Country by country: a case study

Posted on

Tim Worstall has raised questions with regard to country by country reporting. As he says:

Country by country reporting is being held up as the solution to so many international tax woes. But could someone please tell me quite how it does so? Tax should be paid where the economic substance of the activity takes place. That’s the mantra isn’t it? It’s just that I’m not really sure what that means.

He proceeds to give an example of a situation he says he knows of:

Take for example this little international business I know about. I’ve changed the country names, the product name and the margins but the substance of the following is true.

OK, first purchase in Tajikistan some low grade europium oxide. Purchase at $1,000 per kg. Ship it to Warsaw where it is upgraded to high grade europium oxide. This costs $1,000 per kg raising the value to $2,000 per kg. Then ship it to Canada where it sells for $3,000 per kg.

Now, yes, I can see that the whole process creates value of $2,000 per kg. But what I find very difficult to understand is how we’re going to allocate that over the different places.

He then complicates the issue:

For there’s a fourth location involved as well. Tunisia. Where resides the information about how to put this all together. The intellectual rights if you wish. Just about no one other than the man in Tunis knows how to do this. They don’t know where to get the low grade oxide. They don’t know where to get it refined. They don’t know who in Canada is willing to pay for it. Quite literally (put it this way, the Canadian factory hasn’t asked for a competitive bid for europium oxide for a decade, let alone received any unasked for approaches) no one else knows how to put this process together.

He then asks:

OK, so where is the economic substance of this series of transactions? I can see that the original purchase gets taxed in Tajikistan. I’m not sure at all about the $1,000 on the upgrade in Warsaw. For that wouldn’t happen at all without the involvement of Tunis, so surely some of that value should be allocated there?

And the next $1,000. That’s not value added in Canada, is it? Or Warsaw? So does it get taxed in Tunis? The place that created this rise in value? Or does it go entirely the other way around? $1,000 in Tajikistan, $1,000 in Warsaw and $1,000 in Canada? But then what of the value added by Tunis? Where, given the proprietary nature of the knowledge involved in all of this, arguably the real economic substance is taking place?

As he says:

Note that that there is nothing taking place "offshore". No special structures, no hiding anything. It’s just that I cannot see any hard and fast rules about where the value is added, where the economic substance resides. So I’m not sure what benefit there is to country by country reporting.

I trust he’ll forgive being quoted at length: he said he wanted answers. Let me provide them.

First, let’s be clear that there are many more reasons for country-by-country reporting than tax. I have detailed those here. Let’s not assume that corporation tax is the sole issue here. Holding the government of Tajikistan to account for the royalty due on the europium oxide would be another one. As would the shareholder concern for exposure to the politically vulnerable regime in that plae be another one. Tax is not the be all and end all of this issue.

But it is important: I’ll concede that. So let us consider the issue Tim raises. Firstly, and rather oddly, he does give no indication as to whether the parties (let’s assume they’re all companies) in Tajikistan, Poland, Canada and Warsaw are under common control. If they aren’t there is no tax problem here. Each will be operating at arm’s length. We have to assume unrelated companies are negotiating the best deals available and as it is said that there are no offshore shenanigans going on in that case the resulting profits will, we have to assume, be taxed in the right place. The profit in extracting the ore will be paid in Tajikistan. The profit on processing it will be paid in Poland. The fee paid to the Tunisian owned company that provides the know how (presumably from the Polish company who appears to be at the core of this whole deal in that it buys, sells and processes the product and therefore presumably is the user of the Tunisian ‘intellectual property’) will logically be taxed in Tunisia. The processed ore is sold to Canada according to the stated facts: so what one might ask? In this case does the location of the customer determine the place of a taxable profit? There is no indication what the customer does with the purchased ore and so the answer has to be open ended.

We have in this scenario country-by-country reporting. Each entity is paying tax where they are located. If accounts are required to be placed on public record then some trace of the transaction will be open to public inspection. If not, then there is need for local reform, but that is the end of the story if the parties are unrelated. If that is what Tim Worstall had in mind (and it seems possible) then hardly surprisingly he does not understand what country by country reporting is meant to achieve for it adds nothing in this case, and is not meant to.

Now take an alternative hypothesis: let us assume all these companies are under common control. Let’s assume that the Canadian enterprise acquires them all — including the intellectual property previously in the possession of the Tunisian. Let us also assume that prior to acquisition by the Canadian company the trades looked like this:

Tajikistan

Poland

Tunisia

Canada

Notional total

$

$

$

$

$

Sales

- 3rd party

1,000

3,000

300

3,000

- Intrgroup

Purchases

- 3rd party

700

1,000

3,000

700

- Intragroup

Processing costs

700

700

Intellectual property

- 3rd party

300

- Intragroup

Finance costs

Profit

300

1,000

300

1,600

Tax

90

300

90

480

Profit after tax

210

700

210

1,120

Total value added is $1,600. It is not the sum of $2,000 Worstall assumes as I have assumed there are costs of buying / extracting the ore in Turkistan and I have allocated a payment to the intellectual property.

After the Canadian purchaser of the processed ore has acquired all the companies things could look very different. Three things are likely to happen:

1. The Canadian company will have costs of acquisition — let’s call them $300 of interest a year;

2. The intellectual property will be moved to a tax haven;

3. The IP charge will be extended to cover the whole supply chain.

Transactions might now look like this:

Tajikistan

Poland

Tax haven

Canada

Total

$

$

$

$

$

Sales

- 3rd party

3,000

3,000

- Intrgroup

1,000

2,300

700

Purchases

- 3rd party

700

700

- Intragroup

1,000

2,300

Processing costs

700

700

Intellectual property

- 3rd party

- Intragroup

200

400

100

Finance costs

300

300

Profit

100

200

700

300

1,300

Tax

30

60

0

90

180

Profit after tax

70

140

700

210

1,120

Mysteriously the figures for notional real added value within the transaction are the same: $1,120 is earned after tax on processing the ore either way. But note that in the consolidated group (the total column for which represents the data that would be published in the group accounts) the profit is reduced by $300, but so too is the tax bill.

How is that? The group accounts would not, of course, say. The accounts for the tax haven subsidiary would be totally hidden from view. They would not be published in the tax haven either.

But if the group accounts had to include a country by country report that report would look like this (and I’m assuming processing costs are labour):

Tajikistan

Poland

British Virgin Islands

Canada

Total

$

$

$

$

$

Sales

- 3rd party

3,000

3,000

- Intrgroup

1,000

2,300

700

Purchases

- 3rd party

700

700

- Intragroup

200

1,400

2,400

Employee costs

700

700

Finance costs

300

300

Profit

100

200

700

300

1,300

Tax

30

60

0

90

180

Profit after tax

70

140

700

210

1,120

Now the user of the accounts would see the supply chain. They would see where it was located. They would see that intra-group trading exceeded third party sales. They would see that a significant part of profit was located in the BVI, and probably is not available for distribution to them as shareholders as a result. They would see that the Canadian operation is just a sales front: the activity is in Poland and Tajikistan, places they might think a lot less stable. They would realise the whole supply chain was dependent on one source of supply. And that all processing takes place in one location with a government that is volatile (and some think Poland’s is). They would see that the tax charge is obviously open to challenge by the governments of Tajikistan, Poland and Canada.

How could that tax charge be challenged? In these ways:

1. That the transfer pricing to the BVI was wrong;

2. That the transfer pricing between Poland and Canada was wrong;

3. That the finance cost in Canada was for a business activity undertaken elsewhere and so not allowable (I am speculating here for example’s sake: I do not know Canadian law on this point);

4. That in the locations where real economic activity occurs, that is Tajikistan and Poland in particular, where ore is extracted from the ground and real people are employed to undertake real labour in what is, presumably a real processing plant, relatively low profits are earned.

This would, of course, be highlighted by a unitary apportionment formula allocation of profit (for which there is not enough data here, but there would be in a full set of country by country accounts). Under such a unitary formula approach, which is widely used to allocate profits between companies operating in different states within the USA and has therefore been extensively tried and tested, the total group profit is allocated to locations on the basis of a formula. The classic formula is called the Massachusetts apportionment and it allocates profit on the basis of a formula that gives equal weighting to third-party sales, employees and physical fixed assets made from or located in a jurisdiction. One of the reasons for requiring employee and fixed asset information to be disclosed under country by country reporting is to ensure that sufficient data is available to allow a unitary apportionment formula allocation to be undertaken with regard to any group of companies to which country by country reporting would apply to determine whether its profit allocation looks reasonable or not on that basis.

Using a formula apportionment method means that artificial reallocations of activity within a group can be largely eliminated from consideration when deciding the likely profit arising within a group. This is especially so if third party sales are stated net of intra-group purchases to prevent their artificial reallocation to locations like Ireland. As a result the objectives of the ‘arms length principle’ or transfer pricing that the OECD promotes as the ideal solution to solving transfer pricing disputes is achieved but with much less effort and accounting being required than is the case under its chosen bilateral approach.

So is there a reason for country-by-country reporting? I think so. Can it be used to assist profit allocation between states in a more effective fashion than the existing OECD rules? I think so. Will this data provide invaluable information for shareholders as to the risk they take? I think the answer is ‘undoubtedly’. Will it discourage tax haven abuse? I’m sure so.

I could write more: but the point is, I hope made. Country-by-country reporting can and will work.


Thanks for reading this post.
You can share this post on social media of your choice by clicking these icons:

You can subscribe to this blog's daily email here.

And if you would like to support this blog you can, here: