Estimates, Fair Values and Complex Audit Areas

Introduction and context

Auditing estimates is among the most judgment-intensive work performed by an external auditor. Unlike cash or a completed invoice, an estimate often concerns an uncertain future outcome. Management must select a method, identify assumptions, obtain data and determine a measurement date. The auditor must then assess whether the resulting amount is reasonable, fairly presented and supported by sufficient appropriate evidence.

This chapter addresses accounting estimates, impairment, provisions, fair value, expected credit losses (ECL), related parties, group audits and balances requiring specialists. The discussion is grounded primarily in IFRS concepts and International Standards on Auditing (ISAs), especially ISA 540 (Revised), ISA 550, ISA 600 (Revised), ISA 620, ISA 701, ISA 705 and ISA 706. Jurisdictional adoption and the applicable financial reporting framework must always be confirmed.

Learning objectives

By the end of this chapter, a student should be able to:

1. Explain estimation uncertainty, management judgment and the risk of management bias.
2. Design audit procedures for estimates, impairment, provisions, fair values and ECL.
3. Evaluate models, data, assumptions, sensitivity analysis and subsequent events.
4. Distinguish related-party identification from related-party disclosure.
5. Explain group-audit responsibilities and the appropriate use of specialists.
6. Formulate defensible conclusions and communicate significant risks and estimation uncertainty.

1. The nature of accounting estimates

An accounting estimate is a monetary amount subject to measurement uncertainty. Examples include useful lives, warranty provisions, loan-loss allowances, pension obligations, inventory net realisable value, asset impairment and the fair value of unquoted instruments.

Three dimensions are particularly important:

Estimation uncertainty is the susceptibility of an estimate and related disclosures to an inherent lack of precision. It increases where the forecast period is long, outcomes are volatile, data are scarce, models are complex or the estimate is highly sensitive to small changes in assumptions.

A useful conceptual representation is:

The expression is a framework rather than a universal accounting formula. f denotes the entity’s selected estimation process; the variables must be consistent with the applicable reporting standard.

Auditors should distinguish:

• Inherent risk: susceptibility to material misstatement before considering controls.
• Control risk: risk that the entity’s controls will not prevent or detect a misstatement.
• Detection risk: risk that audit procedures will fail to detect a material misstatement.
• Management bias: a lack of neutrality in selecting methods, assumptions or point estimates.

Bias is not established merely because an estimate later proves inaccurate. A reasonable estimate can differ from the eventual outcome. Bias is indicated by patterns such as consistently optimistic forecasts, selective use of evidence, unsupported changes in assumptions or estimates deliberately positioned near a covenant threshold.

2. The audit approach under uncertainty

ISA 540 (Revised) requires an auditor to understand the entity and its environment, including internal control relevant to estimates, and to assess risks at the assertion level. The auditor then designs procedures responsive to the assessed risks. Three broad response routes are available:

1. Obtain evidence from events occurring up to the auditor’s report date.
2. Test how management made the estimate, including methods, significant assumptions and data.
3. Develop an auditor’s point estimate or range and compare it with management’s estimate.

These routes may be combined. For a high-risk estimate, evidence should address not only arithmetic accuracy but also the appropriateness of the model and the reasonableness of assumptions.

A disciplined examination sequence

Understand the process. Identify who prepares, reviews and approves the estimate; determine the information system used; and identify relevant controls.

Assess complexity and subjectivity. A simple accrued utility expense differs from a Level 3 derivative valuation or a litigation provision based on uncertain legal outcomes.

Test data. Agree source data to contracts, ledgers, external confirmations or other reliable evidence. Test completeness, cut-off and whether data were used consistently.

Evaluate methods. Consider whether the method complies with the reporting framework and has been applied consistently. Investigate changes in methods or models.

Challenge assumptions. Compare assumptions with historical experience, budgets, industry evidence, market indicators and post-year-end outcomes. Management’s intent is relevant but is not, by itself, sufficient evidence.

Consider sensitivity. Determine whether reasonably possible changes could create a material misstatement or require prominent disclosure.

Look for contradictory evidence. Search for evidence that management may have omitted adverse information or used inconsistent assumptions across estimates.

Conclude and document. Record the evidence, contradictory matters, professional judgments, specialist involvement and the effect on the audit opinion.

3. Discounted cash flow and sensitivity analysis

A discounted cash-flow (DCF) model converts expected future cash flows into a present value. For a finite sequence of cash flows:

where PV is present value in currency units, CF_t is the cash flow at the end of period t, r is the periodic discount rate expressed as a decimal, and n is the number of periods. The formula assumes cash flows occur at period-end and that the rate matches the cash-flow frequency and risk basis.

Worked example. Suppose expected cash flows are CU100, CU110 and CU120 at the ends of years 1, 2 and 3. At a 12% annual discount rate:

• Year 1: CU100 ÷ 1.12 = CU89.29
• Year 2: CU110 ÷ 1.12² = CU87.69
• Year 3: CU120 ÷ 1.12³ = CU85.41

Therefore:

If the undiscounted total is CU330, discounting reduces it by approximately CU67.61. If a different comparison value is CU271.98, the reduction from that value to CU262.39 is approximately CU9.59. The comparison must not be confused with the total discount from undiscounted cash flows.

Audit procedures should test whether the cash flows are nominal or real, whether inflation is included consistently, whether the discount rate is pre-tax or post-tax as required, and whether terminal value assumptions are supportable. A perpetuity-growth terminal value may be represented as:

where TV_n is terminal value at period n, CF_n+1 is the next-period cash flow, r is the discount rate and g is the perpetual growth rate. This special formula requires r > g and a stable-growth assumption; it is not appropriate for every DCF.

Sensitivity analysis should vary one or more key assumptions within reasonably possible ranges. A sensitivity table is not a substitute for testing the underlying model. The auditor should identify combinations of changes that could result in impairment or a material disclosure about estimation uncertainty.

4. Impairment of non-financial assets

Under IAS 36, an asset or cash-generating unit (CGU) is impaired when its carrying amount exceeds its recoverable amount. Recoverable amount is the higher of fair value less costs of disposal and value in use:

FVLCD means fair value less costs of disposal; VIU means value in use. If carrying amount is CA, impairment loss arises when CA > Recoverable amount.

The auditor evaluates indicators such as declining sales, technological obsolescence, loss of customers, regulatory restrictions and increases in market rates. For value in use, important procedures include:

• reconciling approved budgets to the model;
• challenging forecast growth, margins and capital expenditure;
• comparing prior forecasts with actual results;
• testing allocation of corporate assets and goodwill to CGUs;
• assessing the forecast period and terminal assumptions;
• checking consistency between cash flows and discount rates; and
• performing independent sensitivity analysis.

Goodwill impairment is especially judgmental because goodwill does not generate independent cash flows. The auditor must assess whether the CGU is the lowest level at which goodwill is monitored and whether the allocation avoids masking an underperforming operation with stronger units.

5. Provisions and contingent liabilities

IAS 37 generally requires a provision when there is a present obligation from a past event, a probable outflow of resources and a reliable estimate. “Probable” and disclosure requirements must be interpreted under the applicable reporting framework.

The auditor should inspect contracts, board minutes, legal correspondence, claims registers and post-year-end payments. Legal letters are useful but not automatically conclusive. The auditor should evaluate the competence and objectivity of management’s legal advisers and consider whether management has omitted claims or selected an optimistic outcome.

For a population of individually similar obligations, an expected-value approach may be appropriate. For a single obligation, the most likely outcome may be relevant, but the auditor must consider other possible outcomes and the standard’s measurement requirements. Discounting is appropriate when the time value of money is material. The auditor should not treat a provision as a general reserve: it must relate to a qualifying obligation, not merely to management’s future plans.

6. Fair value measurements

IFRS 13 defines fair value as an exit price: the price that would be received to sell an asset or paid to transfer a liability in an orderly transaction between market participants at the measurement date.

The fair-value hierarchy prioritises inputs:

Fair value is market-participant based, not necessarily entity-specific. The auditor should assess the unit of account, principal or most advantageous market, highest and best use where relevant, valuation technique and disclosures.

Common procedures include reconciling quoted prices, testing observable yield curves, checking comparable transactions and re-performing valuation models. For Level 3 measurements, the auditor should test model logic, inspect calibration to transaction prices, challenge liquidity and credit adjustments, and compare assumptions with external evidence. A valuation can be mathematically correct but conceptually wrong if the wrong unit of account or market perspective is used.

7. Expected credit losses

IFRS 9 ECL is forward-looking and ordinarily reflects probability-weighted, discounted cash shortfalls. A general representation is:

where s indexes macroeconomic scenarios, w_s is the probability weight of scenario s with ∑w_s = 1, t indexes future time periods, DF_t is the discount factor for period t, and EL_s,t is the expected cash shortfall in scenario s and period t. The scenario weight is not the same as probability of default.

A simplified credit-risk decomposition may be written:

EAD is exposure at default, PD is probability of default, and LGD is loss given default. These quantities must be defined consistently for the same scenario and period. The expression is a simplified approximation; contractual cash flows, recoveries, discounting, prepayments and dependence among inputs may require a more sophisticated model.

Audit procedures include testing staging criteria, historical loss data, forward-looking macroeconomic forecasts, scenario weights, collateral values, cure assumptions, write-offs and model overlays. The auditor should investigate post-model adjustments, especially where they compensate for model limitations without transparent evidence.

8. Related parties and unusual transactions

IAS 24 requires disclosure of related-party relationships, transactions and outstanding balances in specified circumstances. The existence of a family relationship may cause an individual to be recorded in the entity’s related-party register, but the relationship alone does not automatically mean that a financial-statement disclosure is required. Disclosure depends on the applicable relationship category and whether relevant transactions, balances, commitments or other required information exist.

The auditor should obtain management’s related-party list, inspect minutes and declarations of interest, review shareholder and director records, search unusual ledger descriptions and examine transactions outside the normal course of business. Particular attention is warranted where terms are unusual, approval controls were bypassed, transactions lack a clear business purpose or balances are settled shortly after year-end.

The auditor must distinguish:

• identification of a related party;
• detection of a transaction or balance;
• evaluation of arm’s-length claims; and
• determination of required disclosure.

Management’s assertion that a transaction was at arm’s length does not remove the need for evidence. Related-party transactions may also create fraud risks under ISA 240 and may affect the auditor’s communication with those charged with governance.

9. Group audits

A group engagement partner remains responsible for the group audit opinion, even when component auditors perform substantial work. Under ISA 600 (Revised), planning should begin with a group-wide understanding of components, risks, information systems, consolidation processes and communication channels.

The group auditor should determine component materiality and performance materiality, identify significant components or significant risks, communicate instructions, evaluate component auditors’ competence and independence, and review their work sufficiently. Procedures should address intercompany balances, foreign exchange translation, acquisition accounting, goodwill, consolidation adjustments and the completeness of component reporting.

A component auditor’s work is not automatically sufficient merely because the auditor is reputable. The group team must assess whether the work responds to identified risks and whether additional procedures are needed. Restrictions on access to component information may create a scope limitation and potentially affect the opinion.

10. Specialists and complex balances

An auditor may use an auditor’s expert when expertise outside accounting or auditing is needed, for example in actuarial valuation, mineral reserves, derivatives, engineering estimates or legal claims. Management may also use a management’s expert.

The auditor remains responsible for the audit opinion. Before relying on expert work, the auditor should evaluate the expert’s competence, capabilities and objectivity; agree the scope and objectives; understand methods and assumptions; assess the relevance and reasonableness of findings; and resolve inconsistencies with other evidence.

A specialist does not transfer professional judgment to the expert. For example, an actuary may calculate a pension obligation, but the auditor must still evaluate whether the population, employee data, benefit terms and accounting framework were correctly applied. The auditor should also consider whether the expert’s work is sufficiently detailed for the identified assertion and materiality.

11. Integrated case: Northstar Manufacturing Ltd

This is a fictional illustrative case. Northstar reports under IFRS. Its year-end includes a goodwill impairment model, a warranty provision, an ECL allowance and transactions involving a director’s sibling.

Background and problem

Revenue has fallen in one division, but management’s five-year forecast assumes rapid recovery. The discount rate is 10%, while observable market rates and comparable risk suggest a higher rate. A warranty provision is based on three years of claims data, although a product redesign occurred during the current year. The ECL model applies one macroeconomic forecast but labels its probability as “default probability.” The director’s sibling is listed in the related-party register; the sibling’s company supplied ordinary materials on normal credit terms, but management proposes no related-party note.

Analysis and approach

The auditor first reconciles the impairment model to approved budgets and compares historical forecasts with actual outcomes. The auditor challenges the recovery assumption, tests the discount rate and prepares a sensitivity analysis. If the carrying amount exceeds recoverable amount under reasonable assumptions, an impairment adjustment is required.

For the warranty provision, the auditor separates pre-redesign and post-redesign populations, tests claims after year-end and considers whether the redesign changes the expected failure rate. Using obsolete historical data without adjustment could overstate or understate the provision.

For ECL, the auditor requires management to distinguish scenario weights from PD. For example, a 60% base scenario and a 40% adverse scenario are weights; within each scenario, the model may estimate a different PD_s,t. The auditor tests whether the weighted loss estimate is discounted and whether exposure, default probability and loss severity are consistent.

Regarding the sibling, inclusion in the register is prudent for identification and monitoring. The auditor then determines whether the relationship falls within the reporting framework’s related-party category and whether transactions, balances or commitments require disclosure. The family relationship alone is not treated as automatic proof that a note is required. The supplier transaction is inspected for authorization, terms, completeness and disclosure consequences.

Outcome and learning points

The auditor concludes that the original impairment estimate is materially optimistic, requires a revised warranty model and identifies an ECL model deficiency requiring correction or transparent disclosure. The related-party conclusion is based on the actual relationship category and transaction facts, not simply on register inclusion. Significant estimation uncertainty and the impairment matter are communicated to those charged with governance and considered for key audit matter reporting where applicable.

The case illustrates four principles: test the process, not just the final number; separate model components; investigate bias through patterns and contradictory evidence; and avoid confusing an internal control classification with a financial-statement disclosure conclusion.

12. Governance, technology and professional judgment

Audit committees should demand clear documentation of significant assumptions, model changes, sensitivity analysis and specialist independence. Governance is stronger when management establishes model-validation controls, independent review and escalation thresholds.

Technology can assist by analysing journal entries, comparing forecast-to-actual outcomes, identifying unusual related-party descriptions and testing large populations. Artificial intelligence may help identify anomalies, but it does not establish the reliability of source data, explain a causal relationship or replace professional skepticism. Auditors must assess data lineage, model drift, false positives, privacy and access controls. Automated output is audit evidence only after its reliability and relevance have been evaluated.

Common limitations include unavailable market data, correlated assumptions, model risk, hindsight bias in evaluating estimates and overreliance on management specialists. Best practice is to use triangulation: combine internal records, external evidence, subsequent events, independent recalculation and sensitivity analysis.

Chapter Summary

Accounting estimates require audit attention because their amounts depend on methods, data, assumptions and uncertain future events. The auditor evaluates inherent risk, estimation uncertainty, management bias and the effectiveness of relevant controls. Appropriate responses include testing subsequent outcomes, examining management’s estimation process and developing an independent point estimate or range.

DCF models require consistent cash-flow and discount-rate assumptions; the three-year example at 12% produces approximately CU262.39. Impairment testing compares carrying amount with the higher of fair value less costs of disposal and value in use. Provisions require a qualifying present obligation and a supportable measurement. Fair-value auditing focuses on the hierarchy of inputs, market-participant assumptions and model risk. ECL models must distinguish scenario weights from probability of default and must align exposure, default and loss assumptions.

Related-party auditing requires separate consideration of identification, transactions and disclosure. Group auditors retain responsibility for the group opinion, while specialists provide expertise without replacing auditor judgment. Strong audits triangulate evidence, challenge optimistic assumptions, document contradictory evidence and communicate significant estimation uncertainty to governance bodies.

Key Terms

Accounting estimate: A monetary amount subject to measurement uncertainty.

Estimation uncertainty: Susceptibility of an estimate to inherent lack of precision.

Management bias: Non-neutral selection or application of methods, data or assumptions.

Point estimate: Management’s single selected amount.

Auditor’s range: A range of amounts developed as an alternative benchmark.

Discounted cash flow: A method converting future cash flows into present value.

Discount rate: Rate used to reflect time value and relevant risk.

Terminal value: Value attributed to cash flows beyond the explicit forecast period.

Impairment: Excess of an asset’s carrying amount over its recoverable amount.

Cash-generating unit: Smallest identifiable group generating largely independent cash inflows.

Fair value: Market-participant exit price at the measurement date.

Level 3 input: Significant unobservable input in a fair-value measurement.

Provision: Liability of uncertain timing or amount meeting recognition conditions.

Expected credit loss: Probability-weighted, discounted estimate of cash shortfalls.

Probability of default: Likelihood of default for a defined exposure, scenario and period.

Related party: Person or entity meeting the applicable reporting framework’s relationship definition.

Component auditor: Auditor performing work on a component for a group engagement.

Auditor’s expert: Person or organization with expertise outside accounting or auditing used by the auditor.

Subsequent event: Event occurring between the reporting date and the auditor’s report date, or discovered thereafter under applicable standards.

Review Questions

Short Answer Questions

1. What is estimation uncertainty, and why does it increase audit risk?
2. Identify the three broad audit response routes for accounting estimates.
3. Distinguish fair-value Level 1, Level 2 and Level 3 inputs.
4. Why must scenario weights be distinguished from probability of default in an ECL model?
5. Why does a group auditor remain responsible for the group audit opinion?

Long Answer Questions

1. Explain how an auditor should audit a management DCF model, including data, assumptions, discount rate and sensitivity analysis.
2. Discuss the recognition and audit implications of provisions and contingent liabilities.
3. Explain the audit challenges associated with Level 3 fair-value measurements.
4. Describe the auditor’s responsibilities when using an auditor’s expert.
5. Explain how related-party identification, transaction testing and disclosure evaluation are connected but distinct tasks.

Analytical / Application Questions

1. An entity’s carrying amount is CU271.98 and its three-year discounted cash-flow value, using cash flows of CU100, CU110 and CU120 and a 12% rate, is calculated independently. Determine the correct present value and explain the audit implications of a material difference.
2. An ECL model uses a 70% base scenario and a 30% adverse scenario. Explain why these percentages cannot automatically be described as probabilities of default, and identify two procedures to test the model.
3. A director’s sibling owns a supplier that sold goods to the entity on ordinary terms. Design audit procedures to determine whether the relationship, transactions, balances and disclosures have been appropriately addressed.

Multiple Choice Questions with Answers

Question 1

Using year-end cash flows of CU100, CU110 and CU120 and a 12% annual discount rate, what is the approximate present value?

A. CU262.39

B. CU271.98

C. CU330.00

D. CU297.61

Correct Answer: A. CU262.39

Question 2

In an ECL model, what does a macroeconomic scenario weight represent?

A. The loss given default for a borrower

B. The probability assigned to an economic scenario

C. The exposure at default in a period

D. The discount factor applied to cash flows

Correct Answer: B. The probability assigned to an economic scenario

Question 3

Under IAS 36, recoverable amount is generally:

A. The lower of value in use and fair value

B. The carrying amount plus impairment loss

C. The higher of fair value less costs of disposal and value in use

D. Always the asset’s original cost

Correct Answer: C. The higher of fair value less costs of disposal and value in use

Question 4

Which fair-value category contains significant unobservable inputs?

A. Level 1

B. Level 2

C. Quoted market price

D. Level 3

Correct Answer: D. Level 3

Question 5

Which statement about a family relationship and related-party disclosure is most accurate?

A. Every family relationship automatically requires a note disclosure

B. Register inclusion and financial-statement disclosure are identical

C. Disclosure depends on the applicable relationship and relevant transactions, balances or commitments

D. Family relationships are never relevant to audit procedures

Correct Answer: C. Disclosure depends on the applicable relationship and relevant transactions, balances or commitments

Question 6

Which is an appropriate audit response to a highly uncertain accounting estimate?

A. Accept management’s number without testing because it is judgmental

B. Test the estimation process, assumptions and data, or develop an independent range

C. Replace the estimate with the prior year’s amount

D. Rely exclusively on management’s verbal representation

Correct Answer: B. Test the estimation process, assumptions and data, or develop an independent range

Question 7

What is the group engagement partner’s fundamental responsibility?

A. To transfer responsibility to every component auditor

B. To remain responsible for the group audit opinion

C. To audit only the parent company’s cash balance

D. To avoid reviewing component work

Correct Answer: B. To remain responsible for the group audit opinion

Question 8

Which condition is required for the perpetuity-growth terminal-value formula TV_n = CF_n+1 ÷ (r − g)?

A. The growth rate must exceed the discount rate

B. The discount rate must exceed the growth rate

C. Both rates must be zero

D. Cash flows must be paid at the beginning of each period

Correct Answer: B. The discount rate must exceed the growth rate

Question 9

What remains the auditor’s responsibility when an auditor’s expert is used?

A. Nothing, because the expert signs the valuation

B. Only the expert’s professional license

C. Evaluation of the expert’s competence, work and relevance to the audit conclusion

D. Guaranteeing that the estimate will equal the eventual outcome

Correct Answer: C. Evaluation of the expert’s competence, work and relevance to the audit conclusion

Question 10

A provision is generally recognized under IAS 37 when there is:

A. Only a possible future intention to spend money

B. A present obligation, probable outflow and reliable estimate

C. A budget approved by management without an obligation

D. Any reduction in expected profit

Correct Answer: B. A present obligation, probable outflow and reliable estimate