'Substance' is a widely known tax concept, especially used in cross-border tax situations. Nonetheless the expression 'substance' does not normally appear in the actual text of tax treaties. There, a number of other tests are used, such as 'residency', 'beneficial ownership', 'qualifying persons', 'base erosion' and increasingly anti-avoidance articles such as a 'general purpose' tests. In addition, modern tax treaties increasingly contain a 'general anti-avoidance rule' (GAAR).

In my eighth consecutive contribution to the Euromoney Corporate Tax Handbook I will show that there are nonetheless a number of connections between the official treaty tests used to ensure tax payers qualify for the reduction of foreign withholding (w/h) taxes and the unofficial 'substance' test which is increasingly employed by the revenue services of the world.

General discussion

In many sectors, the use of Special Purpose Vehicles (SPVs) is common practice. The bulk of these entities does not employ the type of staff that would normally be required to manage the money flows which are being collected my MNCs via these SPVs, either from their foreign wholly-owned group companies or participations such as dividends, capital gains, interest payments or royalty payments, or from their foreign customers (interest and royalty payments). SPVs also do not usually employ significant assets and they are often under a contractual obligation to pay their income onwards (interest and royalties) or habitually do so (capital gains and dividends). So even if the term 'substance' lacks international definition, it is clear that SPVs often have no 'substance' at all. Yet they are very widely used and not many people, including tax advisers, seem to fully recognise the dangers connected to this 'loose' approach. This chapter is intended to raise a red flag, because on analysis it will turn out that the substance test does have meaningful connections with the official tax treaty tests, even if the word substance itself is very seldom used in a tax treaty context.

Nations are rapidly increasing their attacks on SPVs; in all major economical regions in the world, their tax authorities are taking a much closer look at substance issues than ever before. As a result, MNCs that employ SPVs should in my view reconsider their position, in order to avoid nasty surprises in the future. Not to mention that setting up new SPVs should no longer be done the 'old school' way, because there are better alternatives available.

Tax treaty requirements and their relationship with 'substance'

In order to qualify for tax treaty benefits, which is what SPV's are all about, the SPV will in most cases have to meet two criteria generally contained in tax treaties:

  1. the SPV must be a tax resident of the State it is registered in; and
  2. the SPV must be the 'beneficial owner' of the income flow.

What does 'substance' have to do with this?

A number of countries (discussed below) do have official 'substance' rules for foreign entities seeking to reduce those countries' withholding taxes. But tax treaties are known to supersede national tax rules, so if treaties do not mention 'substance' as a formal testing criterion to determine whether or not an SPV is entitled to tax treaty benefits, why should one bother? In practice, in my 35 years of experience with SPV tax planning, as regular counsel to several 'heavy users' of SPVs, i.e. the venture capital sector and the fiduciary services sector, not many people seem to bother indeed! So let me tell you how dangerous this might be.

Substance and the tax residency test

The SPV must be tax resident in its country of establishment. This sounds rather simple: the MNC incorporates a subsidiary under the national laws of a well known treaty shopping jurisdiction (Luxembourg, Malta, Cyprus, the Netherlands, to name a few) and that is it.

However, if that entity is effectively managed from abroad (by the parent company of the group in case the SPV was set up to collect intra-group income) or by a foreign UBO (in case the SPV was set up to collect income from his foreign customers via an SPV), tax residency in the country of incorporation of the SPV is not at all automatic. The tax authorities of the countries where the income is being generated could well take the position that the SPV is not a resident of the other treaty country at all because it is effectively managed from another jurisdiction. The fact that in some countries legal entities incorporated under their national laws are deemed for tax purposes to be tax residents does not really help: these countries have tax treaties with other countries which contain the usual 'tie breaker' rule that if these entities are effectively managed from that other treaty country, they are no longer tax resident in their country of incorporation.

'Effective management' (or 'management and control' which is the term used in Anglo-Saxon tax treaties but its meaning is very similar) is essentially a substance issue. If the SPV lacks the directors who are really the decision makers on what to do with the incoming dividends, capital gains, interest payments or royalties, it exposes itself to a 'lack of substance' attack by the foreign tax authority which might lead to loss of fiscal residency in the SPV country. So simply setting up a, say, Luxembourg finance company and then thinking that the mere incorporation of such an entity and having it registered with a fiduciary services organisation in Luxembourg will give the group access to the Luxembourg tax treaty network is naïve, to say the least. The entity will require 'substance' even if the Luxembourg tax treaties do not contain any reference to this concept. In the transfer pricing (TP) area this issue is referred to as 'significant people functions' and is now a basic part of any modern TP analysis.

A rather scary example of how judges deal with this, can be found in the British 'Indofood' case. In this case, a UK judge decided that the Indonesian Government would likely dismiss a proposed Dutch SPV, which would solve the problem under litigation, as being the beneficial owner of certain Indonesian interest payments, under a domestic Indonesian ''substance'' tax approach in combination with the Dutch/Indonesian tax treaty.

Substance and the beneficial ownership test

Things get even worse for SPVs under the beneficial ownership test which can be found in almost any tax treaty. A lot has been written about the beneficial ownership concept itself and interested readers may want to read the thesis by Dr. Charles Du Toit called 'Beneficial Ownership of Royalties in Bilateral Tax Treaties', published by the IBFD in Amsterdam in 1999. The thesis offers an excellent legal view of the subject, but also the articles in European Taxation (also published by the IBFD in Amsterdam) in September and October 2010 by TP specialists Steef Huibregtse and Louan Verdoner from the Transfer Pricing Associates Group, assisted by Dr. René Offermans (IBFD research associate) offer an insight into the economical aspects of beneficial ownership.

An optimist would, also after reading these fine contributions, say "so, the beneficial ownership concept is far from clear and still full of ambiguities, legally and economically, so let's not worry too much about it". Well, I cannot stop anyone from putting their heads in the sand, but this is certainly no survival strategy. The amounts of dividends, capital gains, interest payments and royalties that flow through SPVs every year are so vast that a 'better safe than sorry' approach seems clearly indicated.

From all that has been published about the term 'beneficial ownership' it can be derived that it in fact means that the receiving legal entity, even if it is a tax resident of the other treaty country, must earn the income for itself and not for someone else. The receiving entity should not, either directly or indirectly, have to pay the dividends, capital gains, interest or royalties it receives onwards to a party outside its country of tax residency.

For dividends and capital gains, this test is relatively easy to meet (there are not many situations in which an intermediate holding company in a group is contractually obliged to pay any capital gain it realises or any dividend it receives, onwards to its own parent company). But even if there is no contractual obligation, if the entity habitually does so, this may well be used against the SPV by the tax authorities in the paying countries because the signal is then that the SPV appears to lack 'significant people functions', i.e. directors who would be able to put the dividends and capital gains received to good use at the SPV level itself (e.g by reinvesting the money in other subsidiaries than where the dividends or capital gain came from).

But for interest and royalty payments, the position is even worse. Most if not all SPVs that collect these income types for a group company or UBO (both of which usually employ another SPV established in a tax haven jurisdiction) are under a contractual obligation to pay what they receive onwards, less an appropriate 'commission', to a destination outside their country of establishment.

In a recent Canadian case ('Prevost') that was litigated before the Canadian Supreme Tax Court, the court took a very legalistic approach. A Dutch SPV that owned a Canadian subsidiary and was under a contractual obligation to pay 80% of its dividend income out to its two parent companies (this was JV set-up) and was clearly inserted in the money loop to reduce Canadian dividend w/h tax, was still judged to be the beneficial owner of its dividend income, irrespective of the onward payment obligation.

In this respect the Canadian 'Velcro' case should perhaps also be mentioned because in that case a very lenient lower Canadian Tax Court judged that if a Dutch SPV was contractually allowed to take 30 days before having to pay its royalty income onwards to a Curacao based group company, it was apparently ''able to avail over its royalty income and should be considered the beneficial owner of it''.

I am rather afraid that tax courts in many other countries will not be so lenient even though the Canadian Supreme Court relied on an in-depth analysis of the OECD commentary concerning the beneficial ownership notion in its Model Tax Treaty which was found to be applicable to the Dutch/Canadian tax treaty. And in my view, beneficial ownership of dividends and capital gains is easier to defend than beneficial ownership of interest and royalty conduit entities. SPVs involved in collecting these money flows for their UBOs (group companies or other) are usually obliged to pay their receipts onwards within five working days.

Incidentally, I have been amazed for a long time that tax authorities do not simply ask for a copy of the annual accounts of SPVs (or entities they suspect to be SPVs) for a review of their beneficial ownership status. The reason for my amazement is that in 98% of the cases, such annual accounts immediately reveal that over 90% of the dividend, capital gain, interest or royalty income is paid onwards. Most treaties contain an exchange of information article, so obtaining a copy of the often very revealing annual accounts of the receiving entity would be a simple way of checking for the revenue services in countries that doubt the entitlement to treaty benefits of the foreign entities to which dividends, capital gains, royalties or interest are paid as a 'first tax sanity test'.

Instead, tax authorities have asked the OECD to come up with 'beneficial ownership guidelines' which the OECD is hard at work on, but a first version published several months ago for public consultation was judged to be overly restrictive so a second version will follow, likely before the end of 2012. In addition, tax authorities are putting pressure on their governments to develop national 'substance tests' for foreign entities. Apparently they do not seem to realize that such national tests can be found to contravene tax treaties and the OECD comments and it is likely that tax judges also outside Canada will take a close look at the Prevost case, because of its thorough analysis of the OECD commentary of the beneficial ownership issue.

Over the last two years, one can observe a very substantial increase in the interest of tax authorities in the beneficial ownership test in tax treaties. With the exception of a few countries (the US and Germany) this test used to always get ignored. Tax auditors are usually not lawyers and have always found it difficult to apply that test in practice. The residency certificates which are part of any tax treaty benefit claim used to be good enough to convince them that the reduction of withholding tax was in order. But those days are over for good. The European Commission has now also promised to bring out guidelines on the beneficial ownership issue, soon, another signal that should not be taken lightly.

Anyone who continues to ignore the (indirect but significant) connection between 'substance' and the beneficial ownership test in tax treaties when structuring SPVs might well deeply regret this in a few years time. And because governments are looking for tax money more intensely than ever before due to the financial crisis, as soon as their government, the OECD or the European Commission gives them the weapons, the revenue services will start using them. And it is clearly not enough to say ''well, we will likely prevail in court'' because once a case is being litigated the SPV is 'dead in the water': no CEO or CFO will want to take the risk to continue that SPVs activity or set up a similar one, until the case has been litigated, which will usually take five years or more.

National substance rules for foreign entities (Germany, China)

Back to the basic question: what does 'substance' have to do with 'beneficial ownership'? The scope of this chapter forbids me to discuss a multitude of countries (and I may again refer the reader to the articles in European Taxation in September and October 2010 mentioned above, which do discuss the rules in a variety of countries), but let's look at two tax jurisdictions with substance rules, an old one and a new one.

Germany

As regards Germany, every international tax specialist knows that this country always had the strictest 'substance' test. It should be noted in this respect that these tests have been relaxed lately based on the so-called 'Cadbury-Schweppes' case before the EU Court of Justice in Luxembourg. This case made clear that the German approach was out of line with the Rome Treaty, but the revised German tests are still quite strict and in my view still out of line with the Rome Treaty, but who wants to set up a structure knowing he has to fight it in court?

Germany uses a double approach: legal and economical. The legal rule (art. 39 of the General German Tax Code) is that the recipient is the beneficial owner if he has the right to avail of the payment received as he deems fit. In addition there are administrative guidelines which contain additional substance requirements, such as:

  1. The recipient entity should also have other meaningful business than collecting the German dividends, capital gains, interest payments or royalty payments.
  2. The SPV should have its own office space, ICT systems, etc.
  3. The SPV should have its own personnel, which includes employment contracts with such employees with salary commitments, payroll tax and social security premiums withholdings etc.
  4. It should pay salaries to its employees (which differs from 'incurring salary expenses'). Salary slips are not enough, the payment must be proven by being able to show the actual bank transfers.

Any foreign entity, if investigated by the German tax authorities, that does not meet these substance criteria will be considered to fail the beneficial ownership test. Interestingly, in many German treaties the beneficial ownership test is not present, by the way. The administrative guidelines seem to be a way to adjust for this. So they may not hold up in court. But who wants to try?

Are there still readers who think that setting up a simple SPV in, say Cyprus or Holland, that then starts to collect German dividends or royalty payments, is a workable solution?

China

China has not been a country with major international economical relationships for a long time. But they are catching up with lightning speed, in all respects, also international taxation. Very recently a Circular (nr. 2012/30), was issued which legalised an earlier internal instruction to the State Administration of Taxes (SAT) which contains the following 'substance' rules, before China will accept a foreign (SPV) entity to be entitled to the Chinese tax treaty benefits. No reduction of the Chinese w/h taxes will be given if:

  1. the recipient entity pays the income onward for more than 60%, to another jurisdiction than its own;
  2. the Receiving entity has no 'business substance' (not further defined) and operates with very little assets;
  3. the income is taxed at minimal rates in the hands of the recipient; or
  4. the recipient is under a contractual obligation to pay its receipts onwards as such to parties outside the recipient's jurisdiction.

National substance rules for national entities

The IBFD, on the request of the Dutch Ministry of Finance facing questions from Dutch parliamentarists on the wide spread use of Dutch SPV's by MNC's, investigated in early 2012 whether neighbouring countries in Europe have substance rules for their own legal entities. The investigation showed that this is not the case. Everyone has substance rules for foreign entities but not for local ones, except the Netherlands itself. Under pressure from the European Commission, the Netherlands introduced national substance rules for holding companies ("dividend and capital gains conduit companies") and interest and royalty conduit companies, i.e. the entire list of SPVs (which the Netherlands has more than 10,000 of) in early 2001. A transition rule allowed existing situations five years to adjust so as from January 1, 2006 the Dutch national substance rules can be summarised as follows:

  1. The entity should have sufficient equity (transfer pricing study required).
  2. The equity should actually be at risk (no non-recourse situations).
  3. The entity's gross profit margin should beat arm's length (transfer pricing study required).
  4. At least 50% of the directors should be permanent Dutch residents (nationality irrelevant).
  5. The directors should have proper professional qualifications in order to manage not only the entity but also its money flows; no 'dummies' allowed.
  6. The books must be kept, and the annual accounts should be prepared, in the Netherlands.

The Dutch fiduciary services sector that manages the bulk of the Dutch SPVs which have registered as such with the Dutch central bank (DNB) has more or less adapted to these rules and the Dutch substance tests are not very hard to meet, except perhaps the equity test. In many cases, especially with big SPVs, the Dutch entity will be required to have a share capital of Û2m.

The introduction of these national substance rules have convinced the European Commission in 2001 that Dutch SPVs if they adhere to these rules, are not violating the EU 'Code of Conduct' and are therefore not 'Harmful Tax Competition within the EU'.

Obviously, the Dutch substance rules are not decisive: the other treaty country is free to impose its own substance rules in the areas of accepting Dutch tax residency of the SPV and its beneficial ownership of the income.

The Netherlands, finally, committed to 'Brussels' that if they find, upon a tax audit, that a Dutch entity does not meet the Dutch national substance rules, the Netherlands will inform the foreign tax authorities in the countries where the dividends, capital gains, interest payments or royalties come from, that the Dutch revenue service does not consider the entity to be the beneficial owner of its income. The Dutch Act on the Exchange of International Tax Information was adapted, effective April 1, 2001 to allow for such 'spontaneous' exchange of tax information. To my knowledge such an exchange has not taken place in my country yet, but the threat is there. The only thing that keeps the Dutch revenue service from writing these nasty letters is that the tax handling of most SPVs is concentrated with the advance tax ruling team, a special division of the Dutch revenue service in Rotterdam, which does not employ tax auditors. All they do is a desk review (by itself understandable because SPVs are relatively simple entities by nature).

A curious side effect of these rules is that, in line with the OECD comments on beneficial ownership, a Dutch entity that does not meet the Dutch substance test is deemed to receive the income for someone else, so it will not have to declare the interest or royalty income as income. Capital gains and dividends can be received tax free in the Netherlands (with some planning) anyway so this special non-taxation clause (art. 8c, par 1) Corporate Income Tax Act (CITA) applies especially to interest and royalty conduit SPVs. In case the SPV is collecting group income, it must report a fictitious gross margin based on the cost-plus method, but if the entity is unrelated, it does not have to report its income. However, this is exactly what the OECD commentary implies.

Tax treaties with an LOB provision

The latest trend in the worldwide SPV combat by tax authorities is that countries are starting to copy a special article which they have in their tax treaty with the US to other tax treaties. This special article, called a limitation on benefits (LOB) provision, is then added to the residency and beneficial ownership clauses in a tax treaty and, basically, says the following:

  1. The recipient company of the money flow must be ultimately owned by persons who are residents of the same country as where that recipient company is established ('qualifying persons test'). Unless the entity is owned by residents of a country which also has a tax treaty with an LOB provision with the source country ('equivalent beneficiaries test').
  2. The receiving entity should not, directly or indirectly, pay or accrue, more than 50% of the income it receives as tax deductible payments of any kind to foreign recipients in countries that do not have an equivalent treaty with the paying country as regards the LOB ('base erosion test').

The Netherlands was amongst the first countries that had to accept an LOB provision in order to maintain a tax treaty with the US but since then (1992), the US has renegotiated almost all of its other tax treaties to contain a similar LOB provision. LOB provisions are a new element in tax treaties and a significant factor to reckon with when setting up new SPVs, and recently other countries have also started to demand LOB provisions in their new tax treaties. A good example is the new Dutch/Japanese tax treaty, effective January 1, 2012. The Japanese were willing to give up their withholding tax of 10% on interest and royalties under the old treaty with the Netherlands , provided the Netherlands signed up for an LOB provision (which it did).

To me this is clearly the future of tax treaties: tax authorities have learned by now that beneficial ownership clauses were no guarantee against treaty shopping (although they should have been) and instead of relying on the OECD to come up with guidelines that may not be narrow enough to cover SPVs, they copy 'the American way'. Interesting in this regard is the remark, by the Dutch Ministry of Finance when the Dutch/Japanese treaty was offered to Parliament for ratification, that one of the consequences of this new treaty would be ''that simple SPV structures which give the Netherlands a bad name, will no longer be effective''. I expect to see many more tax treaties with LOB provisions the coming years.

Interesting is the word 'simple' which the Ministry of Finance used in its presentation of the new treaty with Japan. Apparently the Ministry realises that if an SPV is 'dressed up' to meet the beneficial ownership test and the various tests contained in the LOB provision , there may still be possibilities to set up such structures in the Netherlands, even for interest and royalty collections from Japan (and from the US for that matter, because the LOB provisions are rather similar; the Japanese version is in fact a 'light' version of the US one).

Conclusions and recommendations

I hope to have succeeded in putting the tax concept 'substance' into a more legal and economical embedding in the above. Even though the term 'substance' is seldom used in tax treaties, it does play an important role in the background, via the residency and beneficial ownership articles in those treaties. Therefore, lack of substance can in my view sometimes indeed be used by tax authorities to deny an SPV tax residency in its country of incorporation/establishment and lack of substance can also be a powerful weapon for them to attack SPV entities on failing the beneficial ownership test and/or the various LOB tests. The playing field for SPVs is clearly getting smaller, in any jurisdiction.

Are these developments the end of SPVs then? To me, like any other tax planning situation, this is very much a 'cat and mouse game'. If, like the Dutch Government says, the new treaty developments (which will no doubt spill over to other countries) are a threat to simple SPVs, the answer should lie in 'unsimplifying' them. This may cost some extra effort and money but our fiduciary services division already employs techniques that overcome the threats and lead to SPV structures that will pass the residency test, the beneficial ownership test, the Dutch and foreign substance tests, the qualifying persons test and the base erosion test. Based on the fact that we offer free second opinions on international tax structures, I have in the meantime already seen a number of creative solutions. And I myself have not been sitting on my hands either. As an SPV specialist, my final advice would be: do not keep it simple and consider our new solutions. Also in tax, if old roads close, new ones usually open up.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.