The Population That Wasn't Complete
You are the audit senior on Vantage Supplies Ltd, a wholesaler with 12,000 sales invoices in the year and revenue of $22m; materiality is $110,000. You plan a test of details on receivables using monetary unit sampling from the sales ledger. The client extracts the ledger to a spreadsheet. While reconciling the extract to the trial balance you find the extract totals $21.4m, a difference of $600,000. The credit controller explains that $600,000 relates to invoices raised in the last week of March that were 'not yet posted to the ledger' because the month-end close was delayed. These invoices relate to goods despatched on 28–31 March. Separately, you notice the spreadsheet has no sequential invoice numbering and 14 duplicate invoice numbers appear.
1. Explain why the completeness of the population matters for MUS and determine the effect of the $600,000 on your sampling conclusion.
Read the worked answer
Why completeness matters: MUS (probability-proportional-to-size) selects individual monetary units. The sampling conclusion is only valid if the population is complete - every invoice that should be in the ledger is in the extract. If items are missing, they have zero chance of selection, so the sample cannot detect misstatement in them. Completeness of the population is a precondition, not an optional check (ISA 530).
Effect of the $600,000:
- The extract ($21.4m) does not agree to the trial balance ($22.0m). The $600,000 is 5.5x materiality.
- The omitted invoices relate to goods despatched before year end - under accruals/cut-off they belong in the current year. Their omission understates revenue and receivables.
- Because they are missing from the population, the MUS sample gives no assurance over them; the sample is drawn from an incomplete frame.
Required action:
- Obtain the missing invoices and add them to the population, or test them 100% as a separate population.
- Reconcile the extract to the trial balance and investigate all differences, not just this one.
- Re-perform the sample selection on the complete population if the omission is material to the sampling frame.
- Conclude on cut-off: the late posting suggests a cut-off control weakness; extend cut-off testing around 31 March.
2. The duplicate invoice numbers suggest a data integrity issue. What procedures address this, and how does it affect reliance on the client's spreadsheet?
Read the worked answer
Duplicate invoice numbers mean the spreadsheet may contain the same transaction twice (overstating the population) or that numbering is not controlled. Either way the data is not reliable as extracted.
Procedures:
- Run a duplicate test on invoice number and on invoice number + amount + date to distinguish true duplicates from reused numbers.
- Trace a sample of duplicates back to the sales ledger and to the underlying invoice to confirm whether they are one transaction or two.
- Reconcile the count of invoices (12,000 expected) to the extract; investigate differences.
- Obtain the system-generated report rather than a manually prepared spreadsheet where possible; if a spreadsheet must be used, obtain the IT general controls evidence over the report and test the extract's completeness and accuracy (ISA 500 - reliability of information produced by the entity).
- Check for gaps in the invoice sequence - a missing number may indicate a deleted or unrecorded invoice.
Effect on reliance: The client-prepared spreadsheet is not reliable without testing. The auditor must obtain evidence that the data is complete and accurate before using it for sampling. Document the reconciliation, the duplicate testing and the conclusion. If reliability cannot be established, select items by alternative means (e.g. direct from the ledger system, or a different population) rather than relying on the spreadsheet.