At the recent AICPA SEC and PCAOB Developments Conference, James Shnurr, SEC Chief Accountant, noted that preparers have lots of questions about accounting for transactions under the new standard and they need answers in order to implement the standard. If those answers require additional rule making this might cause a delay in the effective date of the standard. He also pointed out that comparability was a hallmark of the standard setting process. You can find more on his comments this Journal of Accountancy article.
In addition to the FASB/IASB Transition Resource Group which is looking at 28 issues and counting, the AICPA is trying to aid comparability through 16 industry focused task forces to assist in understanding the standard and provide examples. These task forces have been hard at work, but are just now coming to the point of sending information up to the AICPA Finance Reporting Executive Committee for review with publication of results not expected to happen any time soon. If comparability is a real key, I think it behooves the FASB to consider a delay to allow these groups to positively impact comparability within each industry as a result of their work.
The FASB is in the middle of outreach to understand where preparers are in implementing the standard and the issues they are encountering. They are considering a possible delay in the standard based on what they find during the research. The FASB has been clear that simply not getting around to the work is not a reason for a delay. Instead, they are looking for real substantive issues preparers are having with implementing the standard in time for a retrospective application. Either way, we should hear from the FASB in the first half of 2015 about a potential delay. Interestingly the IASB is saying they are not hearing any requests for a delay in the standard, but this is probably greatly influenced by the fact that under IFRS you only have to provide two years of income statements rather than three as required by the SEC. If US preparers only had to provide restated income statements for 2016 and 2017 and not 2015, I think you would be hearing a lot less about the need for a delay.
There is so much going on with the revenue standard I can’t possibly cover it all here. Bottom line is that the issuance of the final standard wasn’t the end, but really was only the end of the beginning of the process of changing the accounting for what is the most important number in the financial statements.
In this blog on the revenue recognition standard I want to cover how you determine the allocation of revenue across performance obligations. As I mentioned in previous blogs performance obligation is the new term for deliverable and it includes items that previously were not considered standalone deliverables. Once you determine your performance obligations and total compensation due from the customer (known as Total Transaction Price or TTP), you have to allocate that TTP over the performance obligations.
As a reminder, today we allocate the total compensation over the deliverables using a relative standalone selling prices for each deliverable. If you are getting paid $80 for two deliverables each with a standalone selling price of $50 then you allocate $40 of revenue to each deliverable ($50/($50+$50)*$80). The allocation process under the new standard is the same but the difference comes in how you determine standalone selling price (SSP). Today we have a strict hierarchy we must follow:
- If Vendor Specific Objective Evidence (VSOE) – or what you sell the product for on a standalone basis – is available we must use that to determine the standalone selling price
- If VSOE is not available then we look at Third Party Evidence (TPE) – or what someone else sells the product for on a standalone basis
- If VSOE or TPE does not exist then we use our best estimate of the selling price
The new revenue standard does not include the above strict hierarchy for determining standalone selling price. It does say the best evidence is an observable price when the entity sells the good or service separately, but then it goes on to add “in similar circumstance and to similar customers.” If I have been selling a product in California and launch a new operation in Texas I can probably reasonable argue that the sales price in California is not in a similar circumstance to a similar customer so I do not need to use it to determine the standalone selling price in Texas. The same might also be said for selling a single cellphone to a customer versus a multiple line plan because it can be argued a multi-line customer is not “similar” to a single-line customer.
The new standard goes on to say that if a standalone selling price is not directly observable you need to estimate the SSP by “considering:”
- Market conditions (I take this to mean TPE)
- Entity-specific factors (I don’t think this is VSOE, but your guess is as good as mine)
- Information about the customer
It goes on list ways to estimate a standalone selling price including bringing back the residual approach as a last resort which was eliminated from GAAP a few years back.
When a deliverable today equals a performance obligation tomorrow, I think many companies will initially default to the standalone selling price they already use today. As new products and services are developed and new performance obligations are added that don’t exist as deliverables today, I predict you will see a decline in the use of VSOE.
Another significant change in revenue recognition in the New Revenue Standard is the change in the way sales incentives are handled. Under current GAAP, unless you as the vendor are also a customer of your customer and paying for something in that capacity as a customer, any cash or a cash equivalent such as a bill credit or a gift card redeemable from you is considered a sales incentive that is recorded as contra-revenue. On the other hand if you give the customer an item you do not sell in the ordinary course of business such as bank giving away a free toaster for opening up an account, a wireless service provider giving away iTunes gift cards or a credit card company giving away loyalty points that can be redeemed for airline tickets or other merchandise, those items are still sales incentives, but are recorded as expense.
The new standard changes the accounting for sales incentives. Cash or cash equivalent sales incentives are still recorded as contra-revenue, and the cost of non-cash equivalent sales incentives is still expensed, but under the new standard the non-cash sales incentives are now considered a distinct performance obligation that must have a portion of the revenue from the contract – known as the Total Transaction Price (TTP) – allocated to them. The portion allocated to the non-cash sales incentive is recognized as revenue when the performance obligation is satisfied.
In the case of a service company giving away a different company’s gift card for signing a two year service agreement that means a portion of the revenue will be recognized as soon as that gift card is delivered which will generally mean earlier revenue recognition than under today’s rules. In the case of loyalty programs where points are earned and merchandise is usually delivered (long) after the transaction that resulted in the award of loyalty points, revenue will be recognized (significantly) later than under today’s rules.
No matter which way your company’s revenues are impacted, the one thing that is clear is that change is coming and it will be incumbent on preparers to be able to crisply outline how and why revenue is changing under the new standard to investors, internal management and really everyone in your company.
Probably the second most significant change under the new revenue recognition standard is the requirement to defer incremental costs to obtain a contract and costs (notice the word incremental is not listed here) to fulfill a contract. This change, like many of the most significant changes, impacts service and technology companies more than traditional manufacturers. Under current rules, companies have options when it comes to these costs and most companies chose to only defer such costs up to the amount of any upfront activation fees deferred under today’s revenue rules. Some companies make an accounting policy election to defer additional upfront costs over the life of the customer contract. Those companies will likely see less impact from the new requirement with changes mainly coming from aligning the costs being deferred to the new definitions.
Companies that defer upfront costs today, however, are in the minority. In fact, the irony of the Boards’ position on these costs is that over the last two decades many companies have moved from deferring all upfront costs to only deferring costs up to the amount of deferred revenue, and they received “preferability” opinions in the process of making that accounting policy change. I guess the fact that the accounting firms issued and the SEC accepted those opinions carried little weight with the Boards when it came to their effort to revise the matching of revenue and expense.
Companies are going to need to set up new processes and systems to accumulate and account for such costs. Today, many companies simply look at qualifying costs overall, stop counting when they get to the amount of deferred revenue, and simply book an entry deferring the same amount of cost as they have deferred for revenue. They also use the revenue amortization schedule to amortize the deferred costs. This process results in no need to look at all possible costs and allows the procurement and payable systems to book expenses straight to the general ledger without the need to additional sub ledgers or systems.
Under the new rules, that process won’t be so simple any more. Companies will need to evaluate costs to determine if they are incremental costs to obtain a contract (probably fairly obvious) or upfront costs, such as setting up a customer in your system or connecting them to your service (that include labor and other less obvious and easily isolated costs). Once they have determined the costs to be deferred, amortization schedules will need to be set up and the costs will need to be evaluated for impairment on an ongoing basis.
It is hard to disagree that the proposed process does a better job matching expense to the recognition of the related revenue. The problem is that the better job comes at the cost of a much more difficult accounting process that is also less conservative.
One of the most significant changes to revenue recognition on the new standard may be the elimination of the contingent, or cash, cap requirement. There are those that did not like the contingent cap requirements because they linked revenue recognition to cash payments and that did not seem to follow the spirit of accrual accounting. That view, however, seems based on believing the contingent cap requirement is in place due to a collectability issue with recognizing revenue before the cash is due from the customer. While that view may be applicable in some cases, the primary driver of the contingent cap rules is that you should not recognize revenue until delivery is complete. If the vendor is still required to provide a significant service in order to be entitled to a payment, you cannot recognize revenue under current GAAP.
In practical terms, if I need to provide wireless service in order to be entitled to a payment, I cannot and should not recognize any portion of that revenue even if today’s multi-element rules say a portion of that service charge should be allocated to a deliverable that is already completed (providing the subsidized handset), because if I don’t provide that second deliverable (the wireless service) I will never be entitled to the revenue. I acknowledge that the future service is almost always delivered and there is very little question about the ability to continue to deliver that service in the future. In fact, the business model of subsidizing wireless handsets is built upon delivering that service in the future, otherwise a business would go bankrupt very quickly selling smartphones to hundreds of dollars below cost.
The FASB has decided that if you have a contract, or even an implied contract, then the revenue from performance obligations (the new term for deliverables) is not contingent at all on the future performance because they start with the assumption that both parties will fully deliver on all elements of the contract (the vendor will provide all performance obligations and the customer will make all required payments). Given that is the way it works a significant majority of the time that assumption is not unreasonable even though it is less conservative than the previous assumption.
The result of this position is that revenue will get recorded earlier under the new standard compared to existing GAAP. It also brings into question other “conservative” elements in existing GAAP. For example, the general practice is that no matter how likely a gain contingency is under ASC 450, you simply do not record it until it is legally owed because to do so “might be to recognize revenue before its realization” (ASC 450-30-25-1). On the other hand, you recognize losses as soon as they are probable and estimable. The asymmetrical answer under ASC 450 would seem to be the opposite of the answer the FASB has come up with under the new revenue standard and I wonder if and when the Board will decide to address that difference.
The FASB and IASB issued their converged revenue recognition standard on Wednesday, May 28. The 700 page document can be found here (ASC 606, Other ASC Conforming Changes, Basis for Conclusions and Implementation Guidance). While the standard is not effective until 2017, there is no time to wait to begin your implementation work on this one. If you are at a public company planning a retrospective approach you have just 7 months until you need to start computing those “prior year” amounts for your 2017 financials. You have an extra year if you are a private company and only have to report one year of comparative income statement amounts, but if you enter into long-term contracts you may already be past the time when you need to start gathering information for changes to your revenue recognition related to those contracts.
The new revenue recognition standard may be the most significant change to accounting standards in the 40 year history of the FASB. The exact impact depends on the type of business on which you report. Software, telecom and technology in general are likely to see the most significant impacts, but every type of business is likely to see some level of changes which will require new processes, new controls and possibly new accounting systems. You can see more about the changes in a Journal of Accountancy Article here. If you want to hear more about the telecom industry impact you can listen to a podcast interview of me by JOA editor Ken Tysac here.
The most significant changes in the revenue recognition standard are:
- Elimination of the contingent (cash) cap on recognizing revenue
- Deferral of all costs to obtain and fulfill a contract – this is no longer a policy election (where upfront recognition was considered “better” accounting – how ironic) and the deferral can no longer limited to the amount of revenue that you deferred
- Sales incentives that used to be considered an expense are now considered a (separate) performance obligation and must have revenue allocated to that obligation.
- Change in the hierarchy of support for “stand-alone selling price” (SSP) and the return, in some cases, of the residual method for computing SSP
I will cover each of these changes in more detail in future blogs, but in summary this standard will likely require earlier recognition of revenue and later recognition of expenses than currently allowed in GAAP. Whether that is good or bad is up to you.
After a period of relative quiet, work on the revenue recognition and leasing convergence standards is heating up just like the temperatures in the Summer. The Boards’ staff recently released their proposal for re-deliberations of the revenue recognition standard based on feedback from comment letters and various roundtable and other outreach activities over the past several months. The staff laid out a timeline that will cover the remainder of 2012 as follows:
June – Identification of separate performance obligations
July – Satisfaction of performance obligations
September – Constraining revenue recognized, collectability & time value of money
October – Allocating the transaction price to separate performance obligations
November – Disclosures, transition and effective dates
December – Cost-benefit analysis
The Boards have previously stated a final standard will not be issued until early 2013, but they have yet to delay the earliest projected effective date of 1/1/2015. Their stated reason is that the effective date is dependent on how much change required as well as what the transition requirements are in the final standard. The implication is that if the standard requires retrospective application then they will likely delay the effective date, but if it does not, they may still go with the 1/1/2015 date.
The number of comment letters was down from the first exposure draft, but that seemed mainly due to the lack of letters from the construction industry this time around. The main concerns of the construction industry appear to have been addressed – or at least addressed to the point that they felt that a few letters from industry groups could better address what else was needed rather than an orchestrated writing campaign from a multitude of businesses. On the other hand, the financial services, software and telecommunications industry appear to have several continuing concerns with the exposure draft and let the Boards know through letters and multiple round table appearances.
On the Leasing front it looks like we will be waiting a while longer for the promised second exposure draft. An exposure draft is now expected to be released in the fourth quarter, but it increasingly appears it may be a Christmas present (or will it be coal in the stocking of financial statement preparers), rather than a Halloween trick or treat.
The biggest news on the leasing front is the Board decision to look at additional models for expense recognition. In solving the issue of leases not being on the balance sheet, the Boards essentially treated all leases as financing arrangements which resulted in higher expenses earlier in a lease rather than the straight line expensing as required by current accounting for operating leases which make up the vast majority of leasing activity. The Boards have heard loud and clear that while many constituents thought the leasing rules were broken because they resulted in no recognized liability on the balance sheet, those same constituents were relatively happy with the current rules impact on the income statement. The Boards are now looking at three possible models including the model from the original exposure draft. The model initially favored by the FASB would essentially keep the current operating/financing lease type split for the income statement, but require all leases to be put on the balance sheet. Effectively, an entry would need to be made each reporting period to record a liability and asset equal to the present value of the required cash flows under the lease, while leaving the current income statement treatment for operating leases intact.
It’s important to note that no decisions have been made by the Boards and they have currently only asked the staff to conduct further research on the additional proposed models and report back to the Boards. It will certainly be interesting to watch the developments over the coming months as leasing and revenue recognition once again take the center spotlight on the Boards agendas.