In short
- Analytical procedures are required in risk assessment and in the final review; as substantive procedures they are optional but powerful when the relationship is predictable.
- A substantive analytical procedure needs a precise expectation, a documented threshold and an investigation of significant differences — not just a comparison of this year to last year.
- The reliability of the data and the precision of the expectation determine whether the procedure is sufficient evidence.
- In Cambodia, management accounts, VAT and Tax on Income returns, payroll and inventory records are useful, but they must be tested for completeness and accuracy.
- A plausible explanation from management is not evidence until it is corroborated and documented.
Three uses of analytical procedures
Analytical procedures are used at three stages of an audit. In risk assessment, they help identify unusual transactions, trends or ratios that may indicate a risk of material misstatement. In the substantive phase, they can be used instead of, or alongside, tests of detail when the auditor expects the relationship to be predictable and the procedure can be precise enough. In the final review, they help assess whether the financial statements are consistent with the auditor's understanding of the entity.
ISA 315 requires risk assessment procedures, which include analytical procedures. ISA 520 sets the requirements for substantive analytical procedures and for the overall review. The auditor should not treat the three uses as interchangeable: a high-level risk assessment ratio is not sufficient evidence for a substantive conclusion, and a final review is not a substitute for substantive testing.
The auditor may also use data analytics tools to analyse a full population, such as all journal entries or all sales invoices. The tool changes the volume of data, not the need for a documented expectation, a threshold and an investigation of significant differences.
Developing a precise expectation
A substantive analytical procedure starts with an expectation of the recorded amount. The expectation should be precise enough to detect a misstatement that could be material, individually or in aggregate. A simple comparison with the prior year is usually too imprecise for a significant account; a model based on volumes, prices, headcount, payroll rates, occupancy or production data is more persuasive.
The auditor can use trend analysis, ratio analysis, a reasonableness test or a regression model. The choice depends on the account, the data available and the risk. A payroll expense can be modelled from headcount and average salary; revenue can be modelled from volumes and prices; inventory can be modelled from purchases, cost of sales and closing quantities. The model should reflect the entity's business, not a generic formula.
The data used in the expectation must be reliable. The auditor considers whether the data is independent, complete, accurate and relevant, and may test the general ledger, the payroll system or the sales ledger to support it. Management accounts that have not been reconciled to the ledger are a weak basis for a precise expectation.
| Account | Expectation | Data to test |
|---|---|---|
| Revenue | Volume × price, by product or customer | Sales ledger, contracts, VAT returns |
| Payroll | Headcount × average salary + NSSF | Payroll register, contracts, Tax on Salary return |
| Cost of sales | Opening inventory + purchases − closing inventory | Inventory records, purchase invoices, production data |
| Rent | Square metres × rate × months | Lease agreements, rental tax return |
| Interest | Average debt × rate × months | Loan agreements, bank statements, effective interest schedule |
Thresholds, differences and investigation
The auditor sets a threshold for investigating differences between the expectation and the recorded amount. The threshold is not a fixed percentage; it reflects the materiality of the account, the risk of misstatement and the precision of the expectation. A difference above the threshold is investigated; a difference below it is accepted, with the reason documented.
The investigation starts with an inquiry of management, but an explanation is not evidence by itself. The auditor corroborates the explanation with other evidence, such as a contract, a subsequent transaction, a price list, a payroll register or a third-party confirmation. If the explanation is not corroborated, the auditor performs additional procedures, including tests of detail.
A significant difference that remains unexplained is a misstatement or a scope limitation. The auditor evaluates its effect on the financial statements and the audit opinion, and communicates it to those charged with governance where appropriate.
A plausible explanation is not evidence
Management can explain a margin drop as a one-off discount, a timing difference or a new customer. The auditor still needs the contract, the credit note, the subsequent sale or another corroborating source before accepting the explanation.
Cambodian data and practical examples
Cambodian companies often have monthly management accounts, VAT returns, Tax on Salary returns, withholding tax returns and annual Tax on Income returns. These can be used to build expectations, but they are prepared by the same finance team and may share the same error. The auditor should test the completeness and accuracy of the data, reconcile the management accounts to the ledger and the tax returns, and consider whether the tax data is on the same basis as the financial statements.
A common example is revenue: the VAT return turnover, the sales ledger and the bank receipts should tell a consistent story. A gap between the VAT return and the accounting revenue can indicate a timing difference, an unrecorded sale, a credit note or a classification error. Another example is payroll: the headcount, the payroll register, the Tax on Salary return and the NSSF declaration should reconcile.
Related-party transactions are a higher-risk area. The auditor can compare the related-party balances and the interest or rent rates with the market and the contracts, and investigate unusual movements. The analytical procedure is not a substitute for the ISA 550 procedures, but it can identify the relationship that needs them.
- Reconcile management accounts to the general ledger.
- Compare VAT turnover, accounting revenue and bank receipts.
- Reconcile payroll, Tax on Salary and NSSF data.
- Model inventory from purchases, cost of sales and closing quantities.
- Compare related-party rates and balances with contracts and market data.
Fraud red flags
Analytical procedures can highlight fraud risk. Unusual margin changes, revenue growth without cash, a growing receivable days ratio, round-number journals, a sudden change in the sales mix, or a cost line that does not move with volume are all signals. The auditor should consider these in the context of ISA 240 and the entity's controls.
A management override often appears as an unusual journal entry near the year-end, a manual reclassification, a reversal in the following period or a related-party transaction without a clear business purpose. The auditor can test the journal entries and use the analytical result to select the population, but the analytical procedure alone does not prove or disprove fraud.
Documentation
The audit file should document the expectation, the data and its source, the threshold, the comparison, the differences, the investigation and the conclusion. The documentation should be sufficient for an experienced auditor to understand why the procedure was performed, what it proved and what it did not.
ISA 230 applies to the documentation. The auditor should also document the reason for selecting the analytical procedure, the precision of the expectation and the reliance placed on it. A file that shows only a spreadsheet with a difference and a note 'explained by management' is not sufficient.
- 01Define the assertion and the risk the procedure addresses.
- 02Build the expectation from reliable data.
- 03Set the threshold and document the basis.
- 04Compare and identify the differences.
- 05Investigate and corroborate the explanations.
- 06Conclude whether the evidence is sufficient.
- 07Document the work and the conclusion.
Key facts
- Standard
- ISA 520
- Substantive analytical procedures and final review.
- Expectation
- Precise, data-based, documented
- Not just this year vs last year.
- Threshold
- Based on materiality and risk
- Not a fixed percentage.
- Cambodia
- Management accounts and tax returns
- Test the data before relying on it.
How this article is checked
Figures were verified against the sources listed below in November 2026 and are re-checked each January or when the rule changes. Written by Liv Bonito, Co-founder & Managing Director. General information, not advice on your situation: for that, talk to a partner.