In researching an IFRS question recently, I stumbled across a whole new (for me) source of stealth guidance put out by the IASB's International Financial Reporting Interpretations Committee (IFRIC). I am referring to official descriptions of issues that were presented to IFRIC, but not taken onto their agenda. Notwithstanding that rejection strongly implies insignificance, the IASB has seen fit to compile IFRIC's stated reasons for rejection, going so far as to update references when changes are made to them. Even so, these compilations are all prefaced by a disclaimer that each snippet (by my count there is a about two hundred of them) is for "information purposes only" and does not "change existing IFRS requirements." Right.
What follows is the story of just one "information purposes only" comment that clearly was meant to be more than that — most likely a gift to a highly-valued stakeholders.
A Little Revenue Recognition Question*
During an IFRS course I was teaching to practicing accountants last month, one of the attendees asked me how discounts for early payment should be treated. His company, for example, offered terms of 2/10 net 30, and 99% of the customers took advantage of the discount. However, not all customers actually paid within the discount period; as is common in practice, a significant proportion of customers claiming the discount did so after the discount period lapsed.
If you have studied U.S. GAAP accounting at even a basic level, you would be aware of at least one method for treating early payment discounts, despite the fact that no specific provisions in GAAP dictate the accounting treatment for them. I couldn't even find an answer from a search of PwC's own GAAP accounting and reporting manual–normally a rich source of information that I use frequently. Thankfully though, practically every intermediate accounting textbook covers the basics of the 'generally accepted' alternatives treatments of invoice payment terms. Before we delve into IFRS for its requirements, kindly permit me to stipulate that there are two generally accepted approaches under U.S. GAAP:
- The 'gross method' calls for, assuming a $100 nominal price, recording $100 of revenue upon delivery of goods/services offset by accounts receivable. Before the financial statements are issued, however, two accounts receivable allowance accounts are established: one for future exercises of early payment options granted during the period (with an offset to a contra revenue); and as with all sales on account, another for uncollectible accounts (with an offset to bad debt expense).
- The 'net method' will record the initial sale at $98. If the company collects $100, the difference will be credited to financial revenue – usually interest income. The allowance for uncollectible accounts is handled in the same way as per the gross method.
Neither method is completely satisfactory for reasons that I won't go into here, but they are useful benchmarks against which to judge what I found from researching IFRS. As you will see, IFRS was crystal clear—until the IFRIC made one of its "information purposes only" comment about an issue they decided not to comment on. Go figure.
I initially focused my investigation of the IFRS treatment on the initial recognition aspect, because that was what my questioner was concerned with; also, the appropriate subsequent measurement would likely suggest itself from initial measurement. International Accounting Standard (IAS) 18, Revenue, provides as follows:
"The amount of revenue arising on a transaction is usually determined by agreement between the entity and the buyer or user of the asset. It is measured at the fair value of the consideration received or receivable taking into account the amount of any trade discounts and volume rebates allowed by the entity." [IAS 18.10, italics supplied]
Stripping out some of the qualifying language, the core of paragraph 10 is about as clear and principled as one could ask for: revenue is to be measured at the fair value of the receivable.
As to that qualifying language, I don't know for sure what "usually" means in this context, but since early payment invoice terms are so common, it should be safe to assume that the general rule enunciated in IAS 18.10 is applicable without alteration. The last phrase of IAS 18.10, "…taking into account the amount of any trade discounts and volume rebates …", also does not appear to pose a problem, although I have to admit that I don't have any firm grasp on its function. A fair value measurement should comprehend all potential economic consequences; why two specific factors are mentioned may be an artifact, which would have become outdated when fair value was formally defined by another accounting pronouncement.
IAS 18.10 is not fully consistent, however, with IAS 39, Financial Instruments: Recognition and Measurement, which provides guidance on accounting for financial instruments for initial recognition at fair value. However, I should note that IAS 39's application guidance does provide a materiality exception for measuring trade receivables at fair value:
"Short-term receivables and payables with no stated interest rate may be measured at the original invoice amount if the effect of discounting is immaterial." [IAS 39.AG79]
I'm going to proceed for now as if discounting would be material, and deal with undiscounted receivables separately.
Which brings us to subsequent accounting for the receivable. For that, we need to stay with IAS 39, because, trade receivables are within its scope. Specifically, IAS 39.46(a) provides that receivables are to be subsequently measured on the basis of amortized cost using the effective interest method.
A key difference between U.S. GAAP and IFRS relates to recognition of impairment. Under U.S. GAAP, the receivable may be reduced for an allowance for bad debts on Day One. However, this is not appropriate under IFRS because an allowance for bad debts should have already been taken into account when initially measuring the receivable at its fair value. Subsequent recognition of impairment losses may only be made if there is "…objective evidence of impairment as a result of one or more events that occurred after recognition of the [receivable]." (IAS 39.59)
So, the answer that I was going to send to my questioner was going to go something like this:
IFRS requires that the receivable be initially measured at fair value and subsequently measured at amortized cost until the invoice is paid or becomes past due. The receivable would be considered to be impaired if not paid within 30 days (or perhaps at some earlier date if an event occurs that provides objective evidence of impairment). The difference between total cash received from the customer and the initial valuation of the receivable should not be presented on the income statement in the same manner as revenues from the sale of goods.
However, for reasons of materiality, it is common practice to disregard the effect of discounting in the initial measurement of trade receivables. When that is the case, then the net method appears to be most consistent with the principles enunciated in IFRS. Accordingly, the liberal discount terms would indicate that the original amount of the receivable and the revenue to be recorded would be $98; any amount collected in excess would be financing revenue.
Finally, I must note that IFRS does not address the measurement of impairment if the receivable is measured at its undiscounted amount. Therefore, I believe it must be presumed that no impairment may be anticipated until a subsequent event provides objective evidence of impairment.
IFRIC the Fixer
Then, this little blurb from the July 2004 edition of IFRIC Update (and as I mentioned above, subsequently compiled with other similar snippets) was brought to my attention:
"Items not taken to the IFRS agenda…. Prompt settlement discounts … IFRIC members agreed that prompt settlement discounts should be estimated at the time of sale, and presented as a reduction of revenues. …"
In other words, IFRIC has 'agreed' (without putting the issue on their agenda – and in the absence of a quorum, I should add!) that IFRS calls for the gross method of accounting. Whatever their answer, it sure would be nice to know how they got there! Cynical me thinks that's precisely why this common question, for which IFRS is clearly not clear on, was not put on the agenda.
One of the criticisms that supporters of IFRS adoption in the U.S. make of U.S. GAAP is that the utterances and scribbling emanating from SEC staff members without due process become de facto rules. Notwithstanding disclaimers like the one that IFRIC employs for its snippets, these and similar sources are considered authoritative by issuers and derogatorily referred to by many as "speech GAAP."
But, a critical difference between SEC speech GAAP and IFRIC's version is that the SEC's interpretations almost invariably constrain the accounting possibilities; issuers don't like them. But, it seems pretty clear that issuers would appreciate those otherwise gratuitous comments from IFRIC, which make chicken salad out of even the most principled of international accounting standards.
Perhaps one can also criticize the various promulgators of U.S. GAAP for allowing diversity in practice to persist for such a basic and common question as early payment discounts. But let's suppose for argument's sake that the EITF formally considered putting the issue on its agenda, and then chose not to. In announcing their decision, would they proffer an opinion as to what the appropriate accounting treatment should be? That would be inconceivable. At the very most, IFRIC could publish a statement that they were not going to address an issue because, perhaps, they viewed it as already being settled – or that the existing standards are sufficiently clear. Anything beyond that is merely poorly-disguised "speech GAAP."
One of the many reasons why I am so opposed to IFRS adoption in the US is that I trust the IASB even less than the FASB to stick to its principles or to focus on the needs of investors. Chicken salad speech GAAP from IFRIC is just one more reason to feel that way.
*My analysis of this question was enhanced by contributions from two contributors to the AECM listserv: Pat Walters of Fordham University and Carla Carnaghan of the University of Lethbridge. An earlier version of this posting omitted reference to IAS 39.AG79; sincere thanks to S. Ramachandran for pointing this out to me.
Also, I have slightly altered the circumstances under which I received this question in order to maintain the confidentiality of a client relationship.