Tax Brief 8 April 2011 Part IVA Outbreak It was almost 15 years before the first significant judgment on the operation of Part IVA was handed down. In the last 2 years the Federal Court has delivered ten more, and the Part IVA world became even busier when Treasury released a Discussion Paper in November 2010 on ‘improving the operation’ of Australia’s anti-avoidance rules. This Tax Brief examines these recent developments, both for the light they shed on the current operation of Part IVA and for what they reveal about how Part IVA might be changing in the near future. Background The cases The spate of judgments on Part IVA began in March 2009 with the decision in BHP-Billiton Finance followed later in the year by Ashwick, AXA Asia Pacific Holdings and British and American Tobacco Australia Services Ltd. The cases in 2010 were News Australia Holdings Pty Ltd (May 2010), Trail Bros Steel and Plastics Pty Ltd (July 2010), Citigroup Pty Ltd (August 2010), Futuris Corporation Limited (August 2010) and RCI Pty Ltd (September 2010) and so far in 2011 the Federal Court has handed down the decision in Noza Holdings (February 2011) and the appeal in Ashwick (April 2011). The results of these cases were initially split – the ATO won four (British and American Tobacco, Trail Bros, RCI and Citigroup) while the taxpayers won six (BHP-Billiton, Ashwick, AXA Asia Pacific, News Australia Holdings, Futuris and Noza Holdings). The appeals process in relation to some of the cases has come to an end with the High Court refusing to accept the appeal application (AXA Asia Pacific and BHP-Billiton). In other cases, appeal courts have heard the appeal but decided not to change the result in the lower Court (News Australia Holdings, Trail Bros and Ashwick). At the time of writing, appeals are pending from some of these decisions (British American Tobacco, Citigroup, Futuris, RCI and Noza Holdings). While it is possible that these figures could yet change, the results so far unquestionably show the highly contingent nature of all Part IVA proceedings. Each of the cases involved complex commercial structures and transactions which we will not attempt to describe in detail. But, at the risk of over-simplifying them, in general terms the cases involved: • BHP-Billiton Finance and Ashwick: deducting accruing interest and losses on writing off sizeable bad debts owed by group companies to an in-house finance company; • AXA Asia Pacific Holdings: the sale of a subsidiary in exchange for shares, and electing to apply the scrip-for-scrip rollover provisions to defer tax on the resulting gain; • British and American Tobacco Australia Services Ltd: the intra-group transfer of assets occurring prior to a corporate merger, the subsequent sale of those assets to parties outside the merged group, and offsetting of the resulting capital gains against existing capital losses; • News Australia Holdings Pty Ltd: transactions implemented in order to relocate the headquarters of News Corporation from Australia to the USA; • Trail Bros Steel and Plastics Pty Ltd: contributions to a fund for employees; • Citigroup Pty Ltd: fund-raising through a structured finance product which generated a liability to foreign tax and a claim in Australia for a foreign income tax offset; • Futuris Corporation Limited: the reorganisation of a corporate group prior to the public listing of a subsidiary; • RCI Pty Ltd: the payment of a dividend from an offshore subsidiary which occurred prior to the sale of the subsidiary; and • Noza Holdings: the deductibility to the Australian company of dividends paid on shares that were re-classified as debt under Australia’s debt-equity rules, and the liability to Australian withholding tax of the US-resident shareholders. In these cases, the ATO appears to have decided to adopt the practice of adding a Part IVA claim to each of the substantive issues involved. Whether or not this represents a new tactic in the conduct of tax litigation, it is certainly proving a fertile source for fresh case law on the operation and interpretation of Part IVA. The scope of the ‘scheme’ Many of the early cases on Part IVA involved disputes about the definition of the ‘scheme’ which was said to have been implemented. The results of those cases had led to a sense of resignation – a general impression that there was little to be gained from protracted disputes about the exact size and contents of the ‘scheme.’ The decision at first instance in Ashwick is, therefore, a little surprising in that the judge based his conclusion that Part IVA did not apply on the proposition that lending money at interest to related companies did not amount to a ‘scheme’ at all for the purposes of Part IVA. This approach was not confirmed on appeal, with the Court concluding that there was a ‘scheme’ evident on the facts. 2 Part IVA Outbreak Leaving Ashwick to one side, several of the recent cases have tried new approaches to challenging Part IVA determinations, relying on arguments about the ‘scheme.’ There had always been an intuition that defining the scope of the relevant scheme quite broadly could prove to be important for taxpayers facing a Part IVA determination. The logic was that it would be harder to conclude that tax avoidance was the taxpayer’s principal objective if the taxpayer had done many things under the identified (broad) scheme. Earlier cases had cast doubt on that wisdom. British and American Tobacco Australia Services Ltd was an attempt to see if the reverse argument would fare any better. In this case, the taxpayer was about to be absorbed into another corporate group (which had available capital losses) and was being obliged to divest itself of certain assets (which would generate capital gains when sold to a third party external to the taxpayer’s corporate group) as part of the merger process. The taxpayer’s management decided to sequence these events so that the taxpayer became part of the acquiring group, transferred the assets to another group company (electing to rollover the resulting gain), which then sold the assets to the outside buyer, triggering the gain, which could be grouped with the losses of companies in the acquiring group. Part of the taxpayer’s argument was that the scheme for Part IVA purposes consisted just of making the election to roll-over the asset, and a scheme that consisted of making such an election was immunised from Part IVA under a special provision. This provision gives effect to the obvious notion that if one part of the Act deliberately offers a taxpayer some form of tax relief, Part IVA should not contradict that offer. The taxpayer lost the argument that the ATO’s scheme was too big largely on the basis that its management was alive to the tax issues involved for a long period of time: … the planning for and implementation of the scheme identified by the Commissioner and described above, in relation to the making of the choice by the Taxpayer … commenced many months before the actual disposal by the Taxpayer of the [assets] on 3 September 1999 and continued for some time after that disposal. Thus, the scheme consisted of much more than the mere making of the rollover choice or election. The conclusion is perhaps not surprising and was upheld on appeal to the Full Federal Court. Schemes that are protected from Part IVA. As was noted above, a specific provision in Part IVA switches off its effects where another part of the Act deliberately offers a taxpayer some form of tax relief. Part IVA should not defeat that provision on the assumption that any tax concession was intended. At first instance, the judge in Ashwick took the view that the deductions claimed by the recipients of tax losses, transferred to them by other group companies, were protected from challenge by this provision – the deductions arose from ‘the 3 Part IVA Outbreak making of [an] agreement [that was] expressly provided for by this Act …’ The appeal court approached the matter somewhat differently, relying instead on a finding that the scheme was not carried out with the necessary purpose of securing the tax benefit. Noza Holdings also sought to rely on this section to immunise a transaction against attack under Part IVA. In this case, the taxpayer claimed that the election to consolidate all the Australian subsidiaries of the US parent into a consolidated entity for tax purposes, protected it. The Court disagreed finding that the tax benefit in question – the deduction for dividends paid on shares that are reclassified as debt – arose from a fact that bore no connection to the decision Noza had made to form a consolidated group. Connecting the scheme and the tax benefit. Two of the cases, RCI and Futuris involved a slightly different argument about the scope of the relevant scheme. The question in these cases was, what happens where the scheme identified by the ATO does not actually involve a step which triggers a lesser amount of taxable income? RCI arose out of the reorganisation of James Hardie Industries. An offshore subsidiary paid a substantial tax exempt dividend to its Australian parent in March 1998. In October 1998, the parent sold the shares in the offshore subsidiary. The ATO’s argument was that the payment of the dividend was done to deplete the value of the offshore subsidiary and thus reduce the amount of taxable capital gain realised on the sale of the subsidiary some seven months later. The ATO argued (as its narrow definition of the scheme) that the ‘scheme’ undertaken by the parent consisted just of the payment of the dividend. The taxpayer’s response was that no tax benefit (a reduced capital gain on sale of the shares) arose from that step – any tax benefit arose as a result of the sale of the shares, and the sale was not included in the scheme as the ATO had defined it. The court, however, disagreed. It noted that the tax benefit only has to arise ‘in connection with the scheme.’ The reduced capital gain was sufficiently connected to the payment of the dividend. The same issue arose in Futuris. This case involved the public float of a division of the company and, in particular, the tax consequences of the various steps undertaken in order to consolidate the relevant assets into the float vehicle. These steps involved the movement of assets, the tidying up of cross shareholdings and the capitalisation of the various debts into additional equity. Again, the narrow scheme identified by the ATO focused on the steps involved in tidying up the corporate structure; it did not include the sale of the shares in the float vehicle. Nevertheless, the court considered that any tax benefit that might have arisen on the eventual sale of the shares would still have arisen ‘in connection with’ the restructuring transactions. One scheme or several schemes? RCI also raises a more general question: a dividend was paid by a subsidiary and 7 months later, a parent sold shares in that company, but should these transactions be linked to form a single scheme? A tax benefit must arise in connection with a scheme but why was the sale not its own 4 Part IVA Outbreak ‘scheme’ – that is, just another subsequent transaction? Companies often pay dividends; why was this dividend not just a discrete transaction, unrelated to the sale, especially considering that it was paid 7 months prior to the sale? The court was clearly alive to this issue but decided that the two transactions were in fact connected on the documentary evidence. The dividend and the share sale were both mentioned in the documents emanating from ‘Project Chelsea’ – the name given to the project established to bring about the divestiture of the offshore subsidiary – including documents which pre-dated the payment of the dividend by many months. Purpose The least predictable element in the operation of Part IVA is always determining the sole or dominant purpose of one or more persons who implemented the scheme. The new cases show again just how contentious the process of finding purpose can be, especially when purpose is intended to be determined ‘objectively.’ That is, asking why this taxpayer did what it did is the wrong question. The right question is, why would somebody else have done what this taxpayer did? Significance of the taxpayer’s stated purpose. The dispute in News Australia Holdings arose from the complex series of transactions implemented in order to relocate News Corporation’s headquarters from Australia to the USA and subsequent internal reorganisations. One of the contested issues in the case arose from the restriction that the relocation was to be undertaken subject to a strict requirement that it should generate ‘no tax and no tax risk.’ Ordinarily, in Part IVA cases, the taxpayer will attempt to argue that tax considerations were the furthest thing from its mind. But the taxpayer in News Australia Holdings had clearly made tax considerations a critical element in planning its restructuring and had gone on the record about just how vital tax issues were (albeit limiting tax risk, rather than seeking a tax benefit). So this case presented a novel twist on the usual pattern. Earlier cases had established that the taxpayer’s own assertions that tax was insignificant were not relevant in deciding the case. But what should follow when the taxpayer asserts, instead, that tax was extremely significant in its planning? The ATO was on the horns of a dilemma. It was clearly very tempting to use the ‘no tax; no tax risk’ policy as evidence that the taxpayer had implemented the relocation in a particular manner with the sole or dominant purpose of avoiding Australian tax. The ATO, however, did not rise to the bait. Instead, in the appeal from the decision of the AAT, the ATO put its case entirely on the basis that the taxpayer’s own reasons were irrelevant in applying the purpose test, whether to exonerate the taxpayer or to damn it. In the result, the court concluded that it could find no error in the decision of the AAT (that the “no tax, no tax risk” edict was an objective fact to which reference could be made) and its holding that Part IVA did not apply. 5 Part IVA Outbreak Inconsistent goals. Another issue which arose in RCI was how to reconcile two competing and inconsistent purposes. As was noted above, the case revolved around the payment of a substantial dividend back to Australia by a subsidiary which was sold 7 months later. There was evidence that the taxpayer was suffering losses in Australia and it wanted funds in Australia which would be invested to generate income that could be applied against those losses. This evidence was led, presumably, to suggest that the purpose of procuring the dividend was to derive greater assessable income in Australia rather than less, from the return of those funds to Australia as a dividend. The ATO’s case was that the dividend was paid in order to reduce the capital gain that would be derived when the foreign subsidiary was sold. If that interpretation was accepted, the purpose of procuring the dividend would be to ensure a smaller amount of taxable capital gain. Which of the two conflicting goals should be considered the taxpayer’s dominant purpose? Again, the court resolved this question largely on the evidence. The court concluded that, ‘careful consideration of the evidence’ did not support the argument that the payment of the dividend had its own independent commercial rationale. This finding, of course, makes the conceptual problem disappear. If the dividend was paid principally to deplete the assets of the offshore subsidiary, there is no inconsistency. It seems inevitable that all such arguments will end up being handled in this manner – the court will draw its own conclusion about which of two or more competing purposes was the more influential, and decide the case accordingly. Significance of timing on finding purpose. In British and American Tobacco Australia Services Ltd the Court was obviously influenced by the fact that the taxpayer’s management had consciously waited until the merger was complete before entering the contract to sell the assets to an external party. The terms of the asset sale were largely agreed in April 1999, and formalised in July 1999 but on terms such that the contract for sale would not arise until after the share transaction was complete, which happened in September 1999. It was not just that completion would not occur until after the merger; no contract would arise until after the merger, even though the terms of the sale had largely been finalised in July. The court was obviously influenced by the fact that the taxpayer consciously chose to defer the asset sale for 6 months until the group relationship could be established. The taxpayer’s principal argument – that what it was doing was being done to comply with requirements of the competition regulator – was not decisive. Indeed, it was probably never going to be sufficient because the scheme was regarded as all about how that asset divestiture was done. As the court put it, the relevant question is why the taxpayer went to the effort of creating a situation where it was possible to elect to rollover the gain. Other arguments about the commercial convenience and neatness in having all the assets in a single vendor were dismissed by the court. 6 Part IVA Outbreak On the other hand, the taxpayer in Noza Holdings successfully relied on the timing of the various events to diminish the inference that its dominant purpose was to avoid tax. The taxpayer’s evidence was that it had finalised the relevant share issues and dividend flows at a point in time, and then happened to uncover a new and significant accounting difficulty arising from US GAAP with respect to foreign exchange gains and loss rules. And so, it was in order to solve an accounting problem that the taxpayer then modified its plans and made the foreign exchange issues disappear. As it happened, those steps also triggered the circumstance that a large tax deduction arose in Australia. But the Court was clearly not satisfied that the Australian tax effect had been sought: “When then it is observed (1) that the transactions had been partly effected (2) a problem emerged that required solution and (3) the solution chosen (which it was thought had very favourable taxation consequences) was chosen to avoid disturbance of the arrangements that had already been made and upon which the participants had relied in securing a favourable Private Letter Ruling from the [US tax authorities], the conclusion that [Part IVA] is not engaged must follow.” Seeking a commercial profit v. seeking a tax benefit. The 1996 High Court decision in Spotless Services had already considered whether a taxpayer could defeat a Part IVA determination on the basis that its goal was simply to enhance its profit, rather than secure a tax benefit. The High Court concluded that this was a false dichotomy where the profit only happened from the taxpayer securing the tax benefit. If the tax benefit generated the profit, then pursuing the profit meant the same thing as pursuing the tax benefit. The decision in Citigroup involves the same kind of territory. The taxpayer in Citigroup was apparently keen to give itself higher visibility in a Hong Kong bond market and embarked upon a coupon-stripping transaction as the means of doing so. In the process, it raised about USD50m from external parties. The effect of the coupon strip was to trigger a liability to tax in Hong Kong, the amount of tax being based on the gross proceeds from the sale of the coupons and being payable in the year when the strip occurred. In Australia, however, the amount of income would be spread over several years, leading to a surplus foreign tax credit in the first year, which could be applied against the tax due on other lightly-taxed foreign source income. Judging from the report of the case, the taxpayer’s arguments as to purpose appeared to revolve around three propositions: that the transaction was entirely unexceptional in Hong Kong (indeed it even had the benefit of a tax ruling from the Hong Kong tax authorities); the company could not have been trying to save tax because an Australian tax offset can only arise from paying tax; and the company had little interest in having a tax offset because it did not expect to be earning lightly-taxed foreign income from other sources. Ultimately, the court was unpersuaded by these arguments. The fact that the transaction was untoward in Hong Kong was not relevant to Part IVA; the fact that the group’s total tax cost was unaffected was not relevant – what matters for Part 7 Part IVA Outbreak IVA is whether the Australian component of the tax was reduced; and the evidence did not support the argument that the tax offset was not likely to be useful – there was sufficient evidence to suggest that the taxpayer was probably going to be earning other lightly-taxed foreign income. But the core of the judgment appears to be the judge’s conclusion that securing the tax offset in Australia was critical to making a profit on the transaction. The judge’s analysis was that (i) the transaction was, at best, slightly profitable before tax; (ii) loss-making after the Hong Kong tax was subtracted (but before the Australian tax offset was factored in); and (iii) slightly profitable again once the Australian tax offset was considered. As the judge put it, ‘the post-tax position is negative or in loss … but applying foreign tax credit relief in both situations …, the tables both disclose post-tax positive or profit situations.’ Viewed in this light, the conclusion that securing the tax offset in Australia was an important consideration is probably not surprising. The relevant counterfactual One of the critical questions in any Part IVA case is to determine the relevant counterfactual: what might have happened, but didn’t. The existence of a tax benefit and the amount of the tax benefit is dictated by the counterfactual. But this is obviously a very fraught process because it involves speculating about what the taxpayer might have done. The range of things that a taxpayer might have done – but didn’t – is almost unbounded, and yet it is a vital step in the operation of Part IVA. Until now, apart from Peabody’s case, ascertaining the counterfactual has not been an especially contentious step. Several of the cases being examined in this Tax Brief offer new and important insights into this aspect of Part IVA. Indeed, some of the cases have been won because the taxpayer was able to show that there was no reasonable counterfactual – the alternatives advanced by the ATO were simply unreasonable or failed to accomplish the same commercial outcome. Reasonableness of the alternative. The structure of the legislation imposes some constraints on this enquiry. It requires that the counterfactual be ‘reasonable.’ In AXA Asia Pacific Holdings the taxpayer, a parent company, sold a subsidiary and claimed it was entitled to enjoy the benefit of the scrip-for-scrip rollover provisions. The ATO submitted that the taxpayer would have embarked upon a number of possible alternatives, were it not for the purpose of securing the tax benefit that the scrip-for-scrip provisions afforded, although it offered no evidence about this. The taxpayer won at first instance and on appeal largely because the court disagreed with the counterfactuals proposed by the ATO. The court considered them unreasonable for two main reasons. First, the relevant counterfactual must have been one in which the investment bank would still have been able to collect a sizeable fee on the transaction. Secondly, a counterfactual which involved the investment bank buying the subsidiary was unreasonable because of the contemporaneous documentary evidence indicating that the bank did not want 8 Part IVA Outbreak the subsidiary complicating the process of entry into the consolidation system which the bank was undertaking at about the same time. The court considered it highly likely that the investment bank, ‘would, in my estimation, have had every reason to insist that there was some other, ideally structural, ingredient of the proposed transaction that insulated [the investment bank] from the risk of having [the target company] included as part of its consolidated group.’ In short, while the sellers might have been more than willing to enter the alternative transactions that the ATO offered, the buyer in those hypothesised transactions would not have made the offers upon which the Commissioner’s case depended. That made the alternatives unreasonable and the attack under Part IVA miscarried. The validity of this analysis was upheld on appeal. The same kind of issue arose in Futuris. The taxpayer wanted to float a portion of its business and, in order to do so, transferred a number of assets to other group companies, tidied up some cross shareholdings and capitalised some existing debts. These transactions had the effect of triggering capital gains tax, deferred in some cases by rollovers, and invoking the value shifting rules. The court decided in the taxpayer’s favour on the basis that it simply would not have implemented some of the possible counterfactual transactions. One possibility – that all of the assets would have been transferred to a new float vehicle – was discounted on the basis that it would have triggered substantial stamp duty. Another possibility – that the assets would have been transferred to one of the existing subsidiaries – was discounted on the basis that it would have triggered capital gains tax twice within the corporate group. A third possibility – that the assets would have been sold to a different existing subsidiary – was discounted on the basis that it would have generated an even larger capital gain than the second alternative. Before moving on, it is worth pausing to note the interesting role that triggering higher tax plays in the judge’s reasoning in Futuris. A tax benefit can only arise from an alternative which involves higher tax. But the judge viewed some of the alternatives which triggered higher gains (and thus tax) as being unreasonable for that very reason. There is a tension here. There must an alternative which would have triggered higher tax, but the very fact that it triggers higher tax could make the alternative unreasonable. Perhaps the lasting importance of this case is in what it says about handling Part IVA controversies. It shows the importance of the use of expert evidence and the role that experts may increasingly play – to opine on what was commercial or implausible. Who has the responsibility? Several of the cases have involved disputes about the onus of proving what might have happened in lieu of the actual events. Must the ATO set out a reasonable alternative that leads to higher taxable income, and which the taxpayer must then defeat? Or is it the taxpayer’s responsibility to set out the range of alternatives and then demonstrate why (i) some are not real alternatives – they do not achieve the same commercial outcome; (ii) some are unreasonable – they are simply too expensive to implement; (iii) the rest would 9 Part IVA Outbreak not have involved a higher amount of tax; and (iv) the ones that did involve more tax, would have involved other taxpayers? It now seems reasonably clear that the ATO does not have to show what would have happened instead of the scheme. In Futuris, the taxpayer had applied to the court in earlier proceedings for an order that the ATO provide particulars of the relevant counterfactual being used to calculate the amount of the alleged tax benefit. The court refused to give the order and this was noted in the later litigation in August 2010. In the second case, the court quoted from the earlier decision that, ‘it is for the [taxpayer], who bears the onus, to establish a series of transactions or arrangements which, it contends, would have ... resulted ...’ Further, the court observed that if the ATO does propose a counterfactual, ‘it is not enough for [the taxpayer] to disapprove any alternative ... put up by [the ATO]... [It] must also adduce evidence which satisfies the court that it had an alternative ... which it would have been able to implement and which would have resulted in the same taxable position as ... the scheme ... [or] some other taxable position.’ In short, it is the taxpayer’s responsibility, and the taxpayer is entitled to win if it proposes a reasonable alternative and can convince the court that the alternative was plausible, likely and would have resulted in no more tax, or perhaps even less. And presumably the taxpayer is also entitled to win if it can demonstrate that there is no reasonable alternative that would accomplish the commercial objective that it sought. The possibility that there might be no alternative that could accomplish the same commercial outcome arose in Noza Holdings. In that case, the taxpayer claimed deductions for dividends on shares that were re-classified as debt under Australia’s debt-equity rules. The ATO argued that all the companies in the chain could instead have issued ordinary shares, which would be classified as equity instruments in Australia, and not generated a tax deduction when dividends were paid from Australia. However, one of the offshore companies had already negotiated its position with US tax authorities on the basis that the share issues would be debt for US tax purposes, and the ATO was not able to identify any instrument that could successfully straddle the debt-equity divide in both countries, for both tax and commercial purposes. Is the alternative affected by the circumstances of the taxpayer? It is now well established that, in Part IVA, the search for the ‘purpose’ of the scheme is objectively determined – that is, the issue is not why did this taxpayer do what they did? Rather, the issue is why would somebody else have done what the taxpayer did? When it comes to the counterfactual, the same question is in play. That is, given the broad range of alternatives that might have been implemented, does the court ask, what is it reasonable to expect that this taxpayer would have done instead? Or does the court ask, what is it reasonable to expect that some putative representative taxpayer might have done? To put it a little more colourfully, is it a defence to a Part IVA assessment to argue that this taxpayer is the kind of person 10 Part IVA Outbreak who would have just kept on looking for some other ‘no tax’ option, and would have continued down that path until one was found? Can the taxpayer win its case on the argument that, the idea that it would just have given up, and agreed to implement the transaction in a way that triggered tax, is simply implausible? This question lay behind the dispute in Trail Bros. The case arose out of the enactment of the age-based limits on deductible superannuation contributions in 1997. Faced with a cap on the amount that it could contribute in a tax effective manner, the employer ceased contributing to the superannuation fund and began instead to make contributions to an employee benefit trust, claiming a deduction for the contributions. The Administrative Appeals Tribunal which heard the case at first instance found for the taxpayer on the basis that there was no ‘tax benefit.’ The tribunal came to this conclusion because of its view that, assuming the taxpayer did not make a contribution to the employee benefit trust, it would simply have implemented some other transaction that triggered a similar tax deduction. When the case came to the Full Federal Court, it was affected by the lack of evidence led before the AAT. However, it seems clear that the court considered that the relevant counterfactual is to be determined by looking at this taxpayer. It was unfortunate for the taxpayer, therefore, that it had not led any evidence to suggest what it would have done instead of implementing the employee benefit trust structure. The same kind of approach was adopted in Futuris. The judge was content to conclude that none of the alternative transactions proposed by the ATO was reasonably likely, and so no tax benefit arose. The judge approached the question by asking in effect, what would this taxpayer have done instead? And he answered that question by concluding it was unreasonable to expect that the taxpayer would have implemented some of the options that were more expensive in terms of tax. A similar issue arose in News Australia Holdings where the court appeared to accept the taxpayer’s argument that if it were not able to implement the restructuring transaction without tax, it would have persisted with a tax inefficient structure until a “no tax” means of restructuring could be identified. A similar approach is evident in AXA Asia Pacific Holdings where the first instance judge placed significant reliance on testimony of representatives of the investment bank and the taxpayer that the alternative transactions identified by the ATO were not ones they would have contemplated. However, in Futuris the court cautioned that, ‘the counterfactual must not itself be a scheme with a dominant tax purpose.’ This seems to suggest that the relevance of the taxpayer’s own preferences is limited. So there appears to be a tension that has to be managed. On the one hand, the taxpayer is free to argue that it would have pursued some other tax-effective option, but the argument will not help the taxpayer if the tax-effective option is considered by the court to be just a different tax avoidance scheme. 11 Part IVA Outbreak What survives to be considered? There is another conceptual conundrum underlying the process of identifying what might have happened, but didn’t. How much of what actually happened is to be disregarded? Does Part IVA require everything that happened to be ignored, or can some of the things that did happen survive to form part of the counterfactual? This issue was argued in AXA Asia Pacific Holdings. The taxpayer argued that an earlier decision, Lenzo, put serious obstacles in the ATO’s path because the counterfactuals being offered involved some of the steps that actually occurred. The taxpayer had argued that the Lenzo decision stood for a series of propositions: that Part IVA (a) required the ATO to cancel everything that did happen, and (b) to set out what would have happened but didn’t, but that (c) in doing step (b), the ATO could not put up as part of the counterfactual an alternative which involved any of the steps which had actually occurred. In the words of the legislation, the taxpayer is to be dealt with on the basis that would have applied ‘if the scheme had not been entered into or carried out.’ The issue arose because the counterfactual presented by the ATO involved AXA selling its subsidiary to a specially-incorporated company, but that transaction was also part of ‘the scheme’ that the ATO had identified. If the counterfactual required the elimination of everything identified as forming part of the scheme, then the buyer would not have existed and the ATO’s counterfactual collapses. The court was not convinced. The Court said it agreed with the lower court that the relevant counterfactual could – and likely would – involve some of the steps of the scheme which the taxpayer had actually implemented. The lower court put it this way: examining what the taxpayer did, ‘is likely to shed much light on what they would have done in the absence of the scheme ...’ There is a flip-side to this question. Can the taxpayer insist that everything that did happen has to be included in the counterfactual? This was at issue in the appeal in AXA. As was noted above, the lower Court had concluded that the Commissioner’s counterfactuals were not reasonable because they would have deprived the investment bank of its fee. On appeal, therefore, the Commissioner argued that the relevant counterfactual should exclude the contracts that gave rise to the fee, or else, if the fee mattered, the Court should simply assume that there was some other basis for which a fee might have been charged. The Court rejected both arguments as being contrary to the evidence that was led, and not supported by any other evidence. So, the counterfactual hypothesis need not be an entirely dissimilar transaction. Rather it is quite likely to be a transaction which involves some of the events which actually occurred, and perhaps some of the critical steps which occurred. The implication of this is it becomes more likely there will be no tax benefit – the alternatives will end up looking very much like what actually happened. A counterfactual with consequences. In some situations, in order for a tax benefit to arise, the counterfactual transaction has to involve an amount of income or capital gain being derived. That is, for the ATO’s case to bite, it is not 12 Part IVA Outbreak sufficient simply to say that the taxpayer would have done something else; it is also necessary to show that the ‘something else’ would have led to a larger amount of income or gain. Where the tax benefit involved is a tax deduction, no such problem exists – the ATO needs only to assert that the deduction would not have been available in order for a higher amount of tax to appear. The court was alive to this in AXA Asia Pacific. It noted that the relevant counterfactual had to involve some form of transaction which would have triggered additional capital gain and tax. In AXA that meant some kind of sale of the target. The court said, ‘the Commissioner needed to replace the presumptively eliminated scheme with some positive transaction ...’ The ‘do nothing’ option. The decision in News Australia Holdings raised consideration of a counterfactual where the taxpayer simply does nothing. It has often been argued that, in many cases, the most likely counterfactual is that the taxpayer will simply sit on its hands if it cannot undertake a particular transaction without triggering a tax cost. Part IVA operates by taxing the taxpayer as if it had undertaken the alternative transaction instead of the one it did. If the most likely alternative is one that does not trigger an amount of income or capital gain, the argument is that Part IVA simply miscarries. The issue arose in News Australia Holdings because of evidence the taxpayer gave that it would not have executed one of the final steps in its relocation if doing so would have triggered tax. Rather, it would have been prepared to live with an inefficient corporate structure. Unfortunately, the court did not have to address the substantive merits of this argument. It simply noted the argument and moved on. It is worth pointing out that this argument has concerned the ATO for some time. Indeed, there was a recommendation by the Ralph Review in 1999 that would have attempted to overcome this argument. That recommendation has never been acted upon. It is worth noting that the drafting of Part IVA recognises this issue in the special regime for dividend stripping. A taxpayer with a redundant company that can only eliminate the company through receiving a liquidating dividend or other distribution will in fact often do nothing with the company, as the existence of many redundant companies testifies. In such a case, the ‘do nothing’ argument is very plausible: if the taxpayer cannot take out the dividend, the company would simply be sitting there. However, this argument is effectively bypassed by deeming a tax benefit to arise where dividend stripping occurs and a dividend paid just before the scheme would have been included in assessable income. There is no need to prove what would have happened instead of the dividend strip. The ‘do (almost) the same thing’ alternative. A variant on this was also presented in News Australia Holdings. The taxpayer argued that whatever counterfactual the ATO proposed, it had to be one which satisfied the ‘no tax, no tax risk’ condition which had been imposed on the entire relocation project. The 13 Part IVA Outbreak argument was that, given the taxpayer’s commercial goals and restrictions, the most likely alternative, assuming that some course of action would have happened, would still have to be one which met the ‘no tax, no tax risk’ requirement. So, even if the ‘do nothing’ alternative was not available, nevertheless, anything else would have had to lead to no tax consequences as well. Again the issue was noted by the court but the court was able to resolve the appeal without needing to address the validity of this argument directly. Timing. Finally, there is a discussion in AXA Asia Pacific about the time at which all this conjecture and speculation has to be undertaken. Obviously, something may be implausible at one point in time but very plausible at another. So, what is the point of time at which one looks to determine whether the counterfactual is reasonable or not? The issue arose in AXA because the transaction was being put together at about the time that the investment bank was trying to put together a consortium of investors to buy the target. The fact that the parent company would have lost a fee on the transaction only became an impediment once the contract that created the entitlement to the fee had been signed. The transaction was also occurring at about the same time as the investment bank was organising to enter the consolidation regime. At some other time, neither impediment would have been important. The court did not address this issue in abstract, although it was clearly in the mind of the judge. There is, however, some intuition that can be gained from looking at the court’s method of dealing with the issue. With regard to the fee income issues, the court considered the timing dimension by looking only at a single point in time. For example at one point in the judgment the court said that evaluating what was reasonable, ‘must be asked, notionally, at 3 June 2002, or some other point in time thereabouts ...’ In other words, reasonableness was to be determined at a single point in time. This leaves open the more difficult question whether the ATO is entitled to win if it can point to any time when the counterfactual might be considered reasonable. With regard to the consolidation issue, however, the court considered a range of times but with a terminal point. The investment bank was planning to elect to consolidate with effect from 1 October 2002. The transaction documents, which had been signed in June 2002, suggested that completion of the sale of the target might occur at any time between August 2002 and mid October 2002. The court was apparently convinced by the evidence of one of the experts that the existence of the target within the group as at 1 October 2002 would not have been acceptable. Given that completion could have occurred prior to 1 October 2002, with the effect that the target company might have been on hand on the date of consolidation, the court considered this made the ATO’s counterfactual unreasonable. Presumably this same logic would have applied if the hypothesised sale might not have occurred until some time after 1 October 2002. 14 Part IVA Outbreak A tax benefit for this taxpayer. Finally, it is worth noting that the tax benefit must be one that would have arisen for the taxpayer which the ATO has assessed. The ATO failed in its attempt to apply Part IVA in Futuris in part because the Court found that under the only acceptable counterfactuals, other companies in the group would have obtained a tax benefit and not the taxpayer. This result harks back to the Peabody case – the first High Court decision on Part IVA – where the ATO failed because it issued the assessment to the wrong taxpayer. It remains to be seen whether the outcome of the result in Futuris is that the ATO will issue multiple Part IVA assessments to every conceivable taxpayer, and then promise at trial to abandon the unsuccessful overlapping and conflicting assessments once a judgment is handed down. Part IVA and other anti-avoidance rules Futuris raises another question which has dogged the operation of Part IVA – how does Part IVA fit with other anti-avoidance provisions in the Act? In Futuris, the ATO determined that the taxpayer had derived a tax benefit in connection with the float of its subsidiary of $82.9m. This amount was in fact the outcome of an increase to cost base because of transactions which triggered the operation of the value shifting rules. The taxpayer argued that Part IVA could not be invoked to defeat consequence that another anti-avoidance regime had intentionally brought about. It also argued that an increase in cost base caused by the value shifting rules could not be a tax benefit. Further, it argued, it was nonsensical to argue that a series of transactions which enlivened an anti-avoidance rule, was actually undertaken for the sole or dominant purpose of securing a tax benefit. After all, it does seem more than a little surprising to suggest that when a taxpayer succumbs to the effects of an anti-avoidance rule its intention is to gain a tax advantage. The court, however, did not accept the taxpayer’s arguments. Rather, the court took the view that the taxpayer’s scheme could be one which worked by triggering an anti-avoidance rule like the general value shifting rules. It has not been unknown in the past for anti-avoidance rules to be turned to a taxpayer’s advantage and this case is a demonstration of the possibility. Thus, it seems, specific anti-avoidance rules do not preclude the operation of Part IVA. Indeed, specific and general anti-avoidance rules will likely operate cumulatively and sequentially. If, as Futuris shows, Part IVA can still operate where a specific anti-avoidance rule has been triggered, the case that Part IVA can apply must be even stronger when an anti-avoidance rule that might have applied has not actually been enlivened. 15 Part IVA Outbreak Treasury’s Discussion Paper As part of the May 2009 Budget, the former Assistant Treasurer Chris Bowen announced that, ‘the Government [would] shortly release a discussion paper canvassing options to consolidate, streamline and improve the operation of provisions designed to counter tax avoidance.’ That Paper was released by Treasury in November 2010. The two substantive chapters of the Paper give equal attention to possible changes to Part IVA and options for rationalising the many specific anti-avoidance rules in the legislation. So far as Part IVA is concerned, the Paper addresses only one substantive issue: options for expanding the concept of ‘tax benefit’ for the purposes of Part IVA to include other mechanisms by which the amount of tax payable might be reduced or deferred. It is assumed that the current list of possible ‘tax benefits’ is inadequate; the only discussion is on the best means of expanding it. The Paper includes some matters of form such as re-writing and relocating Part IVA into the 1997 Act and how best to switch off s. 260, as it hasn’t applied to tax schemes entered into after 1981. The rest of the Paper examines how to consolidate and co-ordinate the many specific anti-avoidance rules. Submissions on the Paper were due with Treasury by 18 February 2011. For further information, please contact Sydney Andrew Mills Chris Colley Director Director 61 2 9225 5966 61 2 9225 5918 mailto:andrew.mills@gf.com.au mailto:chris.colley@gf.com.au Melbourne Tim Neilson Andrew de Wijn Director Senior Associate 61 3 9288 1054 61 3 9288 1227 mailto:tim.neilson@gf.com.au mailto:andrew.deWijn@gf.com.au These notes are in summary form designed to alert clients to tax developments of general interest. They are not comprehensive, they are not offered as advice and should not be used to formulate business or other fiscal decisions. Liability limited by a scheme approved under Professional Standards Legislation 16 Part IVA Outbreak Greenwoods document 510133664 Greenwoods & Freehills Pty Limited (ABN 60 003 146 852) Sydney Level 39 MLC Centre Martin Place Sydney NSW 2000 Australia Ph +61 2 9225 5955, Fax +61 2 9221 6516 Melbourne 17 101 Collins Street, Melbourne VIC 3000, Australia Ph +61 3 9288 1881 Fax +61 3 9288 1828 Part IVA Outbreak