ACCA AAA September 2026: Why "It’s a Self-Review Threat" Scores Almost Nothing
Every AAA exam contains an ethics requirement. In September/December 2025 there were three – 8 marks in Q1, 10 in Q2, 7 in Q3. That is 25 marks you know are coming before you open the paper.
And the SD25 examiner's verdict was blunt: naming the threat is not the mark. The mark is in the implication. It is "not sufficient to only demonstrate knowledge of the threats" – they must be "appropriately applied and explained in the context of the scenario."
You are writing to the audit partner, not a marker
The briefing notes in AAA are addressed to the engagement partner. The examiner made the point twice: the partner "will not need to be educated on what a self-interest or self-review threat is, for example, but will want to know why this threat specifically arises in the circumstances presented." Explaining what a self-review threat is scores nothing. Every mark sits in the second half of the sentence.
The perspective error that cost Q1 marks
In Q1 (Mistral Co), the audit committee asked the firm to help prepare and provide assurance on a new sustainability report. Most candidates spotted a management-responsibility threat — but explained it from the wrong angle. It was not the provision of assurance that created it. It was the request for the firm to help determine and set the measures being reported on. Get the trigger wrong and you lose the mark even though you named the right threat.
Worse, many offered "use a separate team" as the safeguard. The scenario had already stated a separate team would be used. No credit. The same thing happened elsewhere: candidates burned time on partner rotation rules that the scenario said were already being applied correctly.
Do not invent threats the scenario does not support
Mistral's audit committee offered a fee increase on the basis that the work was favourable and in line with their expectations. That is a contingent fee — specifically restricted for audit clients under the IESBA Code. Yet a significant number of candidates wrote about fee dependency instead, which nothing in the scenario suggested.
Q2 (Kerr Co) was the same story in reverse. Candidates claimed a self-interest threat because the due diligence work earned a fee. The examiner's response: "The fact that the non-assurance engagement will earn Earps & Co a fee is not sufficient to suggest that there is a self-interest threat as, of course, income must be generated for all work performed." Meanwhile the familiarity threat and the risk of assuming management responsibility were overlooked completely. Same in Q3, where few spotted that a new finance director making a personal request of the audit partner is a familiarity risk.
Wrong answer vs right answer
Scores almost nothing: "There is a self-review threat because the firm will be providing assurance on the sustainability report and would therefore be reviewing its own work. A safeguard would be to use a separate team."
Scores the marks: "The audit committee has asked the firm to help determine which sustainability measures are reported. Selecting those measures is a management decision, so the firm would be assuming management responsibility — and would later be providing assurance over information it had itself chosen, making it unlikely to challenge it. The assurance engagement can be accepted, but only if the firm declines any role in setting the measures and the audit committee accepts responsibility for them."
Same threat. One clause of label, three sentences of consequence.
Three things to do before 7 September
1. Write every threat in three parts. The specific fact from the exhibit that triggers it, why it compromises objectivity here, and what the firm does about it. Name the threat in half a sentence and spend the rest on the "so what".
2. Delete any safeguard the scenario has already applied. Separate team, partner rotation, specialist transaction advisory department — if the exhibit says it is in place, it is not your mark. Scan your answer for it before you move on.
3. Stop speculating. No "possible fee dependency", no "the short deadline suggests fraud". The examining team confirmed candidates "will always be provided with sufficient detail to come up with relevant ethical threats rather than speculative ones." If it is not in the exhibit, it is not a threat.
The bottom line
AAA passed 39% in June 2026 — the lowest pass rate of any paper in the sitting. The examiner's closing comment was that candidates "continue to show a lack of what should be assumed knowledge for materiality, ethics and auditor reporting."
Ethics is the one requirement you can see coming. Turn up knowing the implications, not the labels.