit risk as defined in Equation (33) and Formulas for Audit Risk. The Accounting Review.
assuming that all the items of evidence are positive, i.e., all Vol. 67, No. 2, pp. 249-283, April 1992.
−
m ’s = 0, assuming the disclosure requirements are ignored, [3] American Institute of Certified Public Accountants
and assuming the inter-relationships between the balance (AICPA), SAS 107: Audit Risk and Materiality in
sheet accounts and the classes of transactions are ignored, Conducting an Audit, American Institute of Certified
reduces to Srivastava and Shafer audit risk formula at the Public Accountants, New York, 2006.
balance sheet level. Because of the lack of space we do not [4] American Institute of Certified Public Accountants
derive the audit formulas for Levels 3, 4, and 5. However, (AICPA), SAS 106: Audit Evidence. Statement on
one can derive these formulas by using [7] and [8]. Auditing Standards, No. 106, American Institute of
Certified Public Accountants, New York, New, 2006.
Table 2 [5] Rajendra Srivastava, Theodore Mock, and Jerry
Belief masses propagated from higher level3 to lower levels Turner, Analytical Formulas for Risk Assessment for
+ a Class of Problems where Risk Depends on Three
mX4←all X5 = ∏ m+X5 , where X5 ε {CAi.j.k.1, CAi.j.k.2,CAi.j.k.3, Interrelated Variables. International Journal of
X5

CAi.j.k.4, CAi.j.k.5,CDCAi.j.k} and X4 ε {CAi.j.1, CAi.j.2, … Approximate Reasoning Vol. 45, pp. 123–151, 2007.
CAi.j.k}, k depends on the number classes of transactions [6] Rajendra Srivastava, The Belief-Function Approach
associated with each variable at Level 4. to Aggregating Audit Evidence, International
Journal of Intelligent Systems, Vol. 10, No. 3, pp.
+
m X3←all X4 = ∏m + = ∏[1− (1− m+X4← all X5 )(1− m+ ) / K
'
X4 X4 X4
]
329-356, March 1995.
X4 X4

where KX4 is defined in Equation (9) [7] Rajendra Srivastava, Alternative Form of Dempster’s
Rule for Binary Variables. International Journal of
+
m X2←all X3 = ∏m + = ∏[1− (1− m+X3← all X4 )(1− m+ ) / K
'
X3 X3 X3
] Intelligent Systems, Vol. 20, No. 8, pp. 789-797,
X3 X3 August 2005.
where KX3 is defined in Equation (13). [8] Rajendra Srivastava, Prakash Shenoy, and Glenn
X3 ε {Ai.1, Ai.2, Ai.3, Ai.4, DAi} Shafer, Propagating Beliefs in an 'AND' Tree,
+
m B←all X2 = ∏m + = ∏[1− (1− mX2+ ← all X3 )(1− m+ ) / K
'
X2 X2 X2
] International Journal of Intelligent Systems, Vol. 10,
pp. 647-664, 1995.
X2 X2
where KX2 is defined in Equation (17). [9] William F. Messier, Jr., Steven M. Glover, and
X2 ε {A1, A2, … An. DB}. Douglas F. Prawitt, Auditing and Assurance
Services, 7th Edition. McGraw-Hill Irwin, 2010.
V. CONCLUSION
In this paper we have derived a general audit risk formula
at the balance sheet level in terms of belief masses, i.e., m-
values, obtained from various audit evidence gathered by the
auditor. In our derivation, we have considered the possibility
of mixed items of evidence, that is both positive and negative
items of evidence. Also, we have integrated current AICPA’s
auditing standards [3] [4] which separates management
assertions into three levels: Balance Sheet Accounts; Classes
of Transactions; and Presentation and Disclosure. In addition,

5
Proceedings of the 2010 Conference on Belief Functions, Brest, France, April 1-2, 2010.

Level 1: Overall Level 2: Accounts in Level 3: Assertion Level Level 4: Class of Level 5: Assertion Level of
Balance Sheet Level Balance Sheet Level of Accounts in Balance Sheet Transactions Level Class of Transactions
Class of Transactions
Assertion Ai.1 of CAi.2.1 related to &
Account Ai & Assertion Ai.2 Assertion CAi.2.k.1 of Class
Account A1 on {cai.2.1, ~cai.2.1}
& {ai.1, ~ai.1} of Transactions CAi.2.k
Balance Sheet
{cai.2.k.1, ~cai.2.k.1}
{a1, ~a1}
Assertion Ai.2 of
………….. Account Ai & & Assertion CAi.2.k.2 of Class
of Transactions CAi.2.k
Account Ai on {ai.2, ~ai.2}
{cai.2.k.2, ~cai.2.k.2}
Balance Sheet &
{ai, ~ai} Assertion Ai.3 of Class of Transactions Assertion CAi.2.k.3 of Class
Account Ai & CAi.2.k related to & of Transactions CAi.2.k
Balance Sheet ………….. {ai.3, ~ai.3} Assertion Ai.2 {cai.2.k.3, ~cai.2.k.3}
& {cai.2.k, ~cai.2.k}
Fairly Presented (B)
{b, ~b} ………….. Assertion Ai.4 of
Assertion CAi.2.k.4 of Class
& of Transactions CAi.2.k
Account Ai
Assertion DAi.1 {cai.2.k.4, ~cai.2.k.4}
{ai.4, ~ai.4}
{dai.1, ~dai.1}
Account An on Presentation and Assertion CAi.2.k.5 of Class
Balance Sheet & Disclosure (DAi) & Assertion DAi.2 of Transactions CAi.2.k
related to Account Ai {dai.2, !dai.2} {cai.2.k.5, ~cai.2.k.5}
{an, ~an} {dai, ~dai}

Assertion DB.1 {db.1, ~db.1} Assertion DAi.3 Presentation and Disclosure

Presentation and {dai.3, ~dai3} related to the Class of
Disclosure (DB) at Transactions DCAi.2.k
& Assertion DB.2 {db.2, ~db.2}
the Balance Sheet {dcai.2.k, ~dcai.2.k}
Level {db, ~db}
Assertion DB.3 {db.3, ~db.3}

Figure 1: Evidential Diagram for the Audit Process. The circles with a symbol ‘&’ represents the AND relationship between the variable on the left with the variables on
its right. The various assertions are listed in Table 1.