A Bundle of Concerns on New Revenue Rules

I recently wrote about the opportunities and challenges that the new Financial Accounting Standards Nos. 166 and 167 bring to U.S. financial reporting. My fear was that the new standards won’t help improve financial reporting not because of shortcomings in the standards themselves, but because experience suggests that implementation of the standards might not go the way we all would like.

Bear with me, folks, but I’m about to express very similar concerns about another significant change in U.S. Generally Accepted Accounting Principles.

Accounting Standards Update No. 2009-13 changes the guidance in Sub-topic 605-25 of the Accounting Standards Codification on revenue recognition for multiple-element arrangements. The changes will greatly reduce how often companies must defer revenue despite having delivered significant products or services to the customer. In that regard, they ought to lead to accounting that is more often consistent with the economics of the underlying transaction. But as with other standards that depend on good faith application of principles, adoption of the new guidance could go astray in many ways.

Today’s GAAP

Previous guidance on multiple-element revenue transactions was included in EITF Issue No. 00-21. The recognition principle underlying that guidance is that revenue should be recognized if the delivered items provide value to the customer and payment isn’t contingent upon future performance by the seller. The measurement principle is equally clear: The contract price should be allocated based on the relative fair values of the deliverables in the arrangement.

But when EITF 00-21 was written, trust in estimates of fair value was much lower than it is now, and there were concerns that revenue might be recognized too quickly without rules limiting the use of such estimates. Because of that, the consensus allows revenue to be recognized on delivered elements only when there is “objective and reliable evidence of the fair value of the undelivered item(s).” When there isn’t sufficient evidence of fair value of the undelivered items, you can’t recognize revenue. Therefore, a company that has delivered the lion’s share of the value in an arrangement is not able to recognize any revenue if the fair value of the undelivered items isn’t objectively determinable—even if the company has already been paid.

In large part, this exception to the underlying principles of EITF 00-21 is an abuse-prevention clause. Like many exceptions to principles in U.S. GAAP, it has proved troublesome over time. Although it accomplishes its goal of ensuring revenue is not recognized too quickly, it does so with a cost of sometimes recognizing revenue much too slowly.

What Is Changing?

The new guidance eliminates the requirement that the fair value of the undelivered items be objectively and reliably determinable. Instead, revenue is to be allocated (presuming other conditions are met) based on the best estimates of the selling prices of the individual items in an arrangement. There is no “reliability hurdle” on those best estimates. In other words, a lack of objective evidence about the selling price of undelivered items will no longer prevent you from recognizing revenue on the delivered ones.

So the principles are the same, but the new guidance eliminates a significant exception to those principles. This should lead to improved reporting, but I remain concerned.

What Could Go Wrong?

For starters, with the reliability hurdle on those estimates gone, the thinking might be that auditors and regulators will have to accept anything that management represents is its best estimate. Some might view this as an opportunity to manage earnings, by asserting estimated selling prices that they know aren’t realistic.

Some auditors or regulators may continue to push more objective evidence of estimated selling prices, even when the company believes that no amount of additional work can actually produce a better estimate.

This is exactly the kind of behavior that triggered a rewriting of the standards on software revenue recognition in 1997. Prior to that, the rules required companies to defer recognition of revenue if substantive performance obligations remained. Too often, companies asserted that remaining performance obligations were non-substantive even when they were, so the rewrite eliminated reliance on a subjective determination of whether an obligation was substantive.

This kind of behavior, of course, should not be tolerated by auditors, regulators, and users of financial statements. Those who interpret the new guidance as providing them an opportunity to manage earnings shouldn’t be involved with financial reporting. The possibility for this sort of deceitful accounting is what most concerns analysts and other financial statement users; if they believe such shenanigans will happen under this new guidance, any perceived improvements to reporting will be for naught.

Even without a sinister motive, preparers still might look at the new guidance as a way to reduce their workload by allowing them to estimate selling prices without much—if any—analysis. They wouldn’t be intentionally misleading; they’d just estimate based on their knowledge and understanding, without considering whether available evidence is consistent with that estimate. This thought process is also flawed.

The EITF consensus requires the company to consider whether “vendor-specific objective evidence” or “third-party evidence” of selling price exists, and to consider all available information and evidence when determining its best estimate. So while there is no such thing as an estimate that isn’t good enough to allow recognition of revenue for delivered items, the accounting must actually be based on the “best” estimate. Simply saying that a figure is the best estimate, without considering the evidence, won’t be sufficient to satisfy auditors, regulators, or users.

But there is a risk on the other end of the scale that could prove more difficult to avoid. Accountants, auditors, and regulators have been conditioned by the recent past to expect a high level of objectivity in estimates used to allocate revenue. But you can’t achieve that level of objectivity in all situations, and the new guidance requires allocation of revenue regardless. That means that revenue will now be recognized in reliance on information that we’ve been conditioned to think is “not good enough.”

I fear that some auditors or regulators may continue to push for more objective evidence of estimated selling prices, even when the company believes that no amount of additional work can actually produce a better estimate. Alternatively, auditors and regulators might become comfortable with analyses done in one particular way and reject different analyses, even when the differences make sense given the facts and circumstances.

My skepticism here arises in part based on experiences of the last decade. I don’t believe that EITF 08-1 should have been necessary, as I believe that the “objective and reliable” threshold was set in practice higher than it should have been. The accounting community made up and accepted bright-line rules for how close together the prices need to be in separate sales of an item for those sales to constitute objective and reliable evidence of selling prices. The resulting frustration with these “unwritten rules” played a part in the EITF’s decision to rethink the guidance.

The possibility for differing opinions between company and auditor also arises because the EITF rejected use of the “residual method” in Issue 08-1. In the past, the residual method meant that companies really only needed to be concerned with determining the value of the undelivered items. The new guidance requires that an estimate of the selling prices of all items in the arrangement be generated. That means companies will need to estimate selling prices of more items than in the past. These companies will likely see this adding no value, but auditors and regulators will (appropriately) point out that the guidance has changed.

Principles-Based Application

As it is with Statements Nos. 166 and 167, the key to a smooth implementation of the new revenue recognition guidance will be principles-based application by all involved. The EITF should be commended for writing guidance that is very much principles-based, which is something that the U.S. financial reporting community has often asked for. But principle-based standards only produce good financial reporting if they are applied and enforced in a way that recognizes the principles for what they are. If too many see the new guidance as ripe for abuse, we will get more abuse-prevention standards. And if the exercise of judgment necessary to apply the new guidance correctly is curtailed, we won’t get much improvement at all.

EITF 08-01 is neither a free pass to avoid work nor a standard that lends itself to a checklist implementation. The goal here is to allocate revenue in a way that reflects the value delivered to the customer. If everybody keeps that goal in mind, we can generate better financial reporting than we had before. If not, we will be writing new guidance on multiple-element revenue arrangements for years to come.