The 5 mistakes I see most when marking SBR scripts — and what the marker actually wants

The gap between a 46 and a 52 in SBR is almost never knowledge. Most of the scripts I see in the high forties come from candidates who clearly know the standards. Some of them know the standards better than the people who passed. They just spent the exam proving that to me instead of using it.

That is the whole thing, really. The marking guide is a list of points, and most of those points describe a sentence that could only have been written by somebody who read this scenario. A sentence that would work in any answer to any question about IFRS 15 earns nothing, however accurate it is.

Five patterns, then, and what I am looking for when I hit them.

1. Writing out the standard instead of answering the question

The classic version: a requirement about a bundled equipment-and-servicing contract, and the first two thirds of the answer is the five-step model, in order, with headings. Step one, identify the contract. Step two, identify the performance obligations. By the time the candidate reaches the facts they have used eight minutes and picked up maybe one mark.

I do not need the model. I have the model in front of me. What I need is the line that says the servicing is distinct because the customer can benefit from the equipment on its own and servicing is available elsewhere, so the $600,000 is allocated on relative standalone selling prices and roughly $80,000 of it is deferred across two years.

A shape that works: one line of principle, then the application, then a conclusion that commits to something. If your principle line is longer than your application line, you have it the wrong way round.

2. Not touching the numbers

Discussion requirements are not arithmetic-free zones. The examiner puts figures in a scenario the way a novelist puts a gun on the mantelpiece.

I mark a lot of answers on contingent consideration that describe the treatment in perfect general terms and never say what happens to the statement of financial position. Meanwhile the scenario handed over the $5m maximum payout, the probability, the discount rate and the date. Somebody wanted that fair valued and unwound.

Nobody is expecting a full consolidation. A three-line working, or even "this increases goodwill by around $3.5m and creates a financial liability, with $0.3m of unwinding going to finance costs", is usually enough. There is often a mark sitting in exactly that sentence and nowhere else on the page.

3. Never engaging with what the directors actually did

Requirement: explain to the directors why their proposed treatment is not appropriate. Answer: a clean, correct explanation of the appropriate treatment, with no mention of what they proposed.

Those are two different answers, and the second one is missing the diagnosis. The directors said they would capitalise the costs. I want to know which criterion that fails, why, and what would change if they turned out to be right about technical feasibility.

Read the verb and read the audience. "Explain to a director with no accounting background" and "discuss the implications for the consolidated financial statements" want different registers and very different amounts of jargon. And a slightly petty note from this side of the desk: if the requirement says consolidated, answering at the subsidiary level costs marks you will never get back.

4. Treating ethics as something you bolt on at the end

"This breaches the fundamental principles of integrity and objectivity." I read that sentence about forty times a marking session, usually as the final line of an otherwise decent technical answer.

On its own it earns very little. The ethics marks sit in the join between the accounting and the motive. Why does management want those development costs capitalised this year? Because the covenant is tested on EBIT, or the bonus is on profit, or they are refinancing in March. Say that. Then say what it does to the users of the accounts. Then say what the accountant should do, specifically, because "act ethically" is not an action.

The best ethics answers I mark read like they were written by someone who has actually sat in a room with a finance director who wanted something.

5. Forgetting whose eyes you are looking through in the analysis question

This one costs the most marks and gets the least revision time.

The analysis question hands you a stakeholder, usually an investor or an analyst, and asks what the reporting means for them. Plenty of candidates treat it as a technical question in disguise and produce a competent essay on the accounting for the item, with the word "investor" sprinkled through it.

What I want is quality of earnings. Is this profit repeatable? A $12m bargain purchase gain is perfectly real income under IFRS 3 and tells you nothing about next year. Is the comparison fair, when one entity leases its fleet and the other owns it, and the finance cost sits below operating profit for one of them? Has the policy change made the three-year trend meaningless? What would you want disclosed that is not there?

Ratios help, but only when you say what the ratio means for the decision. "Gearing rises from 38% to 51%" is a fact. "Gearing rises from 38% to 51%, which on the face of it breaches the covenant in note 4, although that covenant is defined on the previous lease treatment" is analysis.

The professional marks

Two marks, twice, so four across the paper, and candidates give them away constantly. They go to the answer that looks like something a professional would actually send. Headings that match the issues. A memo that reads like a memo. A recommendation that recommends, rather than listing three options and stopping.

If you cannot find your own conclusion when you skim your answer back, neither can I.

The common thread

All five come down to the same habit. The candidate spends the exam demonstrating knowledge when the paper is asking them to use it. Read the requirement twice. Answer the part where the marks are, first. Put the scenario's own numbers and facts inside your sentences rather than around them.

That is most of the distance between a narrow fail and a comfortable pass, and almost none of it is extra studying.