The Federal Court's recent decision in Chevron Australia Holdings Pty Ltd v Commissioner of Taxation (No 4)  FCA 1092 is an important decision in relation to Australia's transfer pricing law and the application of the arm's length principle to financing issues (such as risk margins and choice of currency). In particular, the decision highlights the importance of having relevant and high-quality expert evidence in transfer pricing disputes.
Background to the issues
This case has been closely followed, as it was the first case in Australia which applied transfer pricing principles in a financing context (and only the third significant Australian transfer pricing case). The financing issues have attracted particular attention, as limited judicial guidance on this issue exists globally (such as the Canadian decisions in The Queen v. General Electric Capital Canada Inc. 2010 FCA 344, which had accepted that passive affiliation could be taken into account in a transfer pricing analysis).
It was also the first case to consider the validity of the transfer pricing rules in Subdivision 815-A of the Income Tax Assessment Act 1997 (Cth). Subdivision 815-A was enacted in 2013 with retrospective effect to the income years starting from 1 July 2004 to 30 June 2013. These rules operated concurrently with existing transfer pricing rules in Division 13 of the Income Tax Assessment Act 1936 (Cth).
From 1 July 2013, both Division 13 and Subdivision 815-A were replaced with Subdivision 815-B of the Income Tax Assessment Act 1997 (Cth).
Background to the matter
The relevant borrower was Chevron Australia Holdings Pty Ltd (Borrower), a wholly owned subsidiary of the US-based Chevron Corporation (Parent).
The Borrower entered into a Credit Facility Agreement in June 2003, under which it borrowed from a special purpose subsidiary (Lender) which had been established to raise funds in US bond markets (with the benefit of a guarantee from the Parent). The key features of this credit facility were that:
- the Borrower borrowed the AUD equivalent of US $2.5 billion;
- interest would be paid at 1-month AUD LIBOR + 4.14% p.a.;
- the loan was repayable in full after 5 years, but could be repaid earlier at the Borrower's option;
- the loan was not subject to financial covenants; and
- no guarantee or security was provided by the Borrower.
In summary, the Commissioner made determinations under both Division 13 and Subdivision 815-A, based on his view that:
- the arm's length rate of interest for the loan was lower than that charged by the Lender; and
- an arm's length loan agreement would have contained other terms which reduced the amount payable by the Borrower (such as that the loan would have been in USD and would have contained financial covenants).
The decision turned on the evidence before the Court. The Court made a number of findings on legal and factual issues that are likely to impact upon future transfer pricing matters, including that:
- Article 9 of the Australia-United States double tax agreement did not operate "independently of the transfer pricing provisions in the domestic legislation" as a basis for a tax assessment;
- the concept of "consideration" in Division 13 was not limited to the interest rate, and included valuable promises of the borrower (such as restrictive covenants and security);
- Division 13 does not treat a taxpayer which is a subsidiary of entity as a stand-alone entity;
- independent lenders do not rely upon published credit ratings, and instead complete their own credit analysis (as such, the practices and policies of rating agencies are not relevant);
- although it was permissible to take the implicit support of a parent entity into account, it "had very little, if any, impact on pricing by a lender in the real world";
- an arm's length loan may have been made in AUD for commercial reasons, despite carrying a higher interest rate than a loan in USD;
- the transfer pricing rules can be applied irrespective of whether the amount of debt is below the safe harbour thresholds under the thin capitalisation rules; and
- Subdivision 815-A was constitutionally valid.
The role of expert evidence
Given the nature of the issues which often arise in transfer pricing matters, the Court was heavily reliant on expert evidence to make the factual findings necessary to determine this matter.
From a practical perspective, experts are selected by the party who wishes to rely upon their evidence. However, the expert is not an advocate for that party, and their paramount duty is to assist the Court to make factual findings which are within their area of expertise.
The selected experts are then briefed with the relevant materials and asked to respond to the questions by the party's legal team. In this case, the Court found that many of the questions that had been asked did not address what the Court interpreted to be the relevant test.
The pricing experts
The Court was not persuaded by the evidence of the Borrower's two expert witnesses who were called to give evidence as to the price that would have been paid by the Borrower for an arm's length loan. Therefore, Justice Robertson found that the Borrower had not discharged its onus of proving that the assessments which the Commissioner had issued were excessive.
The Borrower's two witnesses on this point were former senior officers of major US banks. In considering their evidence, the Court accepted the evidence of each that independent lenders performed their own independent credit analysis (and did not rely on the published reports of rating agencies). Accordingly, the Court did not consider the practices of credit rating agencies to be relevant.
However, as discussed above, the Court did not accept the evidence of either witness as to the arm's length consideration of the loan. This was because the Borrower's first witness:
- "did not address what the consideration would have been if it had been negotiated at arm's length" (which he was not asked to do);
- "had no relevant experience in or detailed knowledge of the" exploration and production industry;
- disregarded credit metrics relevant to this industry;
- assumed there could be multiple lenders (when the transaction undertaken involved only one lender); and
- based opinions on a "form of transaction (that) would not have occurred" (as the Court did not accept the loan could be priced as a Term Loan B).
The Borrower's second pricing witness' evidence was not accepted because:
- the Court found that the question he had been asked did not address the statutory question, which required consideration of a loan negotiated at arm's length;
- he based his opinion on the Borrower having a lower credit rating (being "approximately equal to a B rating by" Standard and Poor's) than the Borrower had asserted (BB or BB+);
- he accepted that there could not have been a single lender (also basing his analysis on there being multiple lenders);
- he agreed that the loan would not have been made in the absence of restrictive covenants (but did not consider how such covenants would have affected the interest rate); and
- he had asserted that his bank would have had a minimum fee of 300 basis points (which was not accepted, after he was confronted in cross-examination with documents showing that his bank extended a lower rate to other borrowers).
As the evidence of the Borrower's pricing experts was rejected, the Borrower did not have any evidence which would establish the arm's length consideration for the loan. As such, as discussed above, the Borrower was unable to satisfy the onus of proving that the assessment was excessive.
The other experts
These were not the only expert reports which the Court rejected. The Court considered the reports of the 20 experts called in the matter, rejecting most of the evidence as it did not address a relevant question.
Much of the evidence held to be irrelevant was led in relation to the credit rating that the Borrower would have been given by a rating agency, which was argued by the Commissioner to be relevant. Once the Court found that the practices of rating agencies were not relevant, it followed that the evidence given about credit ratings was not relevant. One of the immediate and practical implications of this case flows from this finding in that it creates uncertainty as to how the pricing exercise should be performed in practice.
In any event, much of the evidence on credit ratings was found to be unpersuasive for other reasons. As examples, the Commissioner relied on the evidence of:
- a witness who attempted to use a statistical model to determine the Borrower's credit rating (rather than applying the policies and practices of rating agencies);
- a witness who was directed to an abstract question of whether a subsidiary's rating could be equalised with its parent, and whose answer was based on a methodology which could not have reliably answered this question;
- a witness who had not been directly involved in issuing credit ratings for some time prior to the relevant loan being made; and
- a witness who had not been a rating analyst.
One witness called by the Commissioner on this question was confronted in cross-examination with an article which he had written in relation to the impact of implicit parental support (which had concluded that there was no market evidence that a lender would provide a lower interest rate for this reason). The article was consistent with the evidence given by the Borrower's witnesses, and was taken into account by the Court in reaching the same conclusion.
The Commissioner also sought to rely upon the evidence of an expert accountant as to the "ideal" currency of the loan. However, the Court accepted that the choice of currency is a commercial matter and was outside the scope of the accountant's expertise. Though not addressed in the decision, commercial decisions around currency are not black and white. As such, evidence as to the "ideal" currency of a loan is unlikely to address the relevant question for transfer pricing purposes.
Having rejected this evidence, the Court was not persuaded that an arm's length loan would have been made in USD, rather than AUD.
Using expert witnesses
There are a number of key points to take away from this decision for parties relying on expert evidence:
- the evidence must be relevant - this an obvious point, and depends on the Court's view as to what may or may not be relevant. Unless the evidence addresses a question of fact which is relevant to the Court's task, the evidence cannot be taken into account by the Court;
- carefully formulate the questions - the evidence of an expert will not be relevant unless they are directed to the right question. In particular, care should be taken to avoid reformulating a statutory test or requirement;
- consider setting the questions collaboratively - issues regarding the relevance of questions are more easily addressed where the parties have agreed as to the relevant questions to be put to the experts (depending on the matter, the Court may also have a role to play in facilitating this process);
- the expert must have relevant expertise - opinions on issues which are outside an expert's area of expertise will not be given weight by the Court. Consider whether there are features of the case at hand which require specialised experience relevant to that transaction or industry;
- due diligence on your own experts - witnesses in this case were confronted with evidence from their own organisations or prior research which did not sit comfortably with their analysis. Where this occurs, there is the potential for significant damage to the witnesses' credibility (which can be managed better if the expert has an early opportunity to address the allegedly contradictory evidence); and
- consider having experts address your opponent's case theory - only leading expert evidence which is consistent with your case theory may leave you exposed if your opponent's case theory is accepted. Consideration should be given to having the expert consider whether your opponent's case theory would lead to a different conclusion.
Given the central role which expert evidence plays in most transfer pricing cases, these issues take on an increased importance. The potential issues with expert evidence should be addressed early, with a view to preventing disputes as to the utility and relevance of an expert's evidence.
The Commissioner was successful in terms of the outcome of this case. This was because the Court found that the Borrower had not led evidence which addressed the Court's interpretation of the proper statutory question.
Importantly, however, the Court rejected the Commissioner's key contentions that would have affected the arm's length interest rate. In particular, the Court rejected the Commissioner's submissions that:
- an arm's length loan would have been made in USD (rather than AUD); and
- that the arm's length interest rate would have been reduced because the Borrower would receive a higher credit rating due to the implicit support of its Parent.
The rejection of the Commissioner's position on currency highlights the commercial nature of this decision. The question of whether arm's length parties would borrow in different currencies is broader than asking what the ideal currency for a loan would be.
The rejection of the Commissioner's position on the relevance of credit rating agencies and the relevance of implicit support will give rise to uncertainty going forward, as it is unclear how an arm's length interest rate can be practically determined (without resorting to expert evidence). Taxpayers using an approach based on rating agency guidance will need to consider their positions.
The implications of this decision will be thoroughly considered in at least three contexts:
- applying the arm's length principle to establish and analyse transfer pricing positions;
- negotiating advance pricing agreements with the Commissioner; and
- transfer pricing disputes.
The appeal process may change or clarify some of the aspects of this decision. However, there are many aspects of transfer pricing that remain unresolved. Importantly, the case did not require the Court to consider Division 815-B, which sets out a different transfer pricing framework (and applies to all future transactions). Given the legal and factual complexities and uncertainties that underpin transfer pricing disputes, taxpayers and the ATO will continue to try to resolve them without litigation. Although the vast majority of transfer pricing disputes are resolved in this way, some may end up before the courts, where effective, collaborative and focused conduct of the dispute will be of vital importance.