University of New South Wales Law Journal
|
|
Home
| Databases
| WorldLII
| Search
| Feedback
University of New South Wales Law Journal |
|
An important theme that emerges from the final report[1] of the Review of Business Taxation (“Ralph Review”) is the need for greater coherency in the design of Australia’s international tax rules. It is necessary to review these rules to ensure that they are both internally consistent and broadly compatible with similar rules in other countries. This second point is really an aspect of the tax debate relating to globalisation. In simple terms, differences in international tax rules between countries will lead to gaps and overlaps in the assertion of tax jurisdiction. The gaps will be exploited by the owners of mobile capital and the overlaps may discourage foreign investors.
Because of time constraints, the Ralph Review did not undertake a detailed examination of Australia’s foreign income rules. While the Ralph Review made some specific suggestions particularly in relation to the foreign investment fund (“FIF”) regime,[2] it recommended that a comprehensive review of Australia’s foreign income rules is undertaken.[3] As a contribution to that review, this paper deals with an issue of internal consistency in the application of Australia’s foreign income attribution rules, in particular, at the border between the controlled foreign companies (“CFC”) [4] and FIF regimes.
The control and substantial shareholder rules in the CFC regime effectively define the border between the CFC and FIF regimes. The CFC regime applies only to Australian residents who are substantial shareholders in a non-resident company that is a CFC. The FIF regime applies to other Australian residents with interests in non-resident companies.
There are three alternative tests for determining whether a non-resident company is a CFC. Under the main test, a non-resident company is a CFC if a group of five or fewer Australian entities with at least a 1 per cent interest together with their associates hold a 50 per cent or greater interest in the company.[5] This is referred to as the ’strict control test’. There are two alternative tests based on economic control. A foreign company is a CFC if a single Australian entity together with its associates has a 40 per cent or greater interest in the company, unless the company is controlled by a group of entities not including the Australian entity or its associates.[6] This is referred to as the ‘objective de facto control test’. Finally, a foreign company is a CFC if the company is controlled (in an economic sense) by a group of five or fewer Australian entities either alone or together with associates.[7] This is referred to as the ‘subjective de facto control test’. Where a non-resident company is a CFC, the CFC regime applies only to Australian residents who are substantial shareholders, meaning that they and their associates have a 10 per cent or greater interest in the company.[8]
An Australian resident who is a substantial shareholder in a CFC will be subject to the CFC regime. The amount (if any) attributed to the resident will depend on the possible application of two exemptions in the regime. The main exemption is for active income. In broad terms, the CFC regime attributes only income from passive investments (ie, essentially mobile capital) and base company operations. The other exemption is for income that is taxable in a country listed as having an income tax broadly comparable to Australia’s. Income of the CFC that is not excluded by these exemptions is attributed to the Australian resident on a branch-equivalent basis. This involves calculating the taxable income (referred to as ‘attributable income’) of the CFC according to Australian taxation law and including the Australian resident’s share of attributable income in their assessable income.[9]
The FIF regime applies to any Australian resident who has an interest in a non-resident company that is not subject to the CFC regime.[10] In other words, it applies to Australian residents with interests in a foreign company that is not a CFC and to residents with non-substantial shareholdings in a CFC. As with the CFC regime, the amount (if any) attributed to a resident under the FIF regime will depend on the possible application of exemptions in the regime. The main exemption is for active income, although the design differs from that in the CFC regime. The FIF regime does not include a comparable tax exemption.
Attribution under the FIF regime depends on the application of three alternative surrogate measures of the FIF’s income, rather than measurement of actual taxable income using branch-equivalent calculations. Surrogate measurement is based on periodic movement in market value of the shares, the application of a statutory interest rate, or accounting profit. This reflects the fact that the FIF regime is essentially dealing with Australian residents with small interests in non-resident companies who, generally, do not have access to the information necessary to make branch-equivalent calculations.
The fundamental issue for Australian investors in non-resident companies is to determine which regime (CFC or FIF) applies. This involves the application of the control and substantial shareholder rules in the CFC regime. It is likely that the different exemptions in the two regimes will encourage planning around the control and substantial shareholder rules as taxpayers attempt to bring themselves within one regime in preference to the other. Clearly, the design of the rules must take this into account. However, the starting point must be to determine a basic principle for defining the border between the two regimes. The principle should relate to outcome, namely that the CFC regime applies to those who can make branch-equivalent calculations and the FIF regime applies in other cases. The issue is whether the control and substantial shareholder rules in the CFC regime achieve this outcome.
The control and substantial shareholder rules closely follow the equivalent rules in the Canadian CFC regime.[11] However, there is a fundamental difference in the broad design of the Australian and Canadian anti-deferral regimes. While both jurisdictions have two anti-deferral regimes applicable to interests in non-resident companies, the roles of the regimes differ greatly. The Australian CFC regime applies to residents who are substantial shareholders in a CFC and the FIF regime applies to residents with other interests in non-resident companies. In other words, the FIF regime is a separate stand-alone anti-deferral regime applicable mainly in portfolio cases. The Canadian CFC regime applies to Canadian residents who are substantial shareholders in a CFC and the offshore investment fund (“OIF”) regime applies to residents who have arranged their investments so as to avoid the CFC regime.[12] In this way, the OIF regime acts as a ’backstop’ to the CFC regime, rather than as a separate stand-alone anti-deferral regime.
This means that the underlying policy basis of the control and substantial shareholder rules is different in the two countries. In the Canadian context, Professor Arnold argued that the rationale for the control rule is ’fairness’. It is arguably unfair to tax residents on their interest in the undistributed income of a non-resident company if they do not have the power to require the non-resident company to distribute the income.[13] On this basis, anti-deferral measures should not apply to persons who do not control the distribution policy of a non-resident company as those persons are unable to bring about the benefit of deferral. This is the position in Canada as the OIF regime applies only in cases where the situation of non-control is artificially constructed.
The Australian FIF regime is deliberately targeted at non-control cases. Thus, the FIF regime applies even though investors subject to the regime have no control of the company’s distribution policy. Thus, in the Australian context, the control and substantial shareholder rules do not define the border between attribution and deferral (as in Canada), but instead they define the border between branch-equivalent and surrogate taxation as the basis for the elimination of deferral.
The control and substantial shareholder rules in the Australian CFC regime were designed at a time when the precise role of the FIF regime was not settled, as the FIF regime was introduced some two and half years after the CFC regime. Initially, it appeared that the FIF regime would follow the Canadian model and act as a backstop to the CFC regime.[14] In this case, the focus of the control and substantial shareholder rules would have been on controlling distribution policy. Ultimately, however, the FIF regime adopted was a separate stand-alone anti-deferral regime. The control and substantial shareholder rules in the CFC regime were not revised on introduction of the FIF regime. Accordingly, there is a question as to whether these rules are appropriate in the context of Australia’s anti-deferral regimes as a whole.
An initial policy question to be resolved is whether it is only those shareholders who control distribution policy that can make branch-equivalent calculations. Intuitively, the answer is no, as any shareholder with a significant interest in a company should have access to the accounting records of the company (perhaps through the appointment of a director) required to make branch-equivalent calculations. Thus, it is arguable that control of distribution policy is not an effective indicator of the taxpayers to whom the CFC regime should apply.
On this point, the Australian legislation gives inconsistent outcomes. On the one hand, an Australian resident who holds a 50 per cent interest in a non-resident company as a joint venture with a non-resident party is subject to the CFC regime. This is the case even though the Australian resident may not control the distribution policy of the company (ie, distribution decisions require unanimous consent). Clearly, in this case, the Australian resident would be able to make branch-equivalent calculations and, therefore, it is right that the CFC regime applies.
On the other hand, an Australian resident with a 49 per cent interest in a non-resident company is not subject to the CFC regime if the other 51 per cent is held by an unrelated non-resident. In this case, the objective de facto control test does not apply because the non-resident company is ‘controlled’ by the non-resident shareholder. Thus, the Australian resident will be subject to a surrogate basis of taxation under the FIF regime. But is there any difference in substance between this case and the joint venture case described above? In neither case can the Australian resident control distribution policy, but in both cases the Australian resident should be able to make branch-equivalent calculations. Does it make sense to apply surrogate methods of calculation to the 49 per cent shareholder when clearly that person can make branch–equivalent calculations? The answer must be no. Indeed, the same can be said for any Australian resident who satisfies the presumptive limb of the objective de facto control test. Surely, anyone with a 40 per cent interest can make branch-equivalent calculations, so what does it matter who really controls the company?
The strict control test includes not only actual control cases, but also ability to control cases. As noted earlier, a non-resident company is a CFC if a group of five or fewer Australian entities with at least a 1 per cent interest together with their associates have a 50 per cent or greater interest in the company. While these entities may not actually act together to control the company, they can always come together to do so. Thus, the purpose of this aspect of the strict control test is to treat any non-resident company with a controlling interest closely-held by Australian entities as a CFC. Once an Australian controlling group has been identified, only those members of the group with a 10 per cent or greater control interest are subject to the CFC regime (ie, a substantial shareholder rule applies).
Two issues arise in respect of the strict control test. The first is why the test should measure control in terms of Australian entities. Overseas regimes use this test as a measurement of ability to control distribution policy.[15] It is assumed that a small number of shareholders will come together to control distribution policy only if those shareholders have a common objective in determining distribution policy. Here the objective is to achieve a tax advantage, but this will only be a common objective if all members of the controlling group are seeking the same tax advantage. Thus, the control group must comprise similarly situated persons, namely residents of the same country.
The assumption of same residence is questionable where the benefit sought is deferral. Achieving the benefit of deferral is likely to be a common objective of taxpayers in many countries. Given the greater mobility of capital today, investors from different countries can easily come together to seek deferral advantages. Moreover, it is not clear why the same residence assumption should play a role in the design of the Australian rules. The concern in identifying a controlling interest under the Australian system is not distribution policy, but rather access to information to make branch-equivalent calculations. The access to information issues must be the same regardless of whether the non-resident company is closely held by Australian interests or closely held by foreign interests (with some Australian participation). The outcome under the current rules is that an Australian resident with an interest in a non-resident company closely held by Australian interests may be subject to branch-equivalent taxation under the CFC regime, but an Australian resident with exactly the same interest in a non-resident company closely held by foreign interests is subject to surrogate taxation under the FIF regime. There seems little justification for the different outcomes as, in both cases, the Australian resident would have the same access to information.
The second issue is whether it makes sense to have both a concentration requirement in the control rule and a substantial shareholder rule. Professor Arnold argues that the justification for the substantial shareholder rule is that small shareholders are not in a position to access the information necessary to make branch-equivalent calculations.[16] But the substantial shareholder rule is used only in conjunction with a control rule that has identified a non-resident company as closely-held. The non-resident company will be a private company. Is it not reasonable to assume that every shareholder in a closely-held private company will have access to the information necessary to make branch-equivalent calculations? Indeed, Professor Arnold concedes that it will be a rare case where a person will have an interest in a private company closely held by complete strangers.[17] There is likely to be some connection between the shareholders to justify the assumption that all members of the company can make branch-equivalent calculations. Again, it seems to make little sense to subject some members of the controlling group to branch-equivalent taxation and other members to surrogate taxation.
It seems that the current control and substantial shareholder rules do not achieve the appropriate outcome. In reviewing these rules, the guiding consideration must be access to information. One approach would be to set the border between the CFC and FIF regimes by reference only to a substantial shareholder rule. The issue then would be the setting of the threshold. This requires a realistic assessment of the level of shareholding required, in the ordinary case, to have access to information to make branch-equivalent calculations.
A concern with the application of only substantial shareholder rule is that it may not get the right outcome for all residents with interests in closely-held non-resident private companies. One is inescapably drawn back to the original Australian anti-deferral regime proposed in the Consultative Document.[18] This document proposed branch-equivalent taxation for all interests in non-resident private companies (no matter how small) and a substantial shareholder rule (based on 10 per cent) for interests in non-resident public companies. It is suggested that the original Consultative Document proposal should be revisited.[19] In the case of non-resident private companies, the original Consultative Document proposal may have been too broad. Clearly, there will be cases where an Australian resident will have an interest in a widely-held non-resident private company and, thus, will not have access to the information necessary to make branch-equivalent calculations. However, branch-equivalent taxation should apply to every Australian resident with an interest in a closely-held non-resident private company, whether or not there is Australian control of the company. It is reasonable to assume that every member of a closely-held private company will have access to the information necessary to make branch-equivalent calculations (eg, no substantial shareholder rule should apply in this case).
It has been argued that the current design of the control and substantial shareholder rules does not lead to consistent outcomes in the operation of Australia’s anti-deferral regimes as a whole. It is suggested that the CFC regime should apply to an Australian resident with a substantial interest in a non-resident company. It should also apply to an Australian resident with any interest (no matter how small) in a closely-held non-resident private company. The FIF regime should apply in other cases.
Regardless of whether the design above is accepted, the critical point is that these rules must be reviewed to ensure that the CFC regime properly applies to those who can make branch-equivalent calculations and the FIF regime applies in other cases.
[*] Faculty of Law, University of Sydney. The author thanks Professor Rick Krever for comments on an earlier draft of this article. All errors and omissions are the sole responsibility of the author.
[1] Australia, Review of Business Taxation, A Tax System Redesigned: More Certain, Equitable and Durable, July 1999 (“Ralph Review”).
[2] Ibid at 683. The FIF regime is in Income Tax Assessment Act 1936 (Cth) (“ITAA 1936”), Pt XI.
[3] Ibid.
[5] Ibid, s 340(a). See ss 318, 349-55 for the determination of interests.
[8] Ibid, s 361(1)(a).
[9] The tax calculation is essentially the same as for Australian residents with a foreign branch, hence the reference to “branch-equivalent taxation”.
[10] Indeed, every non-resident company (including a CFC) is a FIF (ITAA 1936, s 481). However, where a non-resident company is both a CFC and a FIF, there is a reconciliation rule which gives priority to the application of the CFC regime (ITAA 1936, s 494).
[11] There are some specific differences in design. Canada applies a ’more than 50 per cent interest’ test and does not have de facto control tests (although to some extent that is the role of OIF regime discussed infra in the text). See OECD, “Controlled Foreign Company Legislation” in Studies in Taxation of Foreign Income, OECD (1996) pp 120-2.
[12] Income Tax Act (Canada), s 94.1.
[13] B Arnold, The Taxation of Controlled Foreign Corporations: An International Comparison, Canadian Tax Paper No 78, Canadian Tax Research Foundation (1986) p 415. A similar rationale is offered in the OECD, note 11 supra, p 33.
[14] Australia, Treasurer, The Taxation of Foreign Source Income: An Information Paper, 1989 at 18 and 152-5.
[15] See American Law Institute, International Aspects of United States Income Taxation, ALI (1987) pp 190-1.
[16] Arnold, note 13 supra, p 426.
[17] Ibid, p 427.
[18] Australia, Treasurer, The Taxation of Foreign Source Income: A Consultative Document, 1988.
[19] Albeit, in the context of method of attribution, rather than the context of attribution or deferral as under the Consultative Document proposal.
AustLII:
Copyright Policy
|
Disclaimers
|
Privacy Policy
|
Feedback
URL: http://www.austlii.edu.au/au/journals/UNSWLawJl/2000/44.html