7043 Accept-reject testing

Sep-2022

OAG Guidance

The objective of accept-reject testing, also referred to as attribute testing, is to gather sufficient evidence to either accept or reject a characteristic of interest. It does not involve the projection of a monetary misstatement in an account or population; therefore we only use accept-reject testing when we are interested in a particular attribute or characteristic and not a monetary balance.

• We accept that the test objective has been achieved when the test yields less than the number of exceptions we initially determined we can tolerate.

• We reject that the test objective has been achieved when the number of exceptions identified from the test exceeds the number of exceptions we initially determined we can tolerate. If the test is rejected the source of the exceptions needs to be identified, reported to the client and then additional work performed to achieve the audit objective.  In some cases this may only occur after the situation giving rise to the exceptions is corrected.

OAG Guidance

Generally, if we are testing a monetary value (either alone or in conjunction with other assertions) and not an attribute, audit sampling and not accept-reject testing is the appropriate testing method (presuming that sufficient evidence cannot be obtained from targeted testing). By focusing on the assertions we are addressing, we are able to determine whether the objective of the test is to obtain direct evidence related to monetary values or potential monetary misstatements. When we use tests of details to gather evidence regarding the assertions valuation and allocation (balances) and accuracy (transactions), we are generally attempting to obtain direct evidence related to monetary values and potential misstatements. However, it may be appropriate to test the accuracy of amounts that are not monetary balances, but are in fact used as attributes, by using accept-reject testing.

When the test is focused on other assertions, we typically obtain indirect evidence related to monetary values or potential misstatements. For example, in testing whether all shipments were recorded (an attribute), we know there is a potential misstatement associated with shipped items not recorded, but we are not always able to quantify that amount and estimate a monetary misstatement. In such cases, accept-reject testing can be applied. However, audit procedures are normally designed to test multiple relevant assertions, and if a monetary value is also being tested as part of the testing, sampling is likely the most efficient and effective testing method not accept-reject testing.

Audit sampling provides some important benefits over accept-reject testing. For example, audit sampling may be able to accommodate multiple errors so long as they do not project to an unacceptably high monetary level whereas multiple exceptions in accept-reject testing would lead to a rejection of the test.

Before performing accept-reject testing, determine it is the appropriate test in the circumstances. It is expected that controls testing, substantive analytical procedures and targeted testing are considered before using accept-reject testing.

The table below shows the assertions where non-statistical audit sampling and accept-reject testing may be appropriate depending on the circumstances including whether a monetary balance or attribute is being tested:

* When we use non-statistical sampling to test completeness, our plan for detecting such misstatements needs to involve selecting from a source that will provide appropriate evidence of whether there is understatement.

** It may be appropriate to test the accuracy of amounts that are not monetary balances, but are in fact used as attributes, by using accept-reject testing. For example, agreeing the price used to produce an invoice to the standing data file of approved prices. Also, note that we can use accept-reject to obtain evidence over accuracy of revenues in certain circumstances. Refer to guidance in OAG Audit 7011.1.

*** It is generally not common to use non-statistical sampling to test cut-off due to potential difficulties in projecting the identified misstatements.

Practice aids

The following flowcharts are recommended to assist in determining which of the testing methods—individually or in combination—to use:

• Testing strategies decision tree at OAG Audit 4024.
• Flowchart for selection of tests of details at OAG Audit 7041.

OAG Guidance

Examples of when accept-reject testing is appropriate (i.e., where the projection of a misstatement amount is not possible or appropriate) include the following:

• The mathematical accuracy of the total amount of an accounts receivable listing may be tested by totalling only certain pages or customer balances. If no misstatements are found, we may accept the test and conclude that the listing is accurately totalled. On the other hand, if a mathematical error is found, we ordinarily reject the accuracy of the listing, request the client to correct the reasons for the error, re-total the listing and re-perform the test.

• A population of bank reconciliations may be accepted or rejected on the basis of testing only some of the reconciliations if the processes and procedures applied to all reconciliations are the same (this typically applies to financial services clients).

• Testing the accuracy of census data (e.g., birth date, date of hire) submitted to an actuarial firm.

• Testing the proper aging of invoices and corroborating the entity’s policy to write off invoices over a certain age.

• Tie in of test counts for a physical inventory observation from tags to final summarisation (completeness and accuracy) and from the final summarisation back to the tags (existence or false inclusion and accuracy).

• Sales or accounts receivable cut-off testing for clients that do not have effective sequential controls over invoicing and shipping (i.e., targeted testing is not appropriate).

• Testing bank reconciliations by selecting reconciling items to test, where it is not practical to neither test all reconciling items nor target high value reconciling items. For example, selecting outstanding cheques on an accept-reject basis to see that cheques cleared subsequent to year end.

OAG Guidance

Examples of when accept-reject testing is not appropriate include:

• Testing the existence, valuation and accuracy assertions for accounts receivable, where a sample of items to be tested is drawn from the account balance at year-end or an interim date and the test involves either third-party confirmation or alternative procedures such as vouching recorded receivables to subsequent cash receipts or source documentation (i.e., purchase order or shipping documents).

• Testing of inventory pricing (when testing the final inventory summary compared to inventory count records, accept-reject testing is appropriate but when testing pricing, sampling is appropriate).

• Testing of fixed asset additions to supporting invoices.

• Testing for unrecorded liabilities when targeted testing by monetary amounts or risk (i.e., closer to year end) cannot provide sufficient evidence. If further evidence is required from tests of details, perform audit sampling of the remaining untested population (i.e., below the targeted test scope). Consideration will be given to testing several periods separately (e.g., every 15 days) to see that projected unrecorded liabilities are diminishing to an immaterial amount the further out from period end it gets.

• Testing repairs and maintenance expense.

• Recalculating depreciation expense by selecting individual items for recalculation. Note, in many cases, reasonableness testing through use of analytical procedures at a disaggregated level may provide sufficient evidence or reduce the level of evidence needed from testing individual items.

• Testing net realizable value (NRV) of inventory through reports of gross margin by product.

For each of the above examples use targeted testing or audit sampling.

Practice aid

The tests of details practice aid at OAG Audit 7041 provides further guidance on which type of test of details to use.

OAG Guidance

Accept-reject testing, which is a test of details, may sometimes appear to be similar to controls testing as in both cases are testing an attribute or characteristic. However, they have different purposes. The objective of controls testing is to evaluate whether a control operated effectively. The objective of tests of details is to detect material misstatements at the assertion level. Although errors are not projected the aim of an accept-reject test is still to detect whether the population is materially misstated. If more than the tolerable amount of exceptions is identified the test is rejected as not providing the desired evidence.

The required sample sizes to test controls are not necessarily similar to the required sample sizes used when performing accept-reject testing although the tests may appear similar. This is because they are two different types of testing (i.e., controls vs. a test of details), and we are not necessarily working to precisely the same confidence levels or exception rates.

If the attribute in which we are interested is a control, use the controls testing guidance in OAG Audit 6053.