Audit sampling Q&A: applying ISA (UK) 500 Audit Evidence
17 Feb 2021
Michelle Roberts ACA and John Selwood ACA consider the most frequent questions received on sampling following recent changes to the Croner-i Navigate Audit tools for best practice improvements in this area
17 Feb 2021
ISA (UK) 500 Audit Evidence requires the auditor to design and perform procedures to obtain sufficient appropriate audit evidence to be able to draw reasonable conclusions on which to base his opinion. It provides guidance on various means for selecting items for testing, of which sampling is one.
What is sampling?
Audit sampling is defined in ISA (UK) 530 Audit Sampling as the application of audit procedures to less than 100% of items within a population of audit relevance such that all sampling units have a chance of selection in order to provide the auditor with a reasonable basis on which to draw conclusions about the entire population. However, this need not be an equal chance of selection.
This means sampling is not being used in procedures where individual items which have a particular significance are examined, eg, all items over £10,000, as here the auditor is testing part of the population in its entirety and cannot use the results to draw conclusions about the rest of the population. Similarly, procedures such as walk-through tests are not sampling.
What is sampling risk?
Where sampling is used, the auditor must accept a risk that the sample is not representative of the population from which it is drawn and that the wrong conclusion may be drawn from the test.
Sampling risk may result in:
• a population such as a file of reconciliations, which is acceptable, being rejected because the sample happens to pick a large proportion of the items in error. This normally results in additional, unnecessary audit work but should not affect the validity of the final audit conclusion; or
• a population, which contains a material error, being accepted as satisfactory because the sample happens not to select any of the items that contain errors. This risk is more serious because there is the possibility that an unqualified audit report is issued when the financial statements contain material misstatement (although the auditor should organise their other audit procedures so that it is, in fact, unlikely that such an error escapes detection).
When should sampling be used?
Consider the purpose of the audit procedure and the characteristics of the population from which the sample will be drawn. Consider what a deviation or misstatement may be. Remember, any deviations will be projected across the population, so will it make sense for this to happen?
Be aware of the appropriate population to test the specific audit objectives. This is more difficult when the audit objective is testing for understatement, such as the completeness of creditors, as the sample is chosen from a related population.
The relevant population for such a test of understatement might be suppliers’ statements, post-period payments or a list of all suppliers during the period, rather than unpaid invoices. In these cases, the auditor needs to take steps to ensure that the chosen population is complete, such as choosing suppliers’ statements from a list taken from the purchase ledger.
What determines sample size?
A population will have a risk of material misstatement as assessed by the auditor at the planning stage. The level of risk of material misstatement will have a direct bearing on the sample size used.
Where there is a high risk of material misstatement, the number of items tested from the population will be greater than in situations where the risk of material misstatement has been assessed as low. A higher sample size will result in lower sampling risk, increasing the auditor’s confidence levels that the population does not contain a material misstatement. Other factors will also affect sample sizes.
The value of the population being tested will relate specifically to the test being carried out. For example, when undertaking existence testing on fixed assets, the population would be the net book value of the assets. However, when undertaking tests on additions, the value of the population will be the value of the additions in the period.
Croner-i’s sampling method expects the auditor to identify high value and key items separately for testing. These items are considered sufficiently important to justify selecting all of them. High value items will be identified as transactions that are higher than performance materiality, ie, tolerable error.
Key items are other balances or transactions that are identified as significant in the particular circumstance of the test being carried out. The judgment of the auditor is required in assessing whether or not an item is a ‘key’ item for these purposes. No absolute definition of key items is possible. The residual population is the value of the population after deducting the value of the high value and key items.
Can sample sizes be reduced?
By dividing a population into discrete sub-populations, sample sizes may be reduced without increasing sampling risk. Populations could be subdivided by monetary value or risk profile, meaning greater audit effort can be directed to larger value or high risk items. However, the results of audit procedures performed on a stratified population can only be projected to items in the same stratum; other items will have to be tested separately.
In the Croner-i tools a smaller sampling factor could be used if additional work has been performed, for example if tests of details, and analytical procedures have been performed or tests of control performed by the auditor can use a smaller sampling factor. If both have been performed an even smaller factor can be used.
Performing test of controls can also have the benefit of reducing the assertion risk which would allow for a smaller sampling factor too. More guidance on test of controls can be found in Testing the operational effectiveness of controls.
Professional judgment is always needed to decide whether the sample size produced is appropriate. Accordingly, the suggested sample size may be adjusted, but the reasons for the adjustment must be given on the relevant audit working papers and must be justifiable.
The Financial Reporting Council (FRC) does not consider having a cap on sample sizes to be appropriate. In their Developments in Audit 2020 report they note:
‘Audit teams tend to default to limited sample sizes which can prevent an objective assessment of the actual test results. Limited sample sizes also prevent audit teams from exploring more effective ways of obtaining appropriate audit evidence, such as controls testing, or the use of data analytics.’
How do I pick the items?
Various means are available for selecting the chosen sample from a population. High value and key items will already have been identified. The sample selected from the residual population should be selected so as to cover fairly the whole of the population being tested.
This will involve the use of random, systematic or haphazard means of selection. The auditor should try to avoid the selection of a block of items as this is prone to bias and fails to consider the whole population adequately.
The errors identified will be projected across the population, therefore it is important to use a robust approach for selection to provide reliable data on which to evaluate errors.
The working papers should reflect where the sample was selected from and how the items were selected for testing. Sufficient information must be noted about the items sampled to allow them to be identified in the event that further investigation is required.
What happens if an item picked isn’t appropriate?
The auditor performs the planned procedures on each item in the sample. If any item selected is not appropriate, a replacement item should be chosen. However, the auditor should consider whether the unsatisfactory item indicates an error.
If procedures cannot be completed on a sample item, for example the supporting documentation has been lost, alternative audit procedures should be considered for that item. If no alternative procedures are possible, the item should be considered to be an error.
What happens if an error is found in a sample?
The investigation and evaluation of errors encountered during audit testing is a vital part of the audit. Errors should always be followed up. In no circumstances should they simply be ignored.
When an error occurs, there are two specific questions that need to be addressed, namely:
• Could other errors exist elsewhere within the population?
• Is it possible that the errors could be material to the accounts?
If the answer to both of these questions is yes, then additional audit work must be carried out. It is important to remember that the second question is asking whether it is possible, not whether it is likely.
In normal circumstances, therefore, some additional work must be carried out whenever an error is encountered. The work can be directed specifically to help the auditor answer both of those questions. Alternatively, it may be appropriate to extend the sample size in order to determine whether the level of error encountered is typical of the population as a whole.
How to conclude?
The auditor should assess the results of the sample tested to determine if the objectives of the test have been met.
To assess whether errors in the population might exceed the tolerable error, the projected population error plus anomalous errors (net of adjustments made by the entity) should be compared with the tolerable error, taking into account the results of other audit procedures relevant to that financial statement assertion.
This will allow a conclusion to be made on the test and may result in identifying the need for an adjustment to the financial statements.
Projection of errors is not appropriate for tests of controls. On discovering errors in a test of controls, the auditor should consider whether the error is indicative of a general weakness in the control system.
What happens if test objectives are not met?
If the auditor determines that test objectives have not been met, they could:
• request those charged with governance to investigate identified errors and the potential for further errors and to make any necessary adjustments; and/or
• modify the nature, timing and extent of further audit procedures. For example, in the case of tests of controls, the auditor might extend the sample size, test an alternative control or modify related substantive procedures; and/or
• consider the effect on the audit report.
About the authors
Where can I get further guidance?
Further guidance is available in the Guidance and Methodology area of Navigate Audit.
This article first appeared in Croner-i Audit & Accounting Update > 2021 > Issue 117 26 January 2021 > Analysis > Sampling Q&A
This article was correct at the date of publication. It is intended as an aid and cannot be expected to replace specific professional advice and judgment. No liability for errors or omissions will be accepted. It is the responsibility of those using the information to ensure it complies with the law at the time of use and that it is used in line with relevant rules and regulations governing the subject matter in question.
Except where otherwise indicated, all content is copyright of Croner-i Ltd.
© Croner-i Ltd, 2021
All rights reserved. No part of this publication may be reproduced without prior permission