Revenue cutoff for a show that opens on the last day of the quarter

The show opens on 29 September and closes on 1 October. It bills 9.3 million. Somebody in group finance asks how much of it belongs in the third quarter, and the room discovers that nobody has ever written the answer down, because in every previous edition the dates happened to sit inside one month.

Getting the revenue cutoff for a show right matters more than the amount involved, because whatever you decide this year becomes the precedent and the comparative. A decision taken in a hurry on the Friday before close will be defended for the next five years.

The test that decides it

There is a sequence in the standard and it does not start where most people start.

IFRS 15 paragraph 32 requires an entity to determine at contract inception whether it satisfies the performance obligation over time, under paragraphs 35 to 37, or at a point in time, under paragraph 38. So the first question is about the nature of the obligation, and the calendar has nothing to do with it. You do not look at the dates and then pick a treatment.

Paragraph 35 gives three criteria for over time. The one that engages an exhibition is the first, which asks whether the customer simultaneously receives and consumes the benefits provided by the entity's performance as the entity performs. The FASB set the identical criterion in ASC 606-10-25-27 when it issued Accounting Standards Update 2014-09, so the analysis runs the same way under either framework.

If none of the three criteria is met, paragraph 38 puts the obligation at a point in time and sends you to the control requirements in paragraphs 31 to 34, plus a list of indicators of transfer of control that includes the entity having a present right to payment and the customer having accepted the asset.

For a three day exhibition, an honest reading can go either way, and that is worth admitting rather than pretending the standard settles it. An exhibitor standing on their stand on day two is consuming the benefit as it is delivered, which points at over time. Against that, what the exhibitor contracted for is a whole show, and a show that stopped after day one would count as a failed delivery rather than a partial one. Most organisers land on point in time at completion, which is where Emerald Holding's disclosure sits when it says revenue is recognised in the period the trade show occurs.

Working the straddle

Take the 9.3 million and price out the three answers.

Point in time at completion puts everything in the fourth quarter. Third quarter revenue from this show is zero, fourth quarter is 9.3 million.

Point in time at opening puts everything in the third quarter, 9.3 million against zero. This is a harder position to defend, because on the morning of day one the organiser still owes two more days of a three day contract.

Over time, split ratably by day, gives two of three days in the third quarter, which is 6.2 million, and one day in the fourth, which is 3.1 million.

At a 58 per cent contribution margin those choices move 5.394 million of contribution around as well, either wholly into one quarter or split 3.596 million and 1.798 million. On a division doing 40 million a quarter, that is a swing of between 8 and 13 per cent of quarterly revenue decided by an accounting policy question.

Why is a straight line split by day the wrong answer?

Because if you have concluded the obligation is satisfied over time, the standard asks for a measure of progress that depicts the transfer of the service, and days are a poor proxy for how an exhibition delivers.

Trade show value is front loaded. On a typical three day business to business exhibition, badge scans by day might run 43 per cent, 34 per cent and 23 per cent. Apply that to the 9.3 million and the third quarter takes 77 per cent, which is 7.161 million, and the fourth takes 23 per cent, which is 2.139 million. Compare with the ratable split, which gave 6.2 million and 3.1 million. The difference between the two methods is 961,000, and one of them is measured while the other is assumed.

The ratable split is the worst available option because it carries all the audit cost of an over time position and none of the accuracy. If you are going to split, split on something you can evidence. If you cannot evidence a measure of progress, that is itself an argument that the obligation is not satisfied over time, and you should be at a point in time.

My own view is that point in time at completion is the right answer for a conventional exhibition and that the effort should go into never being in this position, which the last section deals with.

What the auditor will ask for

Three things, and having them ready turns a two week argument into a twenty minute conversation.

A written policy that predates the situation. A memo dated after the quarter end explaining why the treatment that happens to flatter the quarter is correct will be read exactly as unfavourably as it deserves. The policy should record the paragraph 35 analysis for your standard show contract and the conclusion, and it should be signed off once and reused.

Evidence that the contract terms support it. What does the exhibitor contract actually promise, and what are the cancellation and refund rights during the run? A contract giving the exhibitor a refund right if the show is abandoned after day one is evidence about when control transfers, and it is in your own paperwork.

Consistency across the portfolio and across years. If the September show is treated at completion, the December show that straddles the year end has to be as well, including in the year when that costs you the growth number. The one thing certain to attract attention is a treatment that changes direction with the result.

What the cutoff does to the closing balance sheet

Whichever treatment you pick, the period end balance sheet has to agree with it, and this is where a clean policy still produces a messy close.

If the show completes in October and you recognise at completion, then at 30 September the entire 9.3 million sits as a contract liability, even though the doors were open for two days of the period you are closing. A reviewer looking at the balance sheet sees an event that has substantially happened and a liability that says nothing has. The note has to explain it, and the explanation is the policy.

The cost side needs the same discipline and usually gets less. Two days of hall hire, security, cleaning and audiovisual were consumed in September against revenue sitting in October. Those costs have to be deferred to match, which means someone has to identify them in the general service contractor's invoice, and general service contractor invoices do not arrive split by calendar day. In practice this becomes an accrual with a stated basis, and the basis should be the same one you rejected or accepted on the revenue side, applied consistently.

The number that catches people out is the pre-paid element. Cash collected months earlier is already in the bank and in the liability, so the cutoff decision changes no cash at all. It changes only where the revenue and the matching cost appear, which is worth saying out loud in the audit committee paper before somebody asks whether the quarter's cash was affected.

What to fix in the calendar instead

The cheapest solution to a cutoff problem is not having one.

Pull next year's calendar and flag every edition whose run touches a period end, using period end rather than month end, since a quarter or half boundary carries the reporting consequence. Then ask the show director what it would take to move the dates by three days. Often the answer is nothing at all, because the venue has the adjacent week and the pattern was set by habit.

Where a date genuinely cannot move, brief it in advance. Tell the audit committee in the quarter before, with the amount and the treatment, so the disclosure is a planned note rather than a surprise. The same list should reach whoever approves date changes, because a change of two days made for operational reasons can create a straddle that did not exist, and the person making that change usually has no idea. Handling the wider version of this across a portfolio is a reporting discipline of its own.

Where this stops

The analysis above assumes the show runs. A show that is cancelled or abandoned partway through is a different problem, involving refunds, contract modifications and cancellation insurance, and none of the cutoff logic survives contact with it.

There is also a limit on how much precision is worth buying. The difference between the ratable and scan weighted splits was 961,000 on 9.3 million, which is around 10 per cent of the show. If your materiality threshold at group level is well above that, the entire exercise is a control question rather than a numbers question, and the right amount of effort is a clear policy plus a consistent application. Spending three weeks building a scan weighted progress measure for an amount nobody will ever look at is a real cost with no reader.

The other honest limit is that the standard was not written with exhibitions in mind. The control indicators in paragraph 38 talk about legal title and physical possession of an asset, which map awkwardly onto a service delivered in a hall over three days. You are applying analogies, and reasonable preparers reach different conclusions. Write down which analogy you chose and why.

This week, take next year's calendar and mark every edition whose run touches a quarter or half boundary. For each one, write the expected revenue and the treatment you intend, in one line. If that list is empty you have no cutoff problem, and if it has three shows on it you have found next year's audit conversation eight months early, which is the whole point of doing it now. The recognition principle underneath the whole question is worth reading first, and the close mechanics that follow it sit with the month end close after an event in the same event finance set.