Unreal transfer pricing does not solve very real debt-equity issues

7 November 2018

Last week the ATO released its long awaited draft determination on the interaction of the transfer pricing and debt-equity rules. The issue was first the subject of consultation back in February 2015 so it has taken 3 years 8 months to get this far.

The interaction between these regimes is not simple and crops up in complex ways. For example, transfer pricing law can change the price of debt but the size of the "financial benefits" can affect whether an instrument is classified as debt or equity. So, are the debt-equity rules applied first (in which case the price of a debt interest might be changed, but not an equity interest) or are the transfer pricing rules applied first (in which case the interest which was classified as equity under the debt-equity rules might be re-classified as debt and now amenable to being re-priced)?

We made a lengthy submission on the issue to the ATO over two years ago which is attached to this Riposte. The gist of the submission is that the structure of the two sets of rules means that the debt equity rules cannot be applied on the basis of the arm’s length conditions under Australia’s transfer pricing rules – in this regard the argument is similar to that accepted by the ATO in TR 2010/7 that the safe harbours in the thin capitalisation rules which are also based on the debt test in the debt-equity rules cannot be overridden by transfer pricing rules, a result now enshrined in the new transfer pricing legislation.

The submission is based on three main technical grounds:

  1. that the arm’s length conditions in the transfer pricing rules are a different matter to the "pricing, terms and conditions" of the instrument being characterised that determine the application of the debt-equity rules as effectively recognised by the ATO in TR 2010/5;
  2. that transfer pricing rules are applied year by year whereas the character of an instrument as debt or equity is determined on issue until there is a material change in the instrument; and
  3. that the debt-equity rules have a panoply of highly articulated ATO discretions and regulation making powers to deal with the misapplication of those rules.

The ATO picks up these arguments in its alternative views section of the draft determination but makes little real attempt to refute them. In much the same way as during the early part of the technical debate over the relationship of the thin capitalisation and transfer pricing rules, the ATO simply repeats the mantra that transfer pricing prevails over the rest of the legislation.

More fundamentally at the policy level it is doubtful that the arm’s length principle in transfer pricing has anything useful to say on what is a relatively arbitrary (and in the Australian rules mechanical) distinction between debt and equity. This concern is apart from and in addition to the broader and very significant problem transfer pricing encounters in dealing with financing of businesses generally as evidenced by the OECD’s continual failure to come up with guidance on this issue, most recently in its 2018 discussion paper on the topic.

A good example is the treatment of related party at-call loans in the debt-equity rules. The debt-equity and associated rules go to considerable lengths to deal with such loans under provisions which were heavily negotiated between the private sector and government but it now turns out in the ATO view that all this was so much hot air as the rules will apparently often be second guessed by the transfer pricing rules.

This problem is evident in the three examples in the draft. The first and third examples deal with outbound scenarios, and the second example is an inbound case.

The first example is a loan to a distressed subsidiary and, apart from twice confusing which company is the borrower and lender, it assumes arm’s length conditions that are very unlikely to be correct in the situation (a problem which the ATO confesses and avoids by requiring the assumption in all examples that statements about arm’s length conditions are correct) and states that at arm’s length the subsidiary would have borrowed at fixed interest. The ATO turns the loan into a debt interest by assumption so that any interest paid is not entitled to a participation exemption.

In the third example involving a loan to an exploration subsidiary, the assumption by contrast is that the subsidiary could not have borrowed from a third party and so although the interest free loan is a debt interest because of its 9 year term the assumed arm’s length conditions would turn it apparently into shares and an equity interest but this does not occur as no transfer pricing tax benefit flows from the change. So at least this suggests that the ATO view can benefit taxpayers but begs the question of how to decide that a parent company would invest in equity of a subsidiary rather than, say, making an interest-free at-call loan.

The second example is a case where the payment of interest is at the discretion of the borrower subsidiary which would give rise to an equity interest, potentially exposing the payment of interest to dividend withholding tax (up to 30%) and not being entitled to a tax deduction. The assumed arm’s length conditions are a loan at interest which gives rise to a debt interest and payment of interest withholding tax with a (possible) adjustment to give a deduction for the interest. So the ATO could be collecting less withholding tax and possibly giving an interest deduction. One wonders whether the ATO would ever make such an assertion.

In the real world these examples are the kind of cases where an at-call loan would often be made by the parent and would be equity under the debt-equity rules but the ATO studiously avoids an at-call loan example and so does not given any guidance for a very common case. Moreover the ATO examples avoid the timing issue referred to above by having examples dealing only with the issue of the interest and not later years. There is an underlying set-and-forget view of transfer pricing (if not arm’s length at the beginning, never arm’s length and vice versa) that one suspects the ATO will not follow when it produces inconvenient results.

A further surprise from the examples is that the ATO is willing to change the legal form of transactions from debt to shares (and presumably vice versa) as well as the terms of the transactions. The basis on which this is done is hinted at in example 3 in which the reconstruction power in the transfer pricing rules is referred to, without indicating which part of that power is relevant. By contrast in examples 1 and 2 where legal form is not being changed but rather the terms of the transaction other than apparently the price (the interest rate), it would seem that the reconstruction power is being applied but this is not mentioned. In other words the ATO has failed to indicate how the transfer pricing and debt-equity rules interact, which is an important issue that should not be parked in an assumption. Changing matters other than price is a very important boundary in transfer pricing even after the 2013 change in Australia’s transfer pricing rules, the Chevron case and the revision of the OECD transfer pricing guidelines as a result of the BEPS project.

In addition to the draft determination the ATO is also working on a new draft schedule to its Practical Compliance Guideline on cross-border related party financing transactions dealing with interest-free loans. As the draft schedule no doubt will assume the correctness of the draft determination, now is the time for taxpayers to raise its problems. Submissions are due by 30 November 2018.



Cameron Blackwood



David Bond



Professor Richard Vann