What you'll learn

  • You'll understand the criteria in ISA 701.8-10 for determining which matters qualify as KAMs
  • You'll be able to structure a KAM description using the four required components from ISA 701.11-13
  • You'll see fully worked KAM examples across four industries: manufacturing, real estate, financial services, technology
  • You'll know what makes a KAM description pass inspection and what gets flagged as boilerplate

The first time you write a key audit matter paragraph, it reads like a procedures memo. Dense, inward-facing, full of audit jargon that means nothing to the investor reading the annual report. The second time is slightly better. By the fifth, you realise the challenge isn't knowing what you did. It's explaining why it mattered and what you found, in language someone outside the audit team can follow.

A key audit matter (KAM) under ISA 701 is a matter that, in the auditor's professional judgment, was of most significance in the audit of the current period financial statements, selected from matters communicated with those charged with governance (ISA 701.8). The auditor describes why the matter was a KAM, how the audit addressed it, and any key observations or references to related disclosures.

What makes something a KAM under ISA 701.8-10

ISA 701.8 defines the starting population: matters communicated with those charged with governance. You cannot select a KAM from a matter you never raised with the audit committee or supervisory board. The communication requirement under ISA 260 creates the boundary.

From that population, the auditor selects matters of most significance. ISA 701.9 provides three categories to consider. First, areas of higher assessed risk of material misstatement, or significant risks identified under ISA 315. Second, areas requiring significant auditor judgment, particularly estimates with high estimation uncertainty. Third, the effect on the audit of significant events or transactions during the period.

ISA 701.10 narrows the selection further. The auditor determines which of these matters were of most significance in the audit. "Most significance" is relative, not absolute. A matter does not need to be alarming to qualify. It needs to have consumed disproportionate audit attention, involved difficult judgment, or affected the audit approach in a material way.

ISA 701.A9-A16 expand on the concept. A16 notes that the auditor may determine that there are no KAMs to communicate (other than any material uncertainty related to going concern, which follows ISA 570 rather than ISA 701). This is rare for listed entities but not impossible for a stable business with straightforward accounting in a quiet year.

The practical test: if you were explaining this audit to the incoming engagement partner next year, which two or three matters would you spend the most time on? Those are your KAMs.

The four-part structure of a KAM description

ISA 701.11-13 require the auditor's report to include a KAM section with a defined structure. Each KAM must contain these components.

Why the matter is a KAM. ISA 701.13(a) requires a description of why the matter was considered to be one of most significance. Many firms default to boilerplate here ("due to the significance of the balance and the judgment involved"). A good description explains the specific audit risk. What made this estimate uncertain? What about this transaction created the judgment? What was the financial magnitude relative to the financial statements?

How the audit addressed the matter. ISA 701.13(b) requires a description of how the matter was addressed in the audit. This is the procedures section, but written for a reader who is not an auditor. Avoid listing every procedure performed. Instead, explain the audit approach and the key procedures that addressed the specific risk described above. ISA 701.A46 states the description may include aspects of the auditor's response or approach that were most relevant.

Key observations or reference to related disclosures. ISA 701.13(c)-(d) permit (and in some cases require) the auditor to include an observation about the outcome of procedures, or reference the entity's related disclosures. ISA 701.A49 notes that observations should be consistent with the opinion on the financial statements. Saying "we consider the provision to be at the optimistic end of the reasonable range" in a KAM while issuing an unmodified opinion creates obvious tension. But referencing Note 12 to the financial statements and stating that the auditor's procedures did not identify material misstatements in that area is both informative and defensible.

A subheading that identifies the matter. Each KAM gets a descriptive subheading (ISA 701.11). "Revenue recognition" is minimal. "Revenue recognition on long-term construction contracts" tells the reader which revenue stream and why.

Worked example: revenue recognition in manufacturing

Scenario: Visser Machinebouw B.V., a Dutch manufacturer of industrial packaging equipment, has €52 million revenue. Approximately €18 million (35%) comes from long-term contracts with performance obligations satisfied over time under IFRS 15. The remaining €34 million relates to standard product sales.

Subheading in the report: Revenue recognition on long-term construction contracts

Why it is a KAM: Revenue on long-term contracts (€18 million, 35% of total revenue) is recognised over time based on an input method (cost-to-complete). This requires management to estimate total expected costs for each contract, which involves judgment regarding procurement lead times, labour efficiency, and subcontractor pricing. As at year end, four contracts had completion rates below 40%, meaning more than 60% of total cost remained estimated. The risk is that revenue has been recognised in advance of actual performance because costs have been underestimated.

How the audit addressed the matter: The engagement team tested management's cost-to-complete estimates for all contracts with remaining performance obligations exceeding €1 million (covering 89% of long-term contract revenue). For each contract, the team compared estimated costs to the most recent purchase orders, subcontractor quotes, and project timelines. For the four early-stage contracts, the team performed a retrospective comparison of management's prior-year estimates to actual outcomes on completed contracts. The team also tested a sample of revenue cut-off entries around the reporting date against delivery confirmations, customer acceptance certificates, and progress reports.

Documentation note: Working paper references the specific contracts tested, the comparison of estimated to actual costs on prior projects, and the results of cut-off testing. Include the percentage completion and remaining estimated cost for each significant contract.

Key observation and disclosure reference: The auditor's procedures did not identify material misstatements in revenue recognised on long-term contracts. Management's disclosures on the revenue recognition policy and significant estimates are included in Notes 2 and 14 to the financial statements.

Worked example: investment property valuation in real estate

Scenario: Bergen Vastgoed N.V., a Dutch real estate investment company holding a portfolio of 22 commercial properties valued at €185 million. Properties are measured at fair value under IAS 40, with valuations performed by an external valuer.

Subheading in the report: Valuation of investment properties at fair value

Why it is a KAM: The investment property portfolio (€185 million) represents 91% of total assets. Fair value is determined using a discounted cash flow model with inputs including market rental rates, vacancy assumptions, discount rates, and terminal yields. Small changes in these assumptions produce material valuation differences. A 25-basis-point shift in the weighted average discount rate changes the portfolio valuation by approximately €7.8 million. Management engaged an external valuer, but the selection of inputs involves judgment by both the valuer and management.

How the audit addressed the matter: The engagement team evaluated the competence and objectivity of the external valuer (ISA 500.A34-A48, ISA 620.9). The team compared the discount rates and terminal yields used against independent market data from publicly available real estate benchmarks. For five properties where the valuer's assumptions deviated most from market data, the team developed an independent point estimate using comparable transaction data. The team inspected lease agreements for the eight properties with the highest vacancy rates to verify occupancy assumptions. The team also compared prior-year fair values to current-year values and investigated properties where the year-on-year movement exceeded 10%.

Documentation note: Record the valuer assessment, the independent benchmarking analysis with sources, the properties selected for independent estimation and the results, and the vacancy verification. Document the threshold for selecting properties for detailed testing.

Key observation and disclosure reference: The fair values determined by management, with the assistance of the external valuer, fell within the reasonable range determined by the engagement team. Refer to Note 8 in the financial statements, which discloses the valuation methodology, key assumptions, and sensitivity analysis.

Worked example: expected credit loss provision in financial services

Scenario: Rheinberg Kreditbank AG, a German regional bank with a gross loan portfolio of €420 million. The expected credit loss (ECL) provision under IFRS 9 is €14.8 million, of which €9.2 million relates to Stage 2 and Stage 3 exposures.

Subheading in the report: Expected credit loss provision on the loan portfolio

Why it is a KAM: The ECL provision (€14.8 million) requires management to apply a multi-stage model with significant judgment at every level. Stage allocation depends on the assessment of significant increase in credit risk (SICR), which uses both quantitative triggers (days past due, rating migration) and qualitative overlays (sector-specific deterioration, borrower-specific events). The provision calculation requires assumptions about probability of default (PD), loss given default (LGD), and exposure at default (EAD), each calibrated from historical data adjusted for forward-looking macroeconomic scenarios. Management applied three economic scenarios with probability weightings. The provision is sensitive to scenario selection. Using only the base case (without downside weighting) would reduce the provision by €2.4 million.

How the audit addressed the matter: The engagement team tested the SICR triggers by recalculating stage allocation for a sample of 85 exposures across all three stages and comparing the result to management's classification. The team engaged an auditor's expert to evaluate the PD and LGD models, including backtesting model outputs against actual default experience over the prior four years. The team independently assessed the reasonableness of the three macroeconomic scenarios and their probability weightings by reference to published forecasts from the Bundesbank and ECB. For Stage 3 exposures (individually assessed), the team reviewed the four largest exposures (representing €6.1 million of the €9.2 million Stage 2/3 provision) against current borrower financial information, collateral valuations, and workout plans.

Documentation note: Record the sample selection methodology for stage re-performance, the expert's report on model validation (with the engagement team's evaluation of the expert's work), the independent scenario comparison, and the detailed review of Stage 3 exposures. Document the sensitivity of the provision to alternative scenarios.

Key observation and disclosure reference: The auditor's procedures did not identify material misstatements in the ECL provision. Management's disclosures regarding the ECL methodology, key assumptions, economic scenarios, and staging analysis are set out in Notes 3 and 19. The engagement team notes that the provision falls within the range of reasonable outcomes but is positioned toward the lower end of that range. This observation is consistent with an unmodified opinion because the provision remains within the range of acceptable estimates under ISA 540.

Worked example: capitalised development costs in technology

Scenario: Lund Digital ApS, a Danish software company with €28 million revenue. The company capitalised €3.4 million of internal development costs during the year under IAS 38, bringing the total capitalised development asset to €11.2 million (net of amortisation).

Subheading in the report: Capitalisation and impairment of internally developed software

Why it is a KAM: Management capitalised €3.4 million of development costs during the year, increasing the net carrying amount to €11.2 million (19% of total assets). IAS 38.57 requires that development costs be capitalised only when six specified criteria are met, including the intention and ability to complete the asset, and the ability to measure expenditure reliably. The judgment centres on two points: whether the projects have passed the research phase (particularly for the two new product modules still in beta testing at year end, with combined capitalised costs of €1.1 million), and whether any indicators of impairment exist given that one legacy module's revenue declined 30% year on year.

How the audit addressed the matter: The engagement team tested each capitalised project against the six IAS 38.57 recognition criteria. For the two beta-stage modules, the team inspected project documentation (technical feasibility reports, product roadmaps, resource allocation plans) to evaluate whether the research phase had concluded and the development phase had begun. The team verified capitalised labour costs for all projects by tracing timesheets and payroll records to project codes. For the legacy module showing revenue decline, the team tested management's impairment assessment under IAS 36 by reviewing the cash-generating unit's forecast, the discount rate applied, and the reasonableness of revenue projections against historical trends and customer renewal data.

Documentation note: Record the assessment of each project against IAS 38.57 criteria, the evidence obtained for the two beta-stage modules, the timesheet verification, and the impairment assessment for the legacy module including the sensitivity of the recoverable amount to changes in revenue growth and discount rate assumptions.

Key observation and disclosure reference: The engagement team concluded that management's capitalisation of development costs met the criteria in IAS 38.57 and that no impairment was required for any capitalised project at year end. Refer to Note 7, which discloses the accounting policy for capitalised development costs, the carrying amounts by project category, amortisation rates, and the results of management's impairment review.

Getting the tone right: what inspectors look for

The FRC and AFM have both flagged KAM descriptions that read identically year on year with no reference to current-period events. A KAM section that could have been written for any company in the same industry is boilerplate, and inspectors treat it as such.

ISA 701.A29-A44 address the description of KAMs. A31 states the description should be entity-specific. A33 notes the auditor should avoid overly standardised language. A34 reminds auditors to avoid implying that the matter was not resolved to the auditor's satisfaction (unless the opinion is modified).

What does "entity-specific" look like in practice? Numbers. The KAM should reference the financial magnitude of the matter. It should reference the specific feature of this entity's accounting that created the judgment (not "management makes estimates" but "the cost-to-complete model for four early-stage contracts required estimates of procurement costs where quotes were not yet finalised"). It should describe what the engagement team actually did, not what any auditor would do on any engagement.

The gap between a good KAM and a bad one is specificity. A bad KAM says "we tested management's assumptions." A good KAM says "we compared the discount rates used against published market data and developed an independent point estimate for the five properties with the largest deviation."

One additional point: the KAM section should not duplicate the auditor's responsibilities section. ISA 701.A53 warns against this. The responsibilities section describes what auditors do generally. The KAM section describes what this audit team did on this engagement for this specific matter.

Practical checklist

  1. Before drafting any KAM, confirm you communicated the matter to those charged with governance during the engagement (ISA 701.8). If it wasn't communicated, it cannot be a KAM regardless of its significance.

  2. For each proposed KAM, document why it was of most significance by reference to ISA 701.9's categories: higher risk area, significant judgment area, or significant event/transaction. State which category applies and why.

  3. Write the "why it is a KAM" paragraph first. Include the financial magnitude (euros, percentage of a relevant financial statement line), the specific judgment or uncertainty involved, and the sensitivity of the outcome to key assumptions.

  4. Write the "how addressed" paragraph second. Describe the audit approach and key procedures, not every procedure. Reference any experts used. The reader should understand what you did without needing to read the working papers.

  5. Include a reference to the entity's related disclosure (note number in the financial statements) in every KAM. This connects the auditor's report to the entity's own reporting and gives the reader a path to more information (ISA 701.13(d)).

  6. Read the KAM description against last year's report. If nothing has changed year on year except the numbers, redraft the description to reflect current-period circumstances (ISA 701.A44).

Common mistakes

  • The AFM has found KAM descriptions that are identical across multiple clients in the same industry within a single firm's portfolio. ISA 701.A31 requires entity-specific language. Template KAMs copied across engagements without modification will be flagged.

  • Teams include too many KAMs, diluting the concept. ISA 701.A10 indicates that for most audits, the number of KAMs will be relatively small. Two to four KAMs is typical for a mid-market listed entity. Listing seven or eight suggests the auditor could not determine which matters were of most significance.

  • The "how addressed" paragraph omits the outcome or observation, leaving the reader uncertain whether the auditor found anything concerning. ISA 701.13(c) permits inclusion of key observations. Using this permission is good practice and reduces the ambiguity inspectors flag.

Get practical audit insights, weekly.

No exam theory. Just what makes audits run faster.

No spam — we're auditors, not marketers.