Maximise the value, minimise the tax compliance risk

As far back as 2001, the UN noticed the relevance and importance of allocating appropriate value to activities within as well as between multinationals. In that year, the Ad Hoc Group of Experts on International Cooperation in Tax Matters identified that over 60% of international trade was carried out within multinationals, so an actual majority of trade was seen to be between connected parties.

The transfer pricing of goods, services and assets is therefore a hugely important concept for understanding what value is generated where and, therefore, how it should be taxed.

On one side, transfer pricing is important because it can be used by corporations to minimise or avoid their taxable income in high tax countries thereby reducing the tax base of countries already struggling with tax income reduced by the so-called Great Recession.

There are many examples of companies using corporate vehicles or opaque accounting practices seemingly for the sole purpose of reducing tax. One need only look at the recent example of the IRS’ objection to Amazon’s sale of its EU trademark rights to a Luxembourg subsidiary to see tax authorities’ fear of this happening.

On the other side, it is important because transfer pricing can expose multinationals to significant amounts of double tax. For example, in 2006, GSK was forced to pay US$3.1 billion in double taxation of its income when arbitration between US and UK tax authorities could not agree on the amount of value created in building the Zantac and other brands in each of their representative countries.

The key challenge, when it comes to intangibles, stems from the underlying “separate and independent” principle adopted by Article 7 and Article 9 of the OECD’s Model Tax Convention – the model for OECD country bilateral tax treaties. These articles essentially state that profits accruing to a branch or company within a group should be equivalent to what you would expect to accrue to an independent party.

This principle has also been incorporated into the UN’s Model Tax Convention and the default tax conventions of many countries, such as the US. Collectively, these model conventions act as the basis for the clear majority of bilateral tax treaties meaning almost all countries tax inter-company income under this principle.

This means that the functions performed, assets owned and risks borne by group companies should command payments commensurate to what one would expect a third party to receive in an arm’s length transaction.

Herein lies the challenge: how does a tax manager establish where value would be created or charged for in an arm’s length transaction between independent parties when the parties have not been acting independently or at arm’s length for many years?

So how do you set the correct rate of return for the use of a brand?

Action point 8 of the OECD’s BEPS initiative relates to intangibles. Intangibles are identified as one of the highest risk issues in transfer pricing as a result of: the lack of comparable transactions due to the unique nature of such assets; the difficulty of separating earnings specifically related to the IP from the earnings derived from the rest of the business; and the difficulty of determining ownership over their value.

Lack of comparables

The OECD specifies that taxpayers should use the “most appropriate” method for determining a transfer price. This can be any method provided it is agreed that it is the most appropriate. However, as a guide, it specifies five methods that can be used. In practice, only two can be used for the transfer pricing of intangibles:

1. The “Comparable Uncontrolled Price” (CUP) method which determines a price based on similar agreements – usually with an adjustment to make the fee seen in the market transaction more comparable – and;
2. The “Profit Split” method which determines the price by identifying the profits made from the transaction and attempting to split these profits according to each party’s contribution.

Establishing comparability in transfer pricing is notoriously hard. Companies operate in slightly different geographies, sell slightly different products, at different times, and with brands that have different strengths in the minds of consumers. Databases for agreements are small, often out of date, skewed towards particular regions (in particular the US) and contain limited information on the underlying businesses to the transactions. This causes significant issues when using comparables to defend a transfer price. In the UK’s landmark 1997 case, DSG retail v HMRC, every single one of the comparables put forward was rejected for various reasons that made them non-comparable. This inevitably means that most documentation will require a combination of the CUP method and the Profit Split method.

Some practitioners consider the Profit Split method to comprise solely of what is known as the “25% Rule” or “Rule of Thumb” which assumes that most IP licensing negotiations start with an assumption that a licensee will pay between 25% and 40% of their earnings before interest and tax on licenced IP. This ‘rule’ is widely discarded by tax courts due to incomparability and is therefore at best a useful corroborative check after a more fundamental review of the impact of the functions performed, assets owned and risks borne of each party to a transaction.

Determining the value of IP earnings

The Amazon.com, Inc. v. Commissioner 148 T.C. 8 (2017) case confirmed what most intangible asset valuers and transfer pricing consultants already knew: that the best method for determining the value that a brand owner should receive is a royalty on sales. Where a valuation is necessary, this precipitates a discounted cash flow method known as the “Royalty Relief” approach whereby the value of a brand is determined by the present value of hypothetical intercompany royalty payments.

The challenge with this method comes in determining that royalty. Since brands and other IP are unique both in strength and in application, it is generally necessary to at least make comparability adjustments to market rates. In the case of Amazon, it was necessary to review the business model since profits were kept deliberately low through reinvestment in order to drive higher revenues. The brand was therefore valuable for driving growth rather than for driving profitability and therefore it was determined that the final rate on sales must be lower than comparable market agreements since sales were higher and growing faster than in those transactions.

However, rather than making arbitrary changes to non-comparable “comparables”, it is generally preferable – given the paucity of even vaguely comparable transactions – to assess the overall earnings from a brand from first principles. This is also beneficial from a commercial perspective since it can also be used as a tool to help management use their brands to maximise earnings and therefore their business’ value.

In various cases we, at Brand Finance, have established value via the statistical analysis of brand attractiveness data to understand brands’ impact on demand (and some costs), comparing what earnings would be under a “generic” brand and identifying the difference to estimate the yearly uplift. This is a challenging approach but it tends towards a more satisfactory approach when faced with strong, unique and large global brands.

A related challenge occurs when considering the case of rebrands. Head office might decide to rebrand in order to reduce central costs or to create a strong global brand but in the short run this may actually destroy brand value in the rebranded company. The net effect may be neutral or negative and so many tax authorities will argue that a charge to a brand outside their country is unacceptable (since there is a perfectly good brand that could be used and would keep value in their own country).

The case of Suzuki Maruti in India highlights this issue. When Suzuki acquired Maruti it decided to cobrand under ‘Suzuki Maruti’ and charge for the use of the Suzuki brand. Indian authorities argued that the effect of this cobranding was that Suzuki was actually building its own value in India by associating itself with Maruti which was a strong brand. They argued that Suzuki should actually pay Maruti for the privilege – a reverse royalty.

The result ended with no royalty chargeable but the case highlights the fact that taxpayers need to be careful when charging royalties to rebranded subsidiaries either by waiting to charge or by charging a ratcheted royalty dependent on performance.

Determining ownership and the right to make a charge

The new BEPS rules – as well as the domestic rules now being adopted by many states – outline that while legal ownership is the starting point, the ability to charge a share of profit rate of return is more dependent on economic ownership of the asset. Legal ownership refers to contractual rights whereas economic ownership refers to rights created through Development, Enhancement, Maintenance, Protection, and Exploitation (DEMPE) functions that create value in the IP. The idea is that, since a group relationship is not exactly equivalent to a licensor-licensee one, group subsidiaries sometimes spend time and energy adding value to IP that a normal licensee would not be expected to do.

The OECD therefore believes that assets should be owned by (and therefore charged from) the parties that generate the value since, in the absence of tax considerations, that is where they would legally be registered and owned in many cases.

Often, tax authorities take this principle and assume it is equivalent to where money is being spent on marketing and advertising. However, when thinking again of Virgin, one can see why this argument does not hold up to scrutiny. As we have seen, Virgin is involved in strategic decision-making and legal protection but does not spend much itself. Despite this, it is still able to charge high royalties to a large number of parties.

Granted, there are examples of companies where central spend is high – if you consider the amounts spent by Nike on global celebrity endorsements you will see how this can be the case – but typically licensors do not spend much and instead expect licensees to spend large amounts on promotion in their markets, albeit under strict rules.

Therefore, to determine where economic ownership lies, one must establish what constitutes ‘strategic’ spend and activities and what constitutes ‘tactical’ spend and activities, and then split these. Preferably this will have been completed before an audit, because it is easier to state ownership is central when strategic spend or activities (that will often occur in subsidiaries) have already been compensated for.

Taking a different type of IP as an example: if a new type of Mars bar were created in Mars’s UK subsidiary and the group decided it wanted to sell it around the world, it is easier for the centre to claim ownership and charge a share of profit from the centre when it has already paid the routine payments for the R&D facilities that created that product. Otherwise, the UK would be the party that owns it and can charge for it.

In the case of a rebranded enterprise, it should be made clear that as well as making the change in brand there should also be a scaling back of strategic activities to the centre. In practice, this is generally easy since a primary reason of rebranding is often to reduce headcount and cost in marketing and brand functions.

Conclusion

As noted in part one of this article, the management of brand and other IP is often ad hoc at best in large, complicated multinational enterprises. This is very understandable at least partly because of four core issues: investors do not demand scrutiny since they do not see the assets’ value on balance sheets; management knowledge of how brands and other IP impact businesses mean information is often limited; brand managers have historically lacked the knowledge to explain the importance of the assets they manage; and tax authorities did not care much until very recently.

Whilst the Brand Finance GIFTTM study highlights the issue of companies not understanding their IP well, the new BEPS rules highlight its importance. Tax authorities are getting more and more aggressive coming after big companies they deem to be exploiting the principles of transfer pricing to avoid tax. The transfer pricing of companies’ brands is one of their biggest areas of contention.

This highlights the need for companies to confront the complicated issue of charging for brands. The steps to confront it are as follows:

1. Identify where all intellectual property is legally owned and registered and identify where new IP is likely to be developed.
2. Identify where each piece of intellectual property is being used.
3. Review marketing and IP related functions performed, assets owned and risks borne by each group company to establish where “DEMPE” activities are being performed.
4. Analyse customer and other research to determine the likely impact of the brand’s reputation on demand drivers and cost pressures, and translate this into an earnings figure.
5. Review comparable licencing transactions and licencing best practice in your industry in order to identify what are common DEMPE obligations for licensor and licensee and to find comparable rates.
6. Decide on the appropriate royalty rate and determine whether any compensation should be paid for value-adding activities by brand users that would not usually be expected of a licensee.
7. Repeat the process for all other IP, markets and group companies and document in country by country reports.

Although a large task, with the prospect of large audits on the horizon for many companies, the diligence of this approach should avoid the threat of double tax that can rise up in to the billions. While financial accountants still avoid thinking about the huge value of IP that companies control, it is clear managers and tax planners cannot.