Following on from two recent blog posts by Paul Sutton and Harold McClure that related to the recent tax case involving Singapore Telecom Australia, we have completed the trilogy in a webinar with Johann Müller.
Johann is a very experienced expert in international tax. In this discussion he looks at the details of the structure that was used, what Singtel's arguments were and why so many of them failed, and what general learning points TP advisers can take from the case.
You can find a recording of the discussion here, and a lightly edited transcript follows.
Paul Sutton: So, welcome everyone to this webinar. The focus of today is discussing the Singapore Telecom Australia decision, both from a transfer pricing perspective and also touching on some of the legal issuesThe focus of today is discussing the Singapore Telecom Australia decision, both from a transfer pricing perspective and also touching on some of the legal issues, which are really interesting.
My name is Paul Sutton. I'm delighted to have our special guest today, Johann Müller, who is a very experienced expert in international tax. Johann, perhaps you can just take a couple of minutes to explain your background and how you came to focus on this area.
Johann Müller: Okay, sure. I always say I'm a little bit from everywhere. I started my tax practice in the early 90s. actually. I've worked first for big consultancy firms (E&Y), and law firms in Amsterdam, then London, then New York, and then finally in Denmark.
Then I went in-house. I worked for Maersk, the world's largest shipping company. I got very interested in tax advocacy there, and did a lot of work at a time when the OECD was busy with the new Article Seven and business restructurings. And that kind of moved me slowly towards transfer pricing.
In 2012, I joined the Danish government – the Danish Competent Authority for transfer pricing. I worked with them until 2014. I was Denmark’s representative on the OECD working policies. Also, during this period BEPS started. So it was very interesting to see all of that from close by. But my heart was too much with the taxpayers. So I went private again.
And then I spent some time doing work in-house again, for the Scandinavian Tobacco Group, which I thought was very interesting, because I've never worked with a multinational, working with national products. Of course, I was also selling cancer, that was not so good. So then I joined Coloplast, which makes ostomy products and stuff like that. A medtech firm. Again, it was a whole new industry for me: I found it fascinating.
And then two years ago, I started my own business, and started doing full time teaching. And doing interim work, because I want to say in practice. But I focus nowadays on international tax and transfer pricing, and basically helping through online courses: training people all over the world, be it in government or in private practice, or people taking the CIOT ADIT exams. So that keeps me off the street.
Paul: Fantastic. Thank you very much. So I'm going bring up your slides. And if it's okay, Johann, I'll ask you just to walk us through the basic facts of the case. It's a complex one, but just to give the overall view on what happened.
Johann: Sure. I try, on a bi-weekly basis or something like that, to upload one international tax case or transfer pricing case onto my YouTube channel and talk about it. So this will be for later this week. And I think it's a very interesting case. It's highly fact-dense and complex, but it deals basically with whether you can change a loan during its journey, and how much you can tweak it before things go wrong.
So the Singapore government indirectly owned the Singtel telecoms company, the largest in Singapore. And in 2002 they acquired the Australian branch of Cable and Wireless, called ‘Cable and Wireless Optus’ at that time. And they acquired it through the structure here.
The first thing that they did is they set up an intermediate holding company in Singapore itself, referred to as SAI or Singapore Australia Investments. And that company was set up to buy out Cable and Wireless in the UK, which had a 52% stake in the Australian company, and the private shareholders as well. And it was a requirement in the bid offer that it had to be a non-Australian entity which would buy it. I'm not quite sure whose requirement this was. But it was required.
And then SAI set up Singapore Telecom AU, or Investments Pty – STAI – a local Australian company. Then formed a consolidated group with Cable and Wireless. The name was changed. And then the idea – as very often in leveraged takeovers – was that the profits from the Australian operating company were used to pay off the debt to SAI in Singapore.
Except there was a bit of a problem, being that the telecom industry is very capital-intensive. And in this case, the operating company's capital investments outstripped the cash flow every year. So there wasn't really cash to pay the interest. So they had to accrue the interest onto the capital, onto the principle, in order to serve the debt.
The debt went through four stages of change. First, it was payable on demand. Then it was changed to a 10-year period. And that was basically fine.
Then what happened was that there was a second amendment in which they said, ‘we're not going to accrue the interest anymore. We'll wait until the operating company has enough cash flow to actually serve the debt, and then we'll start accruing the interest. And in exchange for not accruing the interest in the first few years, we will add a premium of 4.75% to the loan interest.’ Which was the Australian bank swap rate, plus 1% plus 10 divided by 9 (basically, multiplied by a factor of 1.1). But it was a floating rate.
And then the third amendment that came was in 2009. They changed the floating rate to a fixed rate, and said, ‘the financial crisis and everything, we're too unsure about what's going on. So we're going to fix it to a fixed rate now.’ The basic fixed rate was 6.8%, plus 1%, plus the 4.75% multiplied by 1.09. So it came to about 13.8% that they paid in interest.
And the Australian tax authorities said ‘this is too much’. So they made an adjustment, I think for almost a billion Australian dollars, which was non-deductible interest over 2010, ‘11, ‘12 and ‘13. The assessment was made in 2016, so I assume they couldn't go back further.
And the issues that we're discussing in the case were, first of all, the realistic available options for STAI here. Would an independent party have accepted these changes? And basically the authorities said, ‘the original loan was fine. We don't see the need for introducing this interest accrual clause; we suspect (or that's what the judge said) that the interest accrual clause was only there to defer the withholding tax on interest.’ Because as long as you don't accrue, you don't have to pay, I assume.
And then the argument of ‘we had to take the rate from floating to fixed’ at the time when they did it, was not bought at all. Because actually, if you look at the Australian interest developments during the time, the interest rate was going just down, like it did in most of the western world. Because of the crisis it went from a high of 7% in, I think, 2007, to almost 3% in 2009. And the judge didn't buy that argument either.
What neither the Australian Government nor the judge addressed was actually an interesting point. And that is that at the start of this, the Singapore corporate income tax rate was 24.5%. And by 2009, it went down to 18%. And I assume people knew already that was going down to 17%. Whereas the Australian tax rate had been 30% all along. So while initially there wasn't much of a difference to play with, this difference became quite big by the time they made this huge jump in interest rates.
So that was the background. The Australian Government made the adjustments, and different arguments were put up from a transfer pricing point of view about what was going on here. First of all, the big question was: what is comparable? And the taxpayer argued, ‘the comparable would be the credit debt market in the US, and we can show you that, actually, we only paid 177 basis points above what their base rate was. But if we issued these CDMs, it would have been 400 basis points.’ The Government expert disputed that, and said it would have been much lower. So they didn't accept that.
And then there was talk about implicit support. Whether there would have been implicit support from the Singtel parent or not. And the taxpayer argued that it would be very low, because there was no track record of Singtel providing implicit support to companies in trouble. Whereas the government experts said it could have lifted it up by six to eight notches. Just because this was a very strategic investment, both for Singtel itself and indirectly for the Singapore government. Because Singtel itself went from a very large local player to quite a large regional player in Asia Pacific. That was deemed of importance.
And then the second part was, well, if you put implicit support aside, why did Singtel as a parent not give an explicit guarantee? The taxpayer tried to argue that the explicit guarantee would have been way too expensive. Which is actually an interesting point, because you need to agree what is implicit support before you can calculate the explicit support guarantee fee. Because, for instance, if you can borrow at 6% on your own, but through implicit support you can borrow at 3%, and with explicit support you can maybe borrow at 2. Then you should really only charge the guarantee fee (according to the new Chapter 10) over the 1% difference between 3 and 2. Not over the 4% difference between 2 and 6.
But the interesting case is that neither the government nor the taxpayer brought in the new Chapter 10 in these discussions. And at the end, the judge just said, ‘I assume that in a real hypothetical comparable case, there would have been a parent guarantee, because you wouldn't want to pay over AUD5.2 billion for five extra percent in interest to a third party if you could keep those profits for yourself.’ Which is an argument, I think, which makes a lot of sense.
And I think last point that was interesting – and I put some of the financing structures here in for perspective. You can see that Singtel actually gave AUD10.5 billion equity to SAI, but SAI only gave AUD9 billion, so one and a half billion less, in equity to STAI. And you kind of wonder why that was. Was it to get the maximum leverage of what the Australian target could carry? Or was it another reason?
And then the other issue was that Singtel financed part of this acquisition itself with third party debt. And it issued notes for around 3.5%. And if you then look at the external debt that it could get at 3.5% – and SAI had to pay up to 13%. I mean, there's a huge difference there, which could have been wiped out by a parent guarantee. So the judge ended up following the tax authorities’ point of view and readjusted the loan back to its original state. (The first and second amendments, I should also say, having made retroactive effect… the judge also rejected that.)
Paul: Thank you. So, for me, one of the really interesting factors of the case is just the time periods involved. So we have the structure set up in 2002. Shortly after, or immediately after, the acquisition. It was then varied in the same year, to change it from being repayable on demand to the 10-year term. And then it was varied again in 2003, and 2009. And then it appears that the first adverse tax challenge didn't come until 2016. Which is a considerable period after. And one of the things that jumped out at me when I was looking at the judgment is how the court was trying to reconstruct or understand the commercial rationale: why would anyone agree to those changes in that sequence in the way that happened? It was trying to reconstruct that from emails, for example, between the group and its tax advisers.
And clearly, there was an absence of corporate governance-type documentation from a legal entity perspective (which is obviously my perspective on these kinds of cases) to actually record this in a way that made sense. Plus the fact that the original structure, as put in place, clearly didn’t make sense from a legal entity perspective. How could you have STAI as the borrower, presumably with a single asset (actually, its shares in SOPL), how could it possibly agree to a debt being repayable on demand? It just doesn't make sense.
Johann: It doesn't. I do agree. There's a couple of interesting points in what you say there. So the interest payments skyrocketed in ‘09, right? You would have filed your return in ‘10. So maybe the tax authorities picked it up in ‘11, announced an audit around ‘11, which might have taken, let's say, two to three years. So then you get to ‘13, maybe ‘14, you can't get things fixed, and then you go to court. You make the final assessment, and then you end up with the assessment in ’16. I'm not quite sure which year they picked it up, but I can understand that it would have taken time to pick it up. So that's just a comment I want to make there.
Paul: Yeah, makes complete sense. So when you come back to it from a transfer pricing perspective, on the basis of my understanding, the court is looking at two fundamental factors here. One being the credit rating of the borrower, adjusted to reflect the implicit parent company support. And obviously that was a major area of difference between the experts on both sides. And the other was looking at the legal terms of the debts, and the whole question of comparability.
But, based on my understanding, what the court wasn't trying to do – or wasn't tempted to do – was re-characterise the transaction entirely. So it basically accepted it was a debt transaction. So primarily, it's looking at: what are the legal terms of the debt during the evolution of the structure?. And trying to apply the principles of comparability to that ‘as is’, in effect.
Johann: Yeah. There are two points. One was, you also said something about the structure itself. I think it was quite clear that from the beginning they were going to put some kind of debt-carrying vehicle in Australia to finance the acquisition, right? The idea being to take the operating company's profits, and to use that to pay the interest that STAI got. Those kinds of deals are called leveraged takeovers. And so they do want it that way.
I think you can look at that with mixed feelings. Because on one day, you have a taxpayer paying normal taxes. The next day, it gets taken over by a foreign group, and all of a sudden there's no tax base anymore, right? On the other hand, if you were to do an asset deal, and you were to borrow money to acquire the assets, it would be quite normal in profit allocation rules for permanent establishments to allocate that debt to the permanent establishment and deduct the interest from that as well. So from that point of view, I think it is okay, what they've done.
As far as re-characterisation is concerned, a bit of a curveball came out of BEPS. Before that, we talked about re-characterising transactions, and there were strict rules for when you could and couldn't do it. And then BEPS came in and said ‘Oh, hang on, sometimes you only accurately delineate the transaction.’ And in the accurate delineation, you are actually completely throwing out what you don't like. But you're not changing it, you’re just accurately delineating it. And then an even smaller part: they say, ‘if a transaction is economically irrational, then we will re-characterise it.’
So, no, they haven't gone for re-characterisation under the new definition. But they have set aside the second and third amendments. And said, ‘a third party simply would not have accepted these changes.’ I think they've accepted the first amendment: going from repayable on demand to 10 years. But the second and third amendments are quite clearly out of scope. And from that point of view, it's an argument whether they have re-characterised it or whether they have accurately delineated it.
What is quite clear is that the taxpayer didn't have sufficient backup documentation to argue why they wanted to change from simply capitalising the interest to having this deferral of accrual of interest. And why they went for a very high fixed rate in a time when interest rates were dropping. Those were arguments that, one feels, were made up with hindsight, and nothing at the time itself.
Paul: Yeah, absolutely. So if we're looking at this on a forward-looking basis, what are the key learning points from this case that we can take and apply? One of them you've already mentioned, which is record-keeping. I find it hard enough to remember what I've done last week, let alone five years ago or whatever! So that contemporaneous record-keeping. Any other key learning points from your perspective?
Johann: Yeah, if you're talking about record-keeping, it’s also, like, how do you fill it in? I think what would have helped here on the contractual level was, in the contract itself, in the considerations, to have started by explaining what it is. Why they are doing things. And what I've done – and I've seen other people do at other companies – is when you do a transaction that you feel may be challenged, you make a defence file at the time of the transaction.
There you put in email correspondence that would help you. You put in references, you refer to uncertainties and all those things. I mean, one of the problems with, for instance, this thing about the US debt market and why those comparables got thrown out, was because the judge said, ‘you basically took the historic data, and you worked out an argument for it. No-one in 2009 knew what the interest rates were going to do. So we cannot accept your argument.’
And that is very true. I mean, if you can make your papers in 2009, and make your arguments and document them, you will be so much stronger for the future.
Paul: Yeah. Absolutely. And final question in the time available. We touched on it when we had a chat yesterday, Johann: supposing you are parachuted into this situation in this kind of group. And there's an existing structure put in place five years ago, whatever, which on the face of it looks very aggressive. What do you do about it? Is this something that could be fixed?
Johann: I think it is highly fact dependent. There's a couple of different things with fixing things. The first one is: if you change a structure, it draws attention. The second one is: you might be prepared to pay the right tax going forward, but you might not be prepared to pay the right tax going back.
Then you’ve got to think of things like interest and penalties. I mean, in some countries, it's reasonable. Other countries... if you take a country like India, your penalties can be 300% of the amendment. So you really want to know more. And then you want to look at the tax authorities themselves. Are these tax authorities that are business friendly? Or are these tax authorities that are renowned for being extremely aggressive? I mean, some authorities, you just know they go for whatever they can get. While others don't.
And then if you decide to fix things, how would you do it? Are you going to talk to the tax authorities? If they are known as being aggressive, then probably you’re only going to end up settling on your worst case scenario in any case, so why do that? Or are they not?
And to what extent can you trust your local advisors? Are the local advisors giving you advice to talk to the tax authorities, because they think it will come to a fair solution? Or are the local advisors doing this to cover their legal liability issues? That is always a question that you have to ask yourself, in terms of what to do next.
And then sometimes – I mean in this case here, the loan would have expired in three years’ time in any case. If there is a horizon like that, I might simply just sit still and do nothing. Maybe reverse the ridiculous interest rate, and stick to the 4.75% premium or something like that. If they'd done that maybe this thing would have flown under the radar. Try to make changes that won't immediately draw attention. And then sit it out.
But if you are with a business-friendly government, and you feel like you've got a fair chance of going and saying, ‘we've made an honest mistake, we'd like to fix it in the best possible way. Can we do that?’ Then I would go for that.
But part of the reason why is also: skeletons in the cupboard can be terrible. I've seen taxpayers in totally unrelated matters in different countries being completely extorted and blackmailed and everything. Because they had a skeleton in the cupboard that they didn't want the other tax authorities to find out about. You know, if you don't have skeletons in the cupboard, you can take any future battle – totally open and as hard as you need to as well. But you can't do that if you have things that you hope people don't find out about.
Paul: Yeah, absolutely. Well, that's probably a good note to end on. Johann, for people who want to learn more about the background in terms of the movement of interest rates, I think this is covered by your separate video. Can you just remind people where they can find that?
Johann: Sure, I will upload this to my YouTube channel, called Taxpics, tomorrow. And I will also post about it on LinkedIn. And it is quite a complicated case. So I think I’ll end up with three videos of over 15 minutes each, dealing with the different aspects of the case. But then also, there's lots of other cases on that channel, and also links to my other training courses and things like that.
Paul: Fantastic. So that brings us to the end of this conversation, unfortunately. Thank you very much, Johann, I really enjoyed it. Thank you very much to the viewers for watching. And as always, any questions or comments are very welcome. Please do reach out. Thank you very much.
Johann: Thank you so much. Bye.
Get practical advice & insights on the Legal Implementation of Transfer Pricing for Multinational Groups