Capital gains tax avoidance: can Uganda succeed where India didn’t?

Zain
Uganda is pursuing Zain for $85m capital gains tax on the indirect sale of its Ugandan subsidiary

I’m writing this post from under a mosquito net on a close Kampala evening. Since arriving on Wednesday I’ve had a whistlestop tour of the issues facing Uganda as it embarks on a review of its tax treaties. So far I’ve met with four tax inspectors, two finance ministry officials, four (count ’em) tax advisers, one academic and three NGO people. I also spoke at an event to launch a a report on Uganda”s tax treaties written by Ugandan NGO SEATINI and ActionAid Uganda.

This post is about “indirect transfers” of assets, where a sale is structured to take place via offshore holding companies, thus escaping capital gains tax. It turns out there is an $85m tax dispute on this between Uganda and the mobile phone company Zain. This is just about the biggest issue in Ugandan tax right now: the tax inspectors are even tweeting about it.

“Indirect transfers” were highlighted in the recent (and generally solid, I thought) OECD report to the G-20 development working group [pdf].* It says:

Developing countries report that the profit made by the owner of an asset when selling it (for example, the sale of a mineral licence) is often not taxed in the country in which the asset is situated. Artificial structures are being used in some cases to make an ‘indirect transfer’; for example through the sale of the shares in the company that owns the asset rather than the sale of the asset itself.

Unfortunately, it is pretty lame on the solutions. As far as I can tell from the G-20 response [pdf], what is going to happen on it is this:

(deep breath…)

As part of its multi-year action plan, the G-20 development working group will consider calling on the OECD, in consultation with the IMF, to report on whether further analysis is needed.

(…and exhale)

I don’t hear the sound of tax positions unwinding.in response to that one.

To remind you, the big daddy of indirect transfer cases is the Vodafone-India dispute. In that case,  according to this handy summary:

In 2007, Vodafone’s Dutch subsidiary acquired the stock of a Cayman Islands company from a subsidiary of Hutchinson Telecommunications International Ltd. (the subsidiary was also located in the Cayman Islands). The purchase price was $11.1 billion. The Cayman company acquired by Vodafone owned an indirect interest in Hutchinson Essar Ltd. (an Indian company) through several tiers of Mauritius and Indian companies.

Like India, Uganda is trying to tax the sale of a mobile phone company when the transaction took place via offshore holding companies:

Zain International BV owned Zain Africa BV, which had equity in 26 companies all registered in the Netherlands, but effectively owning the telephone operator business in as many African countries. One of them, Celtel Uganda Holding BV, owned 99.99 per cent of the Kampala-registered Celtel Uganda Ltd. On March 30, 2010 Zain International BV sold its shares in Zain Africa BV to Bharti Airtel International BV. As all three companies are registered in the Netherlands, and as the transaction was a sale of shares rather than assets, the company said it did not attract capital gains tax.

The cases are of course not identical. For one thing, Uganda is going after the firm that actually made the capital gain. But the Indian jurisprudence is being used in the Ugandan case.

Just last week, an appeal court ruled that the Uganda Revenue Authority does have the jurisdiction to assess and tax Zain on the gain. Zain will now argue that the transaction was exempt. One of its core arguments is sure to be the Netherlands-Uganda tax treaty.

In common with 86% [pdf] of tax treaties signed by developing countries since 1997, this treaty does not contain the UN model treaty provision that would have allowed Uganda to tax gains on the sale of shares in Ugandan companies made by Dutch residents. It may be that Celtel Uganda counts as a ‘property rich’ company because of all its infrastructure assets, in which case Uganda would have been able to fall back on the OECD and UN model provision permitting it to tax those…except (oops!) even that isn’t included in its treaty with the Netherlands. Yes, this treaty is worse for Uganda than the OECD model, never mind the UN.

So instead we come to Section 88(5) of Uganda’s Income Tax Act [pdf] . This is an anti-treaty shopping provision, which denies the benefits of the treaty to a company whose ‘underlying ownership’ is mostly in a third country:

Where an international agreement provides that income derived from sources in Uganda is exempt from Ugandan tax or is subject to a reduction in the rate of Ugandan tax, the benefit of that exemption or reduction is not available to any person who, for the purposes of the agreement, is a resident of the other contracting state where 50 percent or more of the underlying ownership of that person is held by an individual or individuals who are not
residents of that other Contracting State for the purposes of the agreement.

Sounds like Uganda has it in the bag, right? Unfortunately, this matter will turn on whether Uganda’s domestic law can override its treaty commitments. It is quite likely (certain, if you ask Zain’s tax adviser) that a court will decide it cannot. What everyone I have spoken with agrees on (apart, perhaps, from Zain’s tax adviser) is that it would be preferable to have some certainty about this unresolved question.

The URA has recently begun denying treaty benefits under section 88(5), and until now taxpayers have accepted its reasoning. But, speaking in genera terms at the SEATINI/ActionAid public meeting, a tax official said that the URA doesn’t know if its position will stand up to a court challenge. Tax advisers in the private sector say that, as well as the question of treaty override, the meaning of “underlying ownership” needs to be clarified. Because the Zain case has so far been fought on technicalities, “we were robbed of the opportunity to see how it [Section 88(5)] would work in practice,” one told me.

Perhaps the next stage of the Zain case will answer this question. If it does, it should give some welcome guidance to developing countries struggling with these indirect transfers. If they can’t use their domestic law to override their treaties, they will need to insert an anti-abuse clause into their treaties, strengthen their source taxing rights, or consider cancelling them.

This brings us back to BEPS, and the action on tackling treaty abuse. The OECD is proposing a limitation of benefits clause based on that used by the US, which is similar to that in Uganda’s domestic legislation, only a lot more detailed about who is ruled in and out. This would do the trick, but the challenge would be getting it into treaties that have been already signed.

To solve that, the OECD is pushing a multilateral convention to modify treaties all at once, built on a flexible level of commitment. It concedes [pdf] that the multilateral instrument “has not been identified as high priority by developing countries.” For it to work for them, I think it would need two things:

1. Genuine flexibility so that developing countries can opt into only the bits they want, such as the anti-abuse clause.

2. Willingness on the part of high-risk jurisdictions for treaty shopping (in Uganda’s case the Netherlands, Mauritius, and perhaps now the UK) to opt in to the anti-abuse clause as well.

For Uganda, it might not make sense to wait for this, since we are only talking about two or three treaties. It could ask its partners for a protocol containing a limitation of benefits clause right now. Or, of course, it protect itself and raise more revenue by strengthening all its treaties’ capital gains articles, as the UN model provides for in the first place.

*thanks to @psaintamans for the link!