Singapore Telecom Australia Investments Pty Ltd v Commissioner of Taxation (Special leave decision)

Judges:
Gageler CJ
Gordon J
Steward J

Court appealed from: Federal Court of Australia (Full Court)

Singapore Telecom Australia Investments Pty Ltd
v Commissioner of Taxation

Citation(s):
[2024] FCAFC 29
2024 ATC 20-897
118 ATR 323
302 FCR 192
[2024] HCAtrans 74

Date of decision: 25 October 2024

Result: Application refused with costs

TRANSCRIPT OF PROCEEDINGS

AT MELBOURNE ON FRIDAY, 25 OCTOBER 2024, AT 9.30 AM

Copyright in the High Court of Australia

MR J.W. DE WIJN, KC : If it please the Court, I appear with MR C.J. PEADON and MR L.D. CURRIE for the applicant. (instructed by PricewaterhouseCoopers)

MS C.A. BURNETT, SC : If the Court pleases, I appear with my learned juniors MR M.T. SHERMAN and MS A.A. LYONS for the respondent. (instructed by Australian Government Solicitor)

GAGELER CJ: Thank you, Ms Burnett. Mr de Wijn.

MR DE WIJN: If the Court pleases, can I deal first with the factual findings the Commissioner relies upon at paragraph 7 of his response. They relate to the prediction that SingTel or a parent such as SingTel might have been expected to guarantee a borrowing by its 100 per cent owned subsidiary. That is the first factual finding my learned friends rely upon to suggest this is not an appropriate case for special leave. The primary judge dealt with this point at paragraph 322 and then at 324 and following. I will come back to those shortly.

The Full Court dealt with this point at paragraphs 235 and 245, basically concluding that no error had been disclosed in the primary judge's reasoning. We say that this very factual finding that the Commissioner relies upon in paragraph 7 of the response exposes the error. That is, the whole factual finding relied upon is based upon a non-independent relationship between SingTel and the taxpayer. And we do not shy away from the finding, we rely upon it.

GAGELER CJ: What is the specific finding that you are referring to?

MR DE WIJN: So, the finding that my learned friends rely upon at - - -

GAGELER CJ: You said paragraph 7 of the response.

MR DE WIJN: - - - paragraph 7. The first finding they rely upon is that, at about point 12:

In the present case, STAI's credit rating expert candidly volunteered in his opening remarks that a parent would make a "rational decision" and, where there was a material differential in a subsidiary credit rating, "then their parent will guarantee the obligation" –

It goes on in the response to say:

This was consistent with other evidence establishing that "the logic of the situation strongly points to a parent guarantee being provided" –

So, that is the first factual point my learned friends rely upon to suggest that this is not an appropriate case for special leave, but, of course, that very factual finding turns upon the non-arm's length or the non-independent relationship between the parent and the taxpayer.

At paragraphs 18 and 322 of the primary judge's reasons, his Honour seems to have acknowledged that independence was required both between the taxpayer, who he referred to as STAI, and the intermediate holding company, SAI, who had sold the shares in Optus to the taxpayer, and the ultimate parent, SingTel.

Again, at paragraph 303 of the primary judge's reasons, he makes it clear – we say, correctly – that a relevant condition to be taken into account in looking at Article 6 included the relationship between the taxpayer and the immediate holding company, SAI, and the taxpayer and the ultimate holding company, SingTel. So much is obviously clear, we would say. This is a case, as your Honours will understand, that involves, essentially, Article 6 of the Australia-Singapore Double Tax Agreement, which is incorporated into Division 815-A of the transfer pricing provisions.

STEWARD J: Mr de Wijn, can I ask you a question. If you take out the parental guarantee that informed Justice Moshinsky's conclusion about pricing, what does that mean for the price – or for the interest?

MR DE WIJN: It means we win, because the unchallenged evidence was that there would be – for an unguaranteed loan, the margin would be 400 basis points. On that point, can I take your Honours to what Justice Moshinsky says at paragraph 165. So, at the beginning of paragraph 165, his Honour Justice Moshinsky sets out the Commissioner's contentions and then he sets out a number of paragraphs. Over at paragraph (k) there is a paragraph starting "however". Do your Honours have that?

STEWARD J: Yes.

MR DE WIJN:

However, if the Court concludes that the opinions of Dr Chambers and Mr Chigas on this issue are to be preferred –

Pausing there, that is exactly what happened. There was a dispute about credit rating and the extent of implicit support. The Commissioner's experts had argued that the implicit support would bring the credit rating of STAI right up to, I think, one notch below that of the parent. That evidence was rejected, and Dr Chambers – and Mr Chigas, for that matter, but principally Dr Chambers – had given evidence that STAI's credit rating with implicit support was BBB- and SingTel's credit rating was AA or AA-.

GORDON J: Mr de Wijn, can I ask a question about that. In the joint report – does not the evidence that was given by the joint report replace and in effect answer that question? So, if you go on to paragraph 290 of the primary judge's judgment, I had understood the experts agreed that they would have used a guarantee and it would have been put in place without a fee. Is that right?

MR DE WIJN: No. Perhaps I will deal with your Honour's point, then go back to the paragraph I was reading. At 290 in the joint report – this is 290 of the judgment – paragraphs of the joint report are set out, and 3.14 says:

STAI –

that is, the taxpayer:

could have utilized a SingTel guarantee, but such financing would have been at a higher all-in cost to STAI after payment of a guarantee fee had been made to SingTel on an arm's length basis.

STEWARD J: Did anybody challenge any of that evidence?

MR DE WIJN: No, and your Honours will see – I am just going in a different order, but I am happy to deal with that – at - - -

GORDON J: And then how does that deal then, just so I understand it, just to complete it – paragraphs 3.13 and 3.14 and 3.15 of that joint report?

MR DE WIJN: Sorry, your Honour?

GORDON J: I am just trying to work out what the state of the ultimate position of the joint experts was.

MR DE WIJN: Well, the state of the joint experts was exactly as set out in the joint report: they agreed that if there had been a guarantee, with a guarantee fee, my client would have paid a higher interest rate, or a higher borrowing cost. Now, the evidence in respect of what might have happened without a guarantee is set out at subparagraphs 263(c) and (d). That evidence was not challenged and was not rejected by the primary judge. Your Honour will see in paragraph (c):

(c)

A DCM transaction with a parent guarantee from SingTel, with a guarantee fee payable by STAI, would have had a total credit spread of 300 to 360 bps.

(d)

A DCM transaction with no parent guarantee would have had a credit spread of 400 bps.

Now, that was the unchallenged evidence of Mr Chigas, and in fact the Commissioner and his Honour relied upon that. During cross-examination at page 585, Mr Richmond in fact adopted Mr Chigas' evidence and said, well, is it not obvious that if there was a 300 basis point difference – that is, without a guarantee that the interest is 400 basis points above the floating rate, with a guarantee it is 100 – and Mr Richmond and then his Honour relied upon that enormous difference to suggest that it was obvious that a parent company would give a guarantee.

GORDON J: Is that reflected in 328 of his Honour's reasons, on page 113? That is where his Honour deals with Mr Chigas' evidence.

MR DE WIJN: Sorry, your Honour, paragraph 328?

GORDON J: On page 113. This is where his Honour seems to reject Mr Chigas' evidence on the basis that it is not certain enough – it is "speculative", he says.

MR DE WIJN: This is on the matter of the guarantee fee. His Honour did say that. But Mr Chigas was not challenged on the question of the amount of the guarantee - - -

STEWARD J: I was going to ask you, did anyone put that to Mr Chigas in cross-examination?

MR DE WIJN: No. He was not challenged at all about the guarantee fee, and, in fact, it was common ground between the experts – as your Honours notice from paragraph 290 – that with an arm's length guarantee fee, the taxpayer would have paid more. There was no challenge to that.

GAGELER CJ: Was this a point made to the Full Court?

MR DE WIJN: It was.

GAGELER CJ: The challenge to the conclusion in paragraph 328?

MR DE WIJN: Yes.

STEWARD J: Do I take it, Mr de Wijn, that you would say that the language of "could have been" in Mr Chigas' report is reflective of the fact that he was trying to price a hypothetical?

MR DE WIJN: Exactly. What Mr Chigas did was he gave detailed evidence about guarantee fees that were payable in the market. He explained that it was a range of between 50 and 70 per cent of the differential. Then, following his reports, there is a report by Mr Johnson – so, Mr Johnson and his report does not challenge any of that, and I want to go back to how this case was run – Mr Johnson did not challenge any of that. There was then a joint experts' conclave, where the experts agreed specifically that with a guarantee fee on an arm's length basis, the taxpayer would have paid more.

That was not just overlooked, because Mr Johnson came along just before the trial and amended that by saying, if such a guarantee was required by Australian law, or words to that effect. So, the whole case proceeded, as his Honour says in paragraph 165(k), that if the evidence of Dr Chambers was accepted on the credit rating – that is, that STAI had a credit rating of BBB-, that the case could only succeed if there was a guarantee by SingTel.

STEWARD J: Can I ask you two questions, Mr de Wijn. Firstly, do you need all three grounds to be successful?

MR DE WIJN: No. All we need to be successful – if it was improper to impute a guarantee fee, we win, because there is evidence about what the interest rate would be without a guarantee, it is 400 basis points. No challenge to that – in fact, relied upon by the Commissioner, because the Commissioner wanted to say – and relied upon by the judge to say – with such a difference, it is obvious that a guarantee is going to be provided by the parent – not by an independent company. That is where the error is.

What Dr Chambers said, it is likely that a parent would provide a guarantee to its 100 per cent owned subsidiary if it was going to save 300 basis points. But the test for Article 6 and for Division 815 is what might happen – what might be expected to happen – between independent parties. The error that his Honour fell into, and the Full Court fell into, was failing to look at the independence between SingTel and the taxpayer. That is something – I mean, it may be perfectly logical for a parent like SingTel to provide a guarantee in circumstances where it is going to save the group and itself, ultimately, 300 basis points on a very large loan.

GAGELER CJ: Am I understanding that you are saying that ground 1(a) is sufficient?

MR DE WIJN: Ground 1(a) is sufficient. And 1(a) is sufficient because you simply would not get to the stage of imputing a guarantee. If you did impute a guarantee, that guarantee would still have to be imputed on an arm's length basis.

GORDON J: Could you put that proposition again, please, Mr de Wijn? I did not quite hear you.

MR DE WIJN: So, we say 1(a) is sufficient, that it was inappropriate to impute a guarantee on a non-arm's length basis. If a guarantee was to be imputed – so, without a guarantee, we win: 400 basis points, as opposed to 100 basis points. There is an Excel spreadsheet that is prepared as part of the judgment. All you need to do – the Commissioner just needs to put in a different interest rate. It is just a matter of mathematical calculation.

STEWARD J: Can I ask another, related question. There was some debate below about the first sentence of paragraph 180 of Glencore. If this case were to be given special leave, do we need to worry about the halo effect, or is that something that is accepted?

MR DE WIJN: No, we do not need to worry about the halo effect because, in fact – that may be a case for another day – Dr Chambers, in coming to his credit rating of BBB-, allowed for a halo effect of some amount, but not as much as Mr Weiss allowed. So, Mr Weiss said the halo effect or the implied support would be enough to get it up to the credit rating of SingTel; Dr Chambers introduced a halo effect, but not as significant. So, your Honours do not need to worry about that.

We say the first part of paragraph 180 in Glencore, is exactly the mistake that his Honour and the Full Court made, because they took into account, in imputing a guarantee to a wholly-owned subsidiary, a relationship which would only apply in a related party relationship. A company like SingTel is not in the business of providing guarantees, but it is not going to provide a guarantee to a company if it is dealing wholly independently with it. And that is the simple mistake that both his Honour Justice Moshinsky made and the Full Court made.

STEWARD J: Can I ask you just one more point of clarification. Your third ground was the rejection of the debt capital markets pricing. Do you nonetheless still need that in order to succeed?

MR DE WIJN: We do not, because it is obvious from the evidence and the way the case proceeded – and your Honours will see that at paragraph 165(k), where the Commissioner's case is put on the basis that if Dr Chambers' evidence is accepted, the Commissioner will only win if the guarantee is imputed. That is the gist of what his Honour says at paragraph 165(k).

Now, as far as your Honour Justice Steward's comment about the DCM, I do not want to leave that totally unanswered, because the agreed evidence between the experts – and your Honours will see this at paragraph 290 – was that the DCM would have been the most likely and most economical, if not the only source of finance for this transaction. Remember, this was a transaction where SAI, the intermediate holding company, originally owned the shares in Optus. That company sold the shares to my client for an issue of shares and debt. Now, that is a wholly non-arm's length transaction, or internal transaction. The pricing was exactly the same pricing as for the actual takeover, so there cannot be any argument about pricing.

So, that whole transaction of equity – 100 per cent equity and debt – was by its very nature a non-arm's length, non-independent transaction. So, the experts – and your Honours will see this at paragraph 290, at paragraph 3.10 from the joint experts' report, the evidence was, and the case proceeded, that:

STAI's most economic and likely only source for the quantum of $5.2 billion of third-party, long-term funds would have been a US DCM new issue –

together with a currency swap. So, in circumstances where the experts have agreed that that is the most economic and probably only source of independent finance, with respect, it is, I should say, nonsense to suggest you have to build into it some non-arm's length vendor financing in the light of that evidence.

So, that sort of falls apart, we say, based upon the joint experts' report. A taxpayer does not have the burden of proving something that is not in issue in a case. We say it has been seven years since the decision in Chevron. It was a controversial decision, it seems that Australia is out of step with the international community, and this is an ideal case for special leave because, based upon the credit rating evidence, the case could only succeed if there was implied a guarantee and a guarantee without a guarantee fee.

GORDON J: But you say you do not need 1(b), which deals with the guarantee fee. You say all you need is 1(a).

MR DE WIJN: We probably only need one, but if, for some reason, one implies a guarantee, you cannot imply a guarantee without a guarantee fee. It is interesting, looking at the evidence from the overseas people –

Dr Collier, for example, said the imputation of a guarantee by a wholly owned parent to a subsidiary is fundamentally at odds with the arm's length principle. It just does not happen anywhere else in the world.

GORDON J: But that seems to me to be raising the question Justice Steward just asked you about the intersection between 179 and 180 of Glencore. Is that not dealing with that issue?

MR DE WIJN: We say Glencore supports us, because the very point of paragraph 180 of Glencore is that you take the circumstances of the taxpayer as you find them, but you ignore conditions and circumstances that are attributable to the parent-subsidiary relationship, the non-arm's length relationship.

So, you do not pretend that STAI is a holding company of a mining operation or something like that. You obviously look at STAI as being involved in the telecommunications interest and being a holding company with commercial objectives that it might have. But when it comes to imputing a non-arm's length guarantee, that is exactly – I did not write paragraph 180 of Glencore, but we say that that is exactly what paragraph 180 of Glencore says: you ignore factors between the parent and the subsidiary that are attributable to the non-arm's length relationship – or, I should say, the non-independent relationship, to use the words of Article 6. That is the fundamental error that has been made.

GAGELER CJ: Thank you, Mr de Wijn.

MR DE WIJN: If your Honour pleases.

GAGELER CJ: Ms Burnett. I think, stripped down, the argument against you on special leave is that the case could proceed on ground 1(a) alone.

MS BURNETT: Your Honour, we certainly do not accept that. We think the applicant needs cumulative success on a number of grounds here – at least grounds 1 and 3, and we say ground 2 as well, given the forensic decisions it has made about how to calculate the deductions it said would be available. This matter, for all those reasons, is no vehicle for issues of principle because the applicant failed on the facts at a number of levels below, not on points of law.

In fact, the way the applicant chose to run its case and the many factual findings against it show a number of evidentiary obstacles that I will describe. Indeed, at the end of the Full Court's reasons at paragraph 309 – that is on page 88 of the High Court copy – their Honours held that if the basis on which they decided the case – which was also the basis on which the primary judge decided the case – was not accepted, then, in relation to the notice of contention:

there would have been much to commend the conclusion that STAI did not discharge its onus because of the –

applicant's forensic choices and various problems with its evidence. Now, I will address your Honours on why this is not a suitable vehicle for the proper identification of the arm's length hypothetical in Division 13 or 815-A, in particular, including whether that should include, in this case, a parent guarantee or a guarantee fee. Firstly, on the parent guarantee point, the applicant would simply be unable to overcome the findings about its evidence as to the relevant hypothetical without a guarantee. This would be before any of the legal issues canvassed by my learned friend this morning would arise.

STEWARD J: Ms Burnett, can I ask you a question. Is there a tension in your argument between, on the one hand, rejecting pricing based on debt capital markets, because that is too far removed from the actual transaction and on the other hand saying that, nonetheless, we should impute a parental guarantee in this case when one did not in fact exist?

MS BURNETT: Your Honour, it is not the case that bond market pricing can never be used in a transfer pricing case about another type of financing. It is undoubtedly - - -

STEWARD J: But it was rejected here on the grounds that it was too far removed from reality.

MS BURNETT: That is right, it is far removed.

STEWARD J: But also, the guarantee is also removed from reality because none was actually provided.

MS BURNETT: But there was a great deal of evidence, lay and expert, for the proposition that it was very compelling to expect a guarantee, in terms of reasonable expectation. The bond market evidence – the problem with that was there was no attempt by Mr Chigas to account for the significant differences between a bond financing, which lead to his pricing, and vendor financing or financial accommodation. The courts below – and in transfer pricing, of course, reasonable adjustments can be made, none - - -

STEWARD J: But did not both experts agree that it was an appropriate proxy?

MS BURNETT: I do not think both experts used that term, but both of the debt capital markets experts were using bond market pricing.

STEWARD J: It says at 3.10:

STAI's most economic and likely only source for the quantum of $5.2 billion ... would have been a US DCM new issue transaction coupled with a CCIRS.

MS BURNETT: That is a joint opinion, but it - - -

STEWARD J: From your expert and Mr de Wijn's expert.

MS BURNETT: Yes. That was not the only source of evidence in the matter. There was, of course, all of the contemporaneous documentary material, which included both emails from the Optus finance officers and a SingTel financial controller, and opinions from Ernst & Young as to the interest rate that was actually used in the original form of the LNIA agreement. But we do not say that bond market pricing is never going to be relevant. In transfer pricing, a degree of flexibility and pragmatism is obviously required, and you do not need a perfect match with a comparable, but you do need to bridge the gap, and there was just no bridging to account for - - -

STEWARD J: Can I ask you, related to that, paragraph 3.14 in Mr Johnson's addendum to the joint report – what did he mean when he said:

if such a fee had been required under Australian transfer pricing regulations –

MS BURNETT: That is referring - - -

STEWARD J: Was he questioned about that? I mean, do we know what he meant?

MS BURNETT: It comes before an agreed part of the sentence of:

on an arm's length basis.

And so, the experts there were assuming that Australian transfer pricing law would have required a guarantee fee, and that is not - - -

STEWARD J: Would it not have been Singapore transfer pricing regulations?

MS BURNETT: That is what we say, your Honour, and that is part of a problem for the applicant on ground 2(a), that there was no evidence that Singapore transfer pricing law required an arm's length guarantee fee to be returned in Singapore, and in fact one would infer that that was not the case, because SingTel had in fact guaranteed Optus Finance, another Australian subsidiary, very shortly before the start of the LNIA in question, and no fee was charged on that transaction itself.

GAGELER CJ: Ms Burnett, it would be of assistance if you were to concentrate on the evidentiary problems if we were to consider ground 1(a).

MS BURNETT: Yes, your Honour. I should make it clear that my learned friend answered a question from your Honour Justice Steward this morning and he referred in that regard to paragraph 165(k) of the primary judge's reasons, and used that for the first time – as far I am aware, it was never put to the Full Court – that it was somehow common ground that there would have been a margin of 400 basis points on a unguaranteed debt if Mr Chambers' evidence was accepted.

That does not emerge from that paragraph, and it was clearly never common ground. If that was the case, it would have been a much shorter case and a much shorter appeal. Indeed, the issues with Mr Chigas' evidence go very deep. Mr Chigas gave evidence of a bond market scenario, as your Honours know, and he opined that on his bond scenario, the applicant would have had to pay a total of $7.7 billion worth of interest over the 10 years. This is on a principal of $5.2 billion. That would have been an extra $2.8 billion more than the actual $4.9 billion of interest that the applicant paid to SAI in total. The Full Court found at paragraph 161 that there was a "very substantial disconnect" between Mr Chigas' evidence and reality, referring also to findings of the primary judge.

Mr Chigas at trial was shown the applicant's profit and loss data for the 10-year period, and he was shown how an extra $2.8 billion of interest coming out over that period would have wiped out all of the Optus Group's profits – the net profits of the Optus Group – and kept the group in losses for the whole 10 years.

STEWARD J: Ms Burnett, can I ask you, on that deep disconnect, do you have an answer to what is said at paragraph 19(b) of the reply?

MS BURNETT: Paragraph 19(b) of the reply refers to, I think, the fixed interest rate point, does it, your Honour?

STEWARD J: It says:

The reference to the "very substantial disconnect" ... is itself unconnected with this aspect of Mr Chigas' evidence. That comment was made in respect of an opinion he expressed –

et cetera.

MS BURNETT: Yes, I have that. Thank you, your Honour. The applicant seems to be there accepting the "very substantial disconnect" point. Indeed, that "very substantial disconnect" comment went to the heart of Mr Chigas' evidence. The applicant here is trying to say it does not relate to the fixing of the interest point. That is one component - - -

STEWARD J: That is what I wanted to ask you about. What do you say about that?

MS BURNETT: Well, the fixing interest point is actually fundamental to the applicant's case. It was fundamental in Mr Chigas' evidence. It is not challenged in the grounds and the application, and it means, in and of itself, that the applicant cannot succeed. So, Mr Chigas' scenario – and this was his only scenario for which there were calculations of the interest that would be payable – involved a floating interest rate for the first few years, and then a fixed interest rate from 1 April 2009, set in October 2008.

This was at a time where interest rates were falling and the Optus Group was actually looking to increase its third party floating rate borrowings. The primary judge and Full Court analysed in great detail the lay and expert evidence about how reasonable it was to expect that an entity like this in a group like this would fix its rate from that time – this is in paragraphs 199 to 234 of the Full Court's reasons, referring back also to the primary judge. None of this is challenged.

As your Honour Justice Steward points out, the applicant's only contention in this regard – and that is at 19(a) of its reply – is that at trial, the Commissioner accepted that if a fixed rate were to be used from that date, then a rate of about 5.61 per cent had been accepted.

STEWARD J: Can I just ask you to clarify this – I know you did not do the trial, Ms Burnett, so you may not be able to answer. Paragraph 263(d) of the trial judge's reasons summarises Mr Chigas' evidence that:

A DCM transaction with no parent guarantee would have had a credit spread of 400 bps.

Was that challenged?

MS BURNETT: It was challenged, your Honour. Many of the components that made that up were challenged, including this – the fixed interest rate was an important element of that.

STEWARD J: But was the main challenge the proposition that you cannot use a debt capital market as a proper proxy?

MS BURNETT: I would say there are five problems with it, which were all challenged. One was the very substantial - - -

GORDON J: Could you just list them for us, so I have an overview of where you are?

MS BURNETT: Yes, your Honour.

GORDON J: Thank you.

MS BURNETT: So, the first was the "very substantial disconnect" and the fact that Mr Chigas, when he was shown the net profit result, said he had not actually analysed net profits and where that extra 2.8 billion could come from. The second was that Mr Chigas' evidence was based on the United States bond market and he made no adjustments to account for the significant differences with the vendor financing – the fact that there is one creditor as opposed to numerous dispersed bond holders, for example, others may have been the fact that bonds, by their nature, are covenant-light, whereas a vendor financing may well have had covenants bringing the rate down. The third point is the fixed rate that was baked into Mr Chigas' calculations. The fourth point was - - -

STEWARD J: What is wrong with that?

MS BURNETT: It is that it would not have been reasonable to expect that a debtor in the position of STAI would have done – at the time of the third amendment, when it in fact, in the actual, non-arm's length world, entered into the third amendment, the effect of which was fixing the interest rate, there were firm findings in a great number of paragraphs that that would not have been reasonable to expect, no matter what one calls the fixed interest rate, whether it is 5.61 per cent or the 6.8 per cent in the actual, that all of the economic factors pointed compellingly the other way.

STEWARD J: So, you are up to number four.

MS BURNETT: Yes. The fourth is that Mr Chigas used hindsight to put his model together and to analogise it to the actual LNIA. He assumed that the second amendment was always on foot, whereas for the first period of the LNIA it was not. Related to this, he ignored the fact that $286 million of accrued interest effectively disappeared with the second amendment. He also used hindsight from later in the term to assume that interest was always deferred but never contingent, but in fact, it was contingent for a significant period of this.

GAGELER CJ: So, on an appeal, this is the level of the argument with which we would be dealing. Is that right?

MS BURNETT: This would all need - - -

GAGELER CJ: It is about inferences to be drawn from his evidence in the context of the totality of the evidence before the primary judge?

MS BURNETT: That is right, your Honour. These are the hurdles that the applicant would need to overcome. The appeal before the Full Court went for a full four days, and that was run at a very fast clip. Most of that would need to be - - -

GAGELER CJ: And these points you are addressing by reference to the evidence of Mr Chigas, neither side seems to be able to point as to actual findings that will address these points.

MS BURNETT: Well, your Honour, all the points I have made, I can point to the numerous paragraphs below of the findings that support our propositions. We do not challenge any of those. So, for example, in relation to the hindsight point that I have just referred to, that is in the Full Court at paragraphs 43, 107 and 148, and the primary judge at 235, 236, 332, 343 and so on. The great detail of these judgments needs to be overcome and disputed by the applicant.

GORDON J: Can I ask whether you have dealt with your fifth point?

MS BURNETT: The fifth point is the semi-annual capitalisation of interest that Mr Chigas also baked into his calculations, whereas – and this served to drive up the interest cost relative to an annual capitalisation. The evidence of Mr Johnson, the Commissioner's expert on this matter, as the Full Court found at 266 to 267, was that the 12-month BBSW base rate, which was used in the actual and also used in the hypotheticals, that was an annual rate, 12-month rate, and so it was appropriate to capitalise annually, not every six months. This point is also unchallenged in the application for special leave.

GORDON J: So, that is an aspect of the hypothetical that was adopted.

MS BURNETT: It is. All of those are aspects of the hypotheticals. The applicant chose to put all of its eggs in that basket, of that model of Mr Chigas. These all mean, we say, that the applicant would not be able to show that the assessments were excessive, regardless of what may be found about a parent guarantee. Indeed, it is important to know that this case is not about whether a subsidiary under transfer pricing should be treated as a standalone. As my learned friend accepted, the halo effect, so to speak, was common ground in this case, that an independent party would take into account, and it is relevant to the hypothetical that the applicant is a member of a multinational group like the SingTel group with a parent like SingTel.

STEWARD J: Can I ask you a principled question, if I can call it that. For the purposes of the Australia-Singapore Double Tax Treaty, why is the provision of a service – namely, a guarantee, for free – a condition that one would not expect to see operative between parties dealing independently with each other, as a matter of principle?

MS BURNETT: Your Honour, you said "as a matter of principle", we say one never gets there on the facts of this case - - -

STEWARD J: Yes, do not worry about the facts of this case, just - - -

MS BURNETT: Yes, but as a matter of principle, your Honour, the transaction here is a financial accommodation.

STEWARD J: Do not worry about the transaction here. Why would it not be the case that the provision of a valuable service, for the purposes of the Australia-Singapore Double Tax Treaty Article 6, for free, would be a condition that one would not expect to operate between parties dealing independently with each other? What is your answer to that?

MS BURNETT: Our answer, your Honour, is that there is not, in fact, a guarantee here. This is a financial accommodation. If the transaction was a guarantee and the applicant was trying to deduct guarantee fees, then Article 6 would clearly apply to that. Whereas here, the transaction is about the financing transaction, and that is the relevant commercial and financial relations. What might hypothetically, in an arm's length world, take place with a third party, being the parent, is not engaged by the Article.

STEWARD J: I thought Justice Moshinsky had said, in reaching his price in conclusion, looking at paragraph 18(b), that there would be a parent guarantee from a company like SingTel.

MS BURNETT: Yes, that is right. But, your Honour, that is a guarantee in the capacity as parent, and so - - -

STEWARD J: Is that not a condition that you would not expect to see operative between independent parties?

MS BURNETT: But the independent parties here are the applicant and SAI, the creditor. One does not have to assume – and this emerges from Glencore and Chevron – that STAI is independent of everybody, that it is independent of its parentage. One has to assume it is independent of the counterparty to the tested transaction.

STEWARD J: Is that not an important point that might need to be resolved, taking us back to the first sentence of paragraph 180 of Glencore?

MS BURNETT: We do observe that paragraph 179 confirms that the membership within a multinational group is still a relevant consideration in the arm's length hypothetical. In the abstract, that might be a question that may arise, but it does not arise for determination in this case. Further cases that may come through may explore it in some more detail, but here what is relevant is that the parent guarantee was found by the primary judge to be given in the capacity as parent.

Why would a parent charge its subsidiary for that? It would only be if Singaporean law required an arm's length fee to be returned in Singapore, and therefore charged in Australia, but there was no evidence of that, and the evidence of the non-charging of a guarantee fee for another guarantee was just quite contrary to it. Also, if your Honours would be assisted in relation to the guarantee fee point in 1(a), there are also substantial hurdles or obstacles in the applicant's way there from the paucity of evidence about a guarantee fee.

GAGELER CJ: We have noted what the primary judge said at paragraph 328. You rely on that?

MS BURNETT: Yes, that is right. And it is not merely because it was a hypothetical inquiry. All transfer pricing exercises are hypothetical, but they do require a less speculative and more applied analysis than what is found in the small amount of evidence. In fact, it is impossible to calculate what the deduction would be for a guarantee fee on the basis of Mr Chigas' evidence. It is a very wide range given.

There is no accounting for the actual characteristics of these particular parties that would clothe the arm's length hypothetical, and there is no analysis of: is the guarantee fee applied to the principal only or the principal and the capitalised interest? What would happen in the early years where no interest was in fact paid? Would there still be a guarantee fee paid in those years? What would be the interval for paying of the guarantee fee?

This is because the applicant put – at trial, it focused hardly at all on a guaranteed scenario. The unguaranteed scenario – the 400 basis point calculation – was the only scenario for which the applicant had mathematical calculations of the amount of interest it would pay and the amount of deduction over the 10 years in total, or year by year. It is actually impossible to calculate the guarantee fee scenario, and then there is also the problem of what the guaranteed interest rate is to which the guarantee fee would be added.

There, this Court would be called on to adjudicate between the three different sets of evidence, the contemporaneous documentary material relied on below, Mr Chigas' model as to a guaranteed interest rate, and then, if necessary – we think it would not come to that – Mr Johnson's model as to a guaranteed interest rate as well. So, there would be a significant factual hurdle, and for most of those scenarios there is not a pathway to success for the applicant.

Lastly, in relation to the joint debt capital market's expert report, to which reference was made this morning, that does not say that there would have been a higher all-in overall cost – as my learned friend said, the word "overall" is not there. What the experts were referring to is that a guarantee would require a guarantee fee, which would increase the cost compared to not having a fee, but only if, on the part of Mr Johnson, Australian transfer pricing regulations required that. So, for all these reasons, your Honour, we say the applicant would not be able to show that the assessments were excessive, regardless of any determinations on points of principle.

GAGELER CJ: Thank you very much, Ms Burnett. Mr de Wijn, do you have a reply?

MR DE WIJN: Your Honour, my learned friend referred to a whole lot of evidentiary matters, none of which, with respect, are relevant to the special leave point. For the purposes of special leave, we accept the no amendment model that his Honour relied upon, and the only variable that needs to be put into that is the interest rate. So, all of these questions about fixing interest rates or not fixing interest rates are all irrelevant at this stage, because his Honour the primary judge accepted the no amendment model, which was exhibited to his judgment, and that involved borrowing at day 1 with interest accruing and being capitalised where appropriate. So, the only variable is a mathematical variable of plugging in an interest rate.

Now, my learned friend referred to five or six reasons why Mr Chigas' evidence should not be accepted. Not one of those reasons related to a challenge to the 400 basis points. There were challenges to the calculations he did. There were challenges to the fixed rate and things like that. But nowhere in the evidence and nowhere at trial was there a challenge to Mr Chigas' evidence that without a guarantee fee, the credit spread would have been 400 basis points.

That is logical when one looks at the admitted fact that the overall cost would have been higher with a guarantee fee. The guarantee fee is never going to be more than the differential between a guaranteed and an unguaranteed amount, so once you have a factual agreement between the experts that with a guarantee fee the overall cost to the taxpayer is going to be more – and that is clearly the way you should read it, my learned friend's suggested reading is just inappropriate - - -

GAGELER CJ: This is a reading of the evidence we are being asked to act upon?

MR DE WIJN: Sorry?

GAGELER CJ: You said "a reading". This is a reading of the evidence?

MR DE WIJN: No – well, it is not a reading of the evidence, it is just looking at the agreed fact. I mean, why would anyone cross-examine – there was no cross-examination on the guarantee fee. Why would anyone do that when there was an agreed evidence between the experts that with a guarantee fee, there is an additional amount? So, we start with the point that without a guarantee, the spread is 400 basis points. No challenge to that evidence. My learned friend did not point to one place in the judgment or in the evidence where it was challenged or where his Honour rejected that finding of 400 basis points. In fact, at paragraph 363(d) he refers to it without rejection, and as I have said, effectively relies on it to show the logic of providing a parent guarantee.

It is the logic of providing the parent guarantee to a wholly owned subsidiary that turns this case. That is the basis on which his Honour decided it. With respect, that is the error. This is an ideal case, we say for the grant of special leave. It is not a case where Singapore law was relevant. The question is, what would independent parties do, not a question of whether Singapore law required a guarantee fee. The fact that SingTel did not charge a guarantee fee to its subsidiaries again demonstrates the error, because that is a non-independent relationship.

Why would a parent ask for a guarantee fee from a 100 per cent owned subsidiary? It is just moving from one pocket to the other. The test under Article 6 as introduced into Division 815-A is: what would happen between independent parties dealing wholly independently with each other? We say this is an important issue. Chevron was decided seven years ago. We appear to be out of kilter with the rest of the world on this issue – the 1(a) issue about whether a guarantee can be imputed – and we say it is an important matter not only for Australian taxpayers but for our role in international community.

If the Court pleases.

GAGELER CJ: Thank you, Mr de Wijn. We will retire momentarily to consider the course we will take.

AT 10.20 AM SHORT ADJOURNMENT

UPON RESUMING AT 10.26 AM :

GAGELER CJ: Having regard to the state of the evidence before the primary judge, we are not persuaded that the appeal would present as a suitable vehicle for this Court to examine the issue of principle sought to be raised by proposed ground 1. Otherwise, we consider that the proposed appeal would have insufficient prospects to warrant a grant of special leave to appeal. Special leave to appeal is refused with costs.

The Court will now adjourn until 10.00 am on Tuesday, 5 November.

AT 10.27 AM THE MATTER WAS CONCLUDED