Directional testing - a methodology

This article seeks to:

• explain and illustrate 'methodology' in the context of auditing
• consider the implications of directional testing for the setting of materiality levels and sample selection, and
• illustrate relationships between methodologies (eg the 'cyclical approach').
  

Directional testing is an audit approach which, among other things, facilitates the design and implementation of audit procedures. This article is therefore relevant to all candidates sitting examination papers which call for audit procedures.

Many examination scripts fail to achieve a pass standard because candidates:

• appear to lack understanding of procedures (eg by regurgitating rote-learned tests or otherwise describing inappropriate audit work)
• are unable to apply general knowledge to specified transactions and balances
• make too few answer points (eg because they have run out of ideas)
• repeat answer points already made (eg because they have run out of ideas).
  

It is worth mentioning at this point that 'procedure' means 'a way in which a task is done'. Answer points of the form 'check', 'confirm', 'validate' etc do not amount to procedures, unless they go on to explain how (to check, confirm, validate, etc). Procedures can also mean 'a series of actions conducted in a certain order or manner' thus answer points should be structured to have a 'flow' and not be presented randomly.

An understanding of directional testing is therefore useful in that it:

• clarifies audit objectives
• directs audit work to risk areas
• links together the results of tests.

In addressing directional testing, this article seeks to:

• reinforce your knowledge of double-entry bookkeeping (which is assumed)
• explain why tests are conducted in a particular order and manner
• show, by illustration and example, how to describe procedures (and generate ideas for structured answer points)

Directional testing is an audit methodology which was developed in the 1980s to provide a framework for the conduct of the individual audit assignment and all audits. Although it is not widely used today (for reasons that will be mentioned later) it is of particular interest because it is:

• a good example of an audit methodology (ie the science of audit)
• conceptually straightforward (based on bookkeeping principles)
• still consistent with current 'best practice' as reflected in ISAs.
  

Current best practice requires, for example:

• planning to perform the audit in an effective manner - ISA 300
• an understanding of transactions - ISA 310
• consideration of materiality (eg in determining the extent of audit procedures) - ISA 320
• the assessment of inherent risk relating to financial statement assertions about transactions and balances - ISA 400.

 
Candidates should note that, in a recent study of audit methodologies in the UK, Canada and US, one of the reservations expressed about the emerging 'business risk' approach is that it is not well described by current auditing standards.

The concept
The first basic principle of bookkeeping is that every debit entry has a credit entry. If the trial balance (ie the list of balances extracted from the general ('nominal') ledger) balances, but a debit or a credit entry is misstated, it automatically follows that there must be a second misstatement to balance the first one. (Figure 1)

Figure 1
If trade receivables ('debtors') are overstated by $1,900 then either:

• another asset is understated by $1,900 (eg if cash received has not been recorded) or
• liabilities are overstated by $1,900 (eg if the bank is in overdraft and cash is unrecorded) or
• income (eg sales revenue) is overstated by $1,900 (eg due to incorrect 'cut-off' or invalid invoices being processed) or
• expenditure is understated by $1,900 (eg discounts given) or
• some combination, amounting to $1,900.
   

By testing debits directly for overstatement (O), the matching credits will be tested indirectly for overstatement. By testing credits directly for understatement (U), the matching debits will be tested indirectly for understatement. Thus, IF:

• all asset balances and expenses for the year (i.e. debits) are tested directly for overstatement and
• all credit balances and income for the year are tested directly for understatement

THEN, misstatement in the opposite directions will be tested indirectly.

Direct and indirect tests are also called primary and corollary tests, respectively. These are illustrated in Figure 2.

Figure 2

The primary tests interlock to give complete audit coverage.

It is important to appreciate at this stage that this article is about directional testing as a methodology - ie as a strategy applied to the conduct of the whole audit.

Activity 1 (solution at the end of the article)
If repairs and maintenance costs are overstated by $13,000, suggest possible misstatements in the double entry bookkeeping.

Primary tests
The concept of directional testing permits the primary testing of:

• debits for overstatement and credits for understatement or
• debits for understatement and credits for overstatement.

However, it is the former 'rule of thumb; that is used. There are several reasons for this.

It addresses some of the more common errors that arise in balance sheet items, eg:

• the understatement of liabilities (e.g. through omission due to oversight)
• the overvaluation of assets (eg through failure to recognise impairment or non-recoverability).

It helps to identify irregularities. In particular, a misappropriation (theft) will often result in an overstatement of an asset or expense. For example, the theft of a physical asset (eg cash or inventory) which has been recorded may be 'covered up' by writing it off (ie Cr asset and Dr expense).

It is more difficult for income to be overstated (by error or irregularity) and it will be detected, if material, indirectly. For example, if revenue is overstated by raising fictitious sales invoices a debit (eg cash or receivable) must be overstated (which will be tested directly).

A primary test for overstatement starts with the 'end result' ie the monetary amount recorded in the financial statements. The direction of testing is then 'backwards' to its 'source'. Such tests seek to confirm:

• the occurrence and measurement of recorded transactions; and
• the existence, valuation and rights to assets.

A primary test for understatement starts at the source (which may be non-monetary and not yet recorded in the books e.g. goods despatch notes) and traces transactions 'forwards' to the financial statements. These tests are aimed at ensuring the completeness (and measurement) of recorded transactions and balances.

A word of caution
Be warned that the term 'directional testing' is frequently misused simply because the majority of audit tests (both of controls and of a substantive nature) necessarily have 'direction' (as determined by the purpose of the test). Merely testing for completeness and existence of certain transactions and balances without consideration of other financial statement assertions (eg occurrence, measurement, valuation, rights/obligations) does not constitute 'directional testing' as an audit 'methodology' (or 'strategy').

A test for overstatement
Consider trade receivables (debtors). The audit objective is 'to ensure that trade accounts receivables are not overstated'. Amounts due from customers will be overstated if, for example:

• cash received has not been posted to a customer's account
• a sales invoice is over-priced or posted twice or invalidly raised
• a credit note due has not been raised
• insufficient allowance (provision) is made for non-recoverability.

Substantive procedures are therefore directed towards ensuring that such errors have not arisen (Figure 3). Thus a sample of customers, selected from the list of balances extracted from the receivables (debtors) ledger, are asked to confirm their balances (and all discrepancies are investigated). For non-replies, the make-up of the balance is agreed to supporting invoices, goods despatch notes and/or customer orders. Cash received after the year end is matched against amounts due at the end of the year to verify recoverability.

You should reflect at this point on the importance of the total of the list of balances extracted from the receivables ledger being agreed (by reconciliation if necessary) to the receivables ledger control account balance in the general (nominal) ledger.

Two tests for understatement
Consider sales revenue and the audit objective 'to ensure that revenue is not understated'. Revenue could be understated if for example:

• goods have been despatched but not invoiced
• receipts from cash sales have not been recorded
• sales invoices are underpriced or undercast or
• sales invoices raised have not been recorded, eg in the sales day book.

If starting at goods despatch as the source of a sale it should be ensured, through tests of controls, that goods cannot be despatched without a document (eg sales invoice or despatch note) being raised. This is to establish the completeness of the population from which a sample of documents can be selected to trace through the accounting system. It may be possible to start with a population of customers' orders as an alternative (see Figure 3).

Figure 3

Consider liabilities and the audit objective 'to ensure that trade payables (creditor) are not understated'. To test from source (ie that which gives rise to the liability) means starting with goods inwards. However, if this is not documented (eg on goods received notes), purchase invoices may provide the most complete population from which transactions can be tested. When a sample is selected from the 'other side' of the entry (eg in this case purchases are debits but test is understated of creditors) it is called the 'reciprocal population'.

Reflect at this point that trade payables are most likely to be materially understated in respect of the largest suppliers who will have been identified in the testing of purchases for overstatement.

Activity 2 (Solution at end of article)
Suggest audit tests for a factory payroll.

A note on inventory
Because the year-end inventory (stock) value appears in the balance sheet and the income statement (profit and loss account) it is always tested for both overstatement and understatement. Consider, for example, that in attending a physical count (stocktake) items are selected for counting from 'floor' (ie physical) to 'book' (ie recorded) amounts and vice versa.

Advanced considerations
The remainder of this article is provided solely to provoke thought in those  students whose imagination has been grabbed by directional testing and those who are not faint-hearted.

The role of materiality
Materiality defines the threshold of error at which the auditor might require the client to make an adjustment to the financial statements to avoid qualification. In the context of the financial statements as a whole, materiality is based on the auditor's assessment of the users' view on 'key components' (eg profit before tax, turnover, total assets, net current assets, etc).

Consider again trade receivables and sales revenue. Suppose $100,000 is regarded as material to revenue, but only $50,000 is considered to be material to debtors. If it is concluded that receivables are not materially overstated (by $50,000) then, as a corollary, sales are not materially overstated (by $50,000). However, in concluding that sales are not understated (by $100,000), it is also asserted that receivables are not understated, but this is by $100,000. So although revenue is not materially misstated, receivables could be materially understated (between $50,000 and $100,000).

There are various means of modifying directional testing to deal with this problem (eg by supplementing corollary tests). In the case of the above, additional tests on sales before the year end would provide further assurance that there was no material understatement of receivables at the year end.
Alternatively, when there are different materiality levels for key components, the level set for directional testing should be at the lowest of any one key component (sometimes called 'monetary precision of the audit'). However, this may result in over-auditing (eg testing larger sample sizes than is necessary) in certain areas.

Relationship with the 'cyclical approach'
Transactions may be categorised into 'cycles' which make up accounts in the balance sheet and related items in the income statement. For example:

• trade receivables, sales revenue and cash receipts make up the 'revenue cycle' and
• trade payables, purchases and cash payments make up the 'purchases cycle'.

The use of directional testing within each cycle (ie debits for overstatement and credits for understatement) has two significant benefits. First, it highlights the complementary nature of the primary and corollary tests. Second, it facilitates the use of a different monetary precision within each cycle and therefore recognises the level of risk attached to each audit area.

Testing the balance sheet in both directions
It may, in respect of certain audit clients, be appropriate to set two monetary precisions: one for the balance sheet and one for the income statement. To maintain the validity of the directional testing concept, the approach should be modified to testing either the balance sheet or the profit and loss account in both directions.

If the monetary precision of the balance sheet is less than that of the income statement, direct tests on income and expenses provide corollary tests on assets and liabilities which are not adequate to conclude on the misstatement of balance sheet items. (See previous argument above for sales and receivables.) However, by conducting direct tests on assets and liabilities in both directions, complete audit coverage can be achieved (Figure 4).

Figure 4

Testing liabilities for overstatement is straightforward because suppliers' accounts can be selected from a listing and vouched back to supporting invoices, goods receipts, etc. To test assets for understatement consideration must be given to how this could arise. For example, trade receivables will be understated if cash credited to an account has not been received, or if a credit note has been incorrectly raised. It will therefore be the credit entries in the asset accounts that are tested for their validity.

Testing the income statement in both directions
By the same argument, if monetary precision of the income statement is less than that of the balance sheet, direct tests on income and expenditure in both directions will achieve complete audit coverage.

To test income for overstatement requires that recorded sales are substantively tested for validity. To test an expense for understatement will involve identification of its source and verification of its completeness. For purchases this may entail tracing goods receipts through the accounting system. However, for many expenses such as rent, rates, depreciation and wages, completeness may be ascertained through analytical procedures such as 'proofs in total'.

Conclusion
In understanding where it [directional testing] is coming from you should now appreciate that, as a methodology, it provides complete audit coverage of the inter-related assets, liabilities, incomes and expenses which comprise financial statements.

Solution 1 - Repairs and maintenance costs

• An asset (eg plant and equipment) is understated (ie expense was capital, not revenue, in nature)
• Another expense (eg insurance) is understated (eg due to misposting)
• A liability is overstated (eg in respect of the future service element of a maintenance contract).
   

Solution 2 - Factory payroll