Here’s an issue recently discussed by CPA Canada’s IFRS Discussion Group

• IFRS 15 Revenue from Contracts with Customers requires an entity to estimate variable consideration when determining the transaction price. The estimated variable consideration should only be included in the transaction price to the extent that it is highly probable that there will not be a subsequent significant reversal. Furthermore, at the end of each reporting period, the entity should update the estimated transaction price (including updating its assessment of whether an estimate of variable consideration is constrained) to represent faithfully the circumstances present at the end of the reporting period and the changes in circumstances during the reporting period.
• New information related to variable consideration may arise between the end of the reporting period and the date when the financial statements are authorized for issue. The Group is asked to consider whether this information is considered to provide evidence of conditions that existed at the end of the reporting period (i.e., an adjusting event). Paragraph 8 of IAS 10 Events after the Reporting Period requires the entity to recognize the amounts in its financial statements reflecting the adjusting events after the reporting period.

The group discussed the following fact pattern:

• Entity A enters into a contract with Customer Z on November 30, 2018, to deliver 100 widgets on December 31, 2018 (which is when the performance obligation is satisfied). The transaction price is $1,000 per widget and there are no other performance obligations in the contract. Entity A has a 30-day return policy. Payment is due on January 31, 2019.
• Entity A has a December year-end and financial statements will be authorized for issue March 31, 2019.
• Historical returns have been 3-5 percent of total widget sales for each respective month.
• In considering the requirements for constraining the estimate of variable consideration, based on historical returns and circumstances present at December 31, Entity A concluded that it is highly probable that a significant reversal in the revenue recognized for Customer Z will not occur if a transaction price of $95,000 is used, applying estimated returns of 5 percent.
• Customer Z returned two widgets (or 2 percent) on January 30, 2019.

The group discussed whether the subsequent return of two widgets is an adjusting event under IAS 10 – in other words, whether it provides evidence of conditions that existed at the end of the reporting period. Most group members thought it isn’t, and doesn’t: “These members considered the management’s process to estimate variable consideration to be rigorous. Absent any abnormal circumstances of the return that can raise questions about the accuracy of the management’s estimate, the year-end estimation does not need to be adjusted. One Group member commented the development in data analytics can improve forecast accuracy and reduce likelihood of differences between estimates and actual results occurring in the future.”

But on the other hand, even if that view can be justified, it might seem that to apply the adjustment based on the subsequent knowledge of returns could only serve to make the reported 2018 revenue more fairly presented. Some group members “focused on the nature of the intervening events that caused the difference between actual and estimated returns. If the intervening event does not relate to circumstances that existed at the reporting date, then the variable consideration should not be adjusted. Otherwise, the estimate may need to be adjusted.” But in the fact pattern provided, that’s not such an easy distinction to apply. If Customer Z expected to use 100 widgets but ended up making do with slightly fewer, is that (say) because of overly cautious planning before entering the contract, or superior execution subsequent to that, and either way, why should it matter to Entity A? Maybe a natural disaster in January 2019 forced the project to an early close, ensuring that the last two widgets would go unused, but explanations aren’t usually that dramatic. Should we assume that 3 to 5% of Entity A’s widgets are usually returned because it just isn’t that good at making widgets? If so, the fact that it did better in this case, and that the return was only 2%, perhaps does speak to circumstances at the reporting date (i.e. that the company was sending out a better quality of widgets than it usually does). And so on, and so on…

Of course, one can always pile more detail into such an example than it’s capable of bearing. But insofar as one can relate this one to the real world, it seems to me it will usually be possible to make a case for adjusting the revenue. And even more to the point, it will usually be quite difficult to argue that the 2018 statements would somehow be “better” if the revenue wasn’t adjusted. As I’ve written before in the context of first-time adoption, sometimes IFRS seems to elevate the avoidance of hindsight almost to a fetish…

The opinions expressed are solely those of the author