The IASB and FASB (the “Boards”) issued their joint exposure draft “Revenue from contracts with customers” in June 2010. The comment period for the exposure draft (“ED”) ran through to October 2010 and the ED attracted nearly 1000 comment letters. Reactions ranged from enthusiastic support through to scathing criticism. Along with construction, technology and pharmaceuticals, respondents from the telecom industry were among the most concerned about the impact of the ED should it be implemented as a standard in its current form.
The most significant reservation, echoed in nearly all of the comment letters from telecom operators, concerns the proposed treatment for recognition of revenue for the device provided to the customer in order for the customer to make use of the operator’s service. The most obvious example is a mobile handset. However, dependent on the nature of the arrangement between the customer and the service provider, the principle may apply equally to devices such as a tablet PC and a cable or satellite receiver. In fact, similar issues apply to any situation where the service provider is involved in the provision of the device necessary to use the service and the service itself, and the two are linked by a contractual relationship with the customer. Examples from other industry sectors include copiers, printers and medical imaging devices where the customer buys the device and a service or the media required for the device to function from the same seller.
A number of the provisions of the proposed standard remain under debate, with the Boards undertaking various outreach activities to connect with users and preparers as they reflect on the feedback received to date.
The comments in this paper are based on the ED and subsequent statements from the Boards through to the end of March 2011. The Boards will consider aspects of the ED that are addressed in this paper at meetings in April and May 2011. It is possible that there will be substantive changes to the proposed standard as a result of the redeliberation process – please refer to our website www.tiag.pwc.com
for news on the latest developments.
So what, exactly, is the problem? The following illustration is a simplified example from the mobile phone industry.
Example 1—mobile handset sold at a discount
Blue, a hypothetical typical mobile operator in Europe, sells handsets and airtime to its customers.
Customers who sign-up for a 12 month contract pay €100 for the Azure handset and €20 per month for 1000 minutes of calls.
Blue offers the same calling plan tariff (1000 minutes for €20) to prepay customers. Blue does not sell the Azure handset without a contract, but it is available from independent retailers for €250.
Blue’s current accounting treatment under IFRS is to recognise the sale of the handset at €100 on inception of the contract and the service charge on a straight line basis over the year, at €20 per month. This treatment is commonly adopted by other operators reporting similar transactions under IFRS and USGAAP and is often referred to as the “residual” method (for IFRS reporters) or the “cash cap” or “contingent revenue cap” method (for USGAAP reporters). In essence, it involves applying the implied discount of €150 received by the customer on the whole of the arrangement (being the difference between the market price and selling price of the handset) solely to revenue recorded for the handset sale.
The revenue recognition profile over the year thus looks like this:
The ED proposes that separable deliverables under a contract with a customer should be valued at their “stand alone selling price” and any discount on the entire arrangement applied pro rata to each element.
Blue’s future accounting under the proposed standard will reflect the total discount received by the customer as a deduction from both the handset and service elements, rather than the handset alone, with the allocation based on relative stand alone selling prices. The total revenue to be recorded remains equal to the cash to be received from the customer, but the pattern of recognition changes significantly to this:
Why operators object
There are a number of reasons why operators object to the treatment mandated by the ED. They fall broadly into two categories—technical and practical.
(i) An operator perspective – the technical objections
Operators expressed numerous reservations on accounting technical grounds when the proposals were at the Discussion Paper stage, and these were generally repeated in responses to the ED. The principal objection is that recognition of an amount that exceeds the up-front cash receipt is inappropriate because the operator is not entitled to that amount of consideration unless and until it delivers an ongoing service. In example 1, the operator charges €100 for the handset. It then charges €20 per month for delivering the promised service to the customer. The ED’s proposed treatment takes an element of the amounts receivable in future and allocates it to the delivered element – the handset. Operators argue they have no right to this additional element of consideration until they have fulfilled their contractual obligation to provide monthly service.
An alternative argument, which results in the same conclusion, is based around the degree to which the fair values of deliverables under the contract can be measured reliably. The mobile industry in many countries is highly competitive and features varied promotional offers and other incentives to customers, which may change very frequently. In these circumstances operators may contend that it is challenging to determine an estimate of “stand alone selling price” for a handset. If the tariff for ongoing service is at “market” rate, then one can measure reliably the total consideration for the arrangement and for the service element. The argument then follows that it is appropriate to determine the consideration relating to the handset (which, remember, cannot be reliably measured in its own right) as the difference between the two. The practical impact is that the whole of the discount on the contract is allocated against the delivered item – being the handset.
It is not the purpose of this paper to examine in detail the merits of the various arguments against the proposals in the ED. It is, however, worth noting that many operators currently reporting under IFRS and/ or USGAAP have applied the treatment described above, considering that it is both a fair reflection of the commercial substance of the arrangements and complies with existing GAAPs. It is also worth noting that there is little indication that the investor community is unhappy with the current accounting for revenues in the industry.
(ii) An operator perspective – the practical objections
These fall into two main categories – complexity and the separation of internal and external KPIs.
– Example 1 is very simple. In reality, the customer is likely to have a variety of tariffs to choose from and a series of choices to make about how many minutes and messages and how much data usage to include in the monthly service package. He or she may choose to take further optional services such as an international call package, music or game downloads and choose from contract terms of 12, 18 or 24 months. The tariffs and options selected may modify the up-front amount charged for the handset. These offers and combinations may vary depending on whether the customer orders from the operator’s web channel, or buys in an operator-owned or independent store. And they are likely to change from month to month, even week to week. Add to this the considerable volumes involved – perhaps millions of customers signing new contracts every year, in many different markets, spread across the world. Many operators already have difficulties managing a legacy estate of billing systems. Establishing a systems solution to cope with the ED proposals is, for some, a daunting prospect and many are likely to rely on off-system workarounds which arguably present a significantly increased risk of error.
– ARPU (average revenue per user) vies with EBITDA as the industry’s favourite performance measure. Most operators use internally, and disclose externally, ARPU measures determined using active customers and service revenues. In simple terms, ARPU is closely related to the amounts billed every month for service and, ignoring bad debts for a moment, the cash received. The accounting proposed by the ED will divorce the revenue number in ARPU from the amounts billed for service and decrease the absolute number reported – neither impact is popular with operators. Our expectation is that management teams will continue to use an ARPU measure based on on-going billings for decision-making purposes but may feel it necessary to disclose externally measures based on the new GAAP-compliant revenue figure.
In practice, the situation is actually slightly worse for the operators than Example 1 indicates. Two further provisions of the ED are likely to impact the revenue recognised by Blue. We consider the impact of each individually in turn:
– this is one of the areas under re-deliberation and, on the basis of tentative decisions to date, likely to be amended in the final Standard. Under the ED’s proposals, revenue should be measured at the amount expected to be recovered. In the case of bad debts, the operator should reduce the revenue recorded to reflect its expectation of non-payment. If Blue typically experiences a default rate of 2%, it would record 98% of the figures shown in Table 2 as its revenues over the year i.e. a total of €333 rather than €340. As currently drafted, the ED would require any true-up to reflect actual experience to be classified as other income or expense rather than revenue. The Boards have recently tentatively decided, however, that revenue should be measured without adjustment for bad debts. An allowance for expected bad debts will be presented in a separate line item adjacent to revenue and any subsequent changes in the estimate of collectability will be recorded in this line.
– the ED requires amounts receivable under the contract to be discounted if the effect of discounting is material. If Blue were to apply a discount rate of 10%, that would reduce the revenue to be recorded in the example by approximately €4 to €336. Only the element of the revenue that is attributed to the deferred payment for the handset requires discounting – the monthly service revenue is unaffected. Assuming that the customer paid the amounts due under the contract in line with the contract terms, the total receivable by Blue would be €340 and the difference of €4 would be recorded as interest income.
In practice, it is necessary to consider both of these factors together. These two effects are additive – the aggregate impact is a revenue reduction in the example of 3%.
To sum up the objections in somewhat emotive terms, operators are saying: “Why take a reporting model where revenue reflects our contractual entitlement to bill, our expectations of cash receipts and derives easily from our billing systems and replace it with a model that accelerates reported revenue, means we record an asset (accrued revenue) that we will only recover if we provide services in future and is massively complex to administer?” There are also concerns, particularly in the US, about how a model that results in the recognition of revenue in advance of the right to receive cash seems to go against views that have been expressed by the SEC in the past.
Possible commercial and practical solutions
The debate around the technical merits and practical implications of the ED is likely to continue for some time. Some operators remain optimistic that the Boards will amend elements to the ED to permit (or at least not explicitly to prohibit) an approach that gives similar answers to current GAAPs. Others reflect on the lack of changes between the original Discussion Paper and ED in their areas of greatest concern and recognise that writing a Standard that is fit for purpose to apply across industry sectors almost inevitably means that some provisions will cause difficulties for some industries.
The remainder of this paper is focused on what operators could do on the assumption that the Standard, when it comes into force, closely resembles the ED either:
- commercially to change their patterns of revenue recognition; or
- practically to apply the provisions of the standard in a compliant but efficient and effective manner.
We have assigned each option ratings (out of three •) dependent upon our views of the relative accounting technical merits and practicality – these are intended to be illustrative only; each operator will have its own views about what may, or may not, be acceptable and viable.
Possible solution A – Change the pricing model in the market:
Technical merit •••
The proposal is that operators move to charging an amount equivalent to “stand alone selling price” for the handset, but retain the total customer contract value at the same amount by reducing the tariff for service.
For illustration, below we take the Blue example and imagine Blue has taken a decision to reflect “market” rates for the handsets and service it sells.
Example 2—handset sold at fair value, service tariff reduced
Blue now charges €250 for the Azure handset, which equates to the amount charged by independent retailers when the unit is sold without a service contract. Blue has reduced the monthly tariff for 1000 minutes to €7.50. Blue’s revenue recognition profile (ignoring bad debt allowances and discounting) now matches its cash receipts and looks like this:
At first sight this may appear to be a very difficult task, but is it achievable? We think the answer is yes, but only if the market acts rationally. Customers are perhaps more sophisticated than industry observers may believe. In markets where heavy subsidies are commonplace (for example, the US and UK), customers are often well aware of the value of the up-front device subsidy – it is relatively easy to determine the price of an identical handset when sold without a contract. Moreover, the customer may be presented with a trade-off decision between initial outlay and on-going service charge. The customer buying a smartphone with a high value data plan will expect to pay less “for the handset” or to receive a more attractive device. There is little illusion that the operator’s business model involves recouping the upfront acquisition costs over the contract period and that a “SIM only” tariff for 1000 minutes is likely to be cheaper than the equivalent service provided with an expensive new handset. Moreover there may be adjacent markets (in Western Europe, Italy is an example), similarly sophisticated more remote markets (say Hong Kong) or rapidly maturing emerging markets (say India) where customers expect to pay a full market price for the device.
It can also be argued that being more open about the economics of the transaction with the customer is an inherently beneficial thing – transparency encourages trust and, perhaps, encourages customer loyalty.
A step in this direction has been taken in some markets by being more open with the customer regarding the composition of the contract charges. For example, some operators are explicit that an element of the on-going service charge relates to deferred purchase of the handset. If Blue retained the original billing profile (€100 on inception plus €20 per month) but identified €12.50 of the €20 monthly charge as relating to the handset purchase, assuming the commercial terms supported it, this would give the same revenue recognition profile as identified in table 3. Although cash and revenue remain de-linked, at least Blue’s billing system would be generating the numbers required to calculate amounts in accordance with the ED’s requirements.
There is, however, a major commercial weakness in this possible solution. It requires “rational” behaviour by industry participants i.e. all operators doing broadly the same, which may strike some as either naïve or contrary to competition in the best interests of the consumers.
Possible solution B – Focus on service provision
Technical merit ••
If an operator divorces itself commercially from the customer’s handset purchase in an appropriate manner, attribution of revenue to the device ceases to be a problem.
Example 3—operator not involved in the handset sale
Operator Green provides service in the same country as Blue. It offers an identical service tariff, being €20 for 1000 minutes per month. Green has no retail stores or web channel for handset sales and its arrangements with independent dealers do not constitute an agency arrangement as far as handset sales are considered. Green pays a commission to independent retailers signing customers up to one of its contracts which depends on the service tariff but is not related to the model of handset the customer may choose to buy from the retailer.
In this case, the only revenue generating transaction relevant to Green’s accounting for the contract is the service tariff. Green’s overall economics are similar to Blue’s – it pays more in commission but avoids the handset subsidy.
How achievable is this? In pure economic terms, it ought to be perfectly achievable and enable operators with and without handset distribution channels to compete on a level playing field. Provided the commercial and contractual relationships with dealers do not indicate an agency relationship, there is no handset sale to account for.
However, there are two significant practical drawbacks – customer behaviour and existing commercial arrangements.
– Operators may feel that having no handset retail function is an unacceptable risk if others in the same market do have such a presence. Customers may feel more confident being able to relate to the operator in a tangible manner and the operator’s brand may be reinforced by the look and feel of its physical presence on the high street or shopping mall. Placing more reliance on independent retailers may also pose commercial risk, particularly where there is concentration of the market into a small number of powerful retailers.
– Many operators have their own store portfolios and the trend has been to increase rather than decrease that footprint. However, there is an argument that good telecom operators do not necessarily make good retailers and that many customers prefer the (at least apparent) impartiality of an independent store.
Is it possible for an operator with an existing store estate to step out of the handset supply chain without a full divestment? In our view it is, provided the arrangement does not constitute an agency relationship for the provision of handsets. Two possible commercial structures are:
Possible solution C – The SAC argument
Technical merit ••• or •
- Inject the store portfolio into a minority owned business. The sale of a controlling stake to a third party would enable the operator to retain an interest in the economics of handset supply and influence on the strategy of the business. The operator would no longer have a handset sale to consider as part of its assessment of the contract with its customers.
- Store sharing. This is the retail outlet equivalent to the network share arrangements which are increasingly popular and accepted. If the share were structured to achieve minority ownership by the sharing operators, the same divorcing of handset supply from service provision as described above would occur.
If the provision of the device is not a revenue-generating transaction, then it does not impact on accounting for revenue under the contract.
Some operators currently regard the provision of a handset as a very different commercial proposition from the provision of service – while recognising that they are commercially linked. These operators argue that they are “not in the business of selling handsets” but rather are focused on providing telecom services – this is their primary activity and the supply of the handset is an ancillary pre-cursor.
In the view of such an operator, the supply of a handset is part of a commercial transaction to acquire the contract with the customer i.e. it is an element (together with dealer commission) of customer or subscriber acquisition cost (“SAC”) rather than a transaction that should be regarded as part of their revenue-generating activities.
Example 4—provision of the device as an element of costs rather than revenues
Operator Red provides telecom services in the same country as Blue and Green. Red has arrangements with a number of independent retailers who receive commission for securing new contract customers. The commission includes an element to enable the retailer to sell handsets to new contract customers at an amount below their stand alone selling price. The contractual arrangements and actual commercial behaviours support Red’s determination that the independent retailers are not acting as its agent in the sale of handsets. Red customers who sign up to 24 month contracts offering 1000 minutes for €20 per month are entitled to buy the Crimson handset for €120. The handset is available without a contract for €300. The 1000 minutes for €20 tariff is available to pre-pay customers who have no minimum contract term.
Red regards the payment of commission (including the amounts to support provision of the handset) as part of the cost of acquiring a customer contract. It treats the subscriber acquisition cost (being the cost of acquiring the customer contract) as an intangible asset in accordance with IAS38.
What are the merits and drawbacks of this approach under the proposed standard?
– although comparatively rare, this rationale and method of accounting is applied by a number of operators currently reporting under IFRS. A strong argument can be made in favour of the assertion that the costs of acquiring a contract do qualify as an intangible under IAS38, notwithstanding the fact that most operators currently expense such costs as incurred. If the supply of a handset or other device is regarded by the operator as consideration to acquire the customer contract rather than as a transaction to generate revenue in its own right, then it would not seem unreasonable to treat its elements as a cost item. For the purposes of considering the impact of the ED, it is less important whether such a cost is recognised immediately in the income statement or deferred and recognised over a period.
– there are two key problems. The first is historical practice. Assuming nothing has changed in the economics of the operator’s business, why would it be appropriate for a transaction formerly treated as a component of revenue in future to be considered as no longer revenue-generating? Many operators have provided rationale for their current accounting treatment to securities regulators – in the absence of a change in commercial substance, it may be difficult to re-characterise the supply of handsets as a nonrevenue transaction.
The second problem is in the wording of the ED, which specifically requires costs of obtaining a contract (for example the costs of selling, marketing, bid and proposal and negotiation) to be expensed as incurred. The ED also states that costs incurred in fulfilling
a contract which give rise to an asset under another standard (for example IAS2, IAS16 and IAS38) should continue to be accounted for under the provisions of that standard. The distinction between obtaining and fulfilling
may help to explain the apparent contradiction with IAS38 which, in appropriate circumstances, requires the cost of acquiring a contract to be capitalised. It is also worth noting the Boards’ tentative decision in February 2011 to amend the provisions of the ED to permit the capitalisation of the costs to obtain a contract if they are “incremental and expected to be recovered”. In fact, the classification of such costs as an asset or expense is relatively less important for the purpose of this argument; it is only necessary for cash receipts related to handsets be treated as part of a cost rather than revenue-generating transaction.
We have given a split rating for the ‘Technical merit’ of this possible solution. From a pure accounting technical viewpoint, we consider the arguments to be relatively robust and to earn a ••• rating. However, if one factors in the long history of debate over how to account for SACs and the prior attitude of many in the industry and the regulators towards capitalisation, a • rating may seem more realistic.
Possible solution D – The portfolio approach
Technical merit •••
If an operator retains its current pricing model for selling communication services and handsets, the operator could apply a portfolio approach to allocate the transaction price to multiple performance obligations. Allocating the transaction price to performance obligations based on a portfolio of similar contracts, rather than at the individual contract level for an operator’s tens of millions of customer contracts, will provide at least some practical relief on applying the proposed guidance.
Example 5 - portfolio approach to allocating the transaction price to performance obligations
Operator Yellow sells handsets and airtime to its customers. Yellow sells differing calling plan tariffs (1000 minute plans and 2000 minute plans) and a variety of handsets (six different handset models) to attract and provide communication services to customers. The stand alone selling prices of the handsets differ based on the handset model; however, Yellow may sell the handset at differing prices based on the contract period signed by the customer. Models X, Y and Z, although each manufactured by different vendors, have the same stand alone selling price. Models X.A, Y.A and Z.A are higher-end models of models X, Y and Z and therefore have higher stand alone selling prices.
The proposed guidance requires Yellow to allocate the transaction price in proportion to the stand alone selling price of the goods and service underlying each performance obligation—the handset and the calling plan tariff. Yellow signed up approximately 500,000 customers during the year and used a portfolio approach to allocate the transaction price among the performance obligations within each of the contracts. Using the contract terms signed by the customer (i.e., contract term and the handset model and calling plan tariff purchased), Yellow systematically classified the customers’ contracts into one of the portfolios described in table 4, which determines the allocation percentage between the handset and the calling plan tariff. For example, Yellow used portfolio 2.1 to determine the allocation for customers who signed up for a 12 month contract and paid €100 for the model Y.A handset and €20 per month for 1000 minutes. This resulted in Yellow recognising €1901 as handset revenue upon delivery of the handset and €12.502 per month as service revenues.
What are the merits and drawbacks of this approach under the proposed standard?
– Applying the portfolio approach, allocating the transaction price to performance obligations based on a portfolio of similar contracts, is likely to enable considerable cost savings, as compared to an allocation at the individual transaction level, and allow for systematic determination of the revenue allocation among the performance obligations.
– The portfolio approach is only likely to be acceptable if it reflects the economics of the transaction and is a reasonable approximation of the (discounted) stand alone selling price of individual contracts. This approach, although less rigorous than an individual contract basis, will still require operators to develop portfolios for the numerous varieties of tariff plans and handsets and implement processes and controls to keep the stand alone selling prices of the handsets and tariff plans underlying each of the performance obligations current, including enhancing existing capabilities of POS and billing systems. Operators who also consider possible solution C (the SAC argument) should also bear in mind that a portfolio approach does not sit well with IAS 38’s requirements for accounting for intangible assets at the contract level.