Inland Revenue - Tax policy Tax Policy

News and information about the Government's tax policy work programme, including:
- proposed changes to the laws that Inland Revenue is responsible for
- updates on the progress of bills through Parliament
- policy announcements

01 - Policy report: Consultation on addressing hybrid mismatch arrangements

Document number 01
Date 9 March 2017
Type Policy report
Title Consultation on addressing hybrid mismatch arrangements
Description

Report on submissions received for the Government's discussion document Addressing hybrid mismatch arrangements (September 2016) – for the submissions see document #21.

Report number: IR2017/133, T2017/0406

Downloads

PDF (1.7 MB; 23 pages)
DOCX (206 KB; 23 pages)

Contents

Tax policy report: Consultation on Addressing Hybrid Mismatch Arrangements


Date: 9 March 2017 Priority: Medium
Security Level: In Confidence Report No: T2017/406
IR2017/133

Action sought

  Action Sought Deadline
Minister of Finance Note the content of this report
Agree to the recommendations in this report
21 March 2017
Minister of Revenue Note the content of this report
Agree to the recommendations in this report
21 March 2017

Contact for telephone discussion (if required)

Name  Position Telephone
Matthew Gan Tax Specialist, The Treasury [Withheld under section 9(2)(a) of the Official Information Act 1982]
Paul Kilford Policy Manager, Inland Revenue
Casey Plunket Special Policy Advisor, Inland Revenue

 

9 March 2017

Minister of Finance
Minister of Revenue

Consultation on Addressing Hybrid Mismatch Arrangements


Executive summary

Discussion document

On 6 September 2016 the Government released a discussion document seeking feedback on proposals to address hybrid mismatch arrangements in line with the recommendations in Action 2 of the OECD BEPS programme (T2016/1319 IR2016/342 refers).

Submissions and subsequent meetings

20 submissions were received on the discussion document. 6 were from corporates and financial institutions, 4 were from industry bodies, 2 were from private individuals and 8 were from professional services firms. A list of submitters is included as Appendix 1 to this report.

Following the submissions, Inland Revenue and Treasury officials have:

  • met with a number of submitters to further discuss their submissions; and
  • embarked on a series of five monthly workshops with Chartered Accountants Australia New Zealand and the Corporate Taxpayers Group.

We have also been discussing hybrid issues with the Australian Tax Office, the Australian Treasury and the OECD secretariat. This report summarises the submissions received and (where relevant) our initial responses to those submissions, noting that consultation is ongoing with some of these matters yet to be covered. It also seeks your agreement to the timeframes for the remainder of the policy process.

Submissions varied significantly in responding to the proposals both in general views and specific coverage. Some submitters were supportive of New Zealand taking action in line with the OECD hybrids package, subject to various provisos including that it was done in a co-ordinated fashion with other jurisdictions and/or that there should be concessions for hybrid regulatory capital. However, a greater number were in favour of adopting a targeted or phased approach to the OECD hybrids package focused on countering hybrid arrangements that are of most concern to New Zealand.

Submissions also covered a number of specific aspects of, and general concerns with, the proposals, including the complexity of the proposals and that New Zealand should not be in the first wave of countries adopting the proposals.

Officials’ response to submissions

Nothing compelling has emerged in the course of consultation to date to suggest that full implementation of the hybrid rules generally as envisaged in the discussion document should not be pursued. We remain of the view that the proposals are likely to be in New Zealand’s best interests.

Post-discussion document developments outside New Zealand

We note that since the discussion document was released, the UK hybrid rules have come into force (1 January 2017), the EU has released a binding directive which requires EU members to expand their hybrid rules so that they apply to transactions with non-EU countries (effective 1 January 2020), and Australia remains committed to introducing hybrid rules (effective 1 January 2018 or 6 months after enactment). In all cases the rules as enacted or proposed are broadly those contained in the OECD’s Final Report on Action 2. No other jurisdictions have proposed implementing the OECD recommendations at this stage so New Zealand may well be within the first wave of adopters. However, the countries that are adopting the rules are significant for New Zealand. For instance, they are the source of approximately 62% of foreign direct investment into New Zealand.

Just before the discussion document was released, the OECD released a Public Discussion Draft titled BEPS Action 2: Branch Mismatch Structures. This document discusses cross border tax mismatches arising in the context of branches. As the title suggests, the OECD sees these mismatches as part of the hybrids project (BEPS Action 2). Accordingly, although most of these mismatches were not discussed in the Government’s discussion document, we seek your approval to consult with the original submitters on them as part of this project, and we expect that submitters will be comfortable with that.

Proposed path for development of policy and legislation

A number of submissions sought further consultation on the content of the hybrid rules (“what” rather than “whether”). Given the novel nature of the proposals and the fact that they will need to cover a wide range of situations and provisions, we agree that further consultation on their content would be useful. We have therefore agreed to conduct the workshops referred to above, which are currently scheduled to occur between now and June. This will help ensure that the proposals are implemented in a manner appropriate to the New Zealand context that minimises additional compliance and administration costs without discouraging productive foreign direct investment. We will prepare materials to facilitate discussion at the workshops, and to record their outcomes.

The timetable for these workshops was set before the date of the pre-election period was known. We therefore seek your views on whether we should:

  • keep to the current timeframe, which would involve seeking final Cabinet approval for policy decisions during the pre-election period – probably in mid-July; or
  • shorten the planned consultation timeframe so that final policy decisions can be made before 23 June. This would inevitably reduce the scope and quality of the consultation, but it would remain a useful exercise.

Officials currently consider that, if final policy decisions are made before the election (irrespective of whether this is just before or after the pre-election period commences), consulting on draft legislation over the election period would be a useful exercise. We currently envisage that the relevant legislation will form part of the first Omnibus Tax Bill following the election. Given the inevitable complexity of legislation on these issues, consultation on draft legislation would likely result in a smoother select committee process after that bill is introduced to Parliament. If you indicate you are comfortable with consultation on draft legislation, we will include a request to that effect in the Cabinet paper seeking final policy decisions.

Recommended action

We recommend that you

(a) Note the contents of this report.

Noted    Noted

(b) Agree either that:

1. officials should continue to consult with submitters on the current scheduled timelines, which would result in a Cabinet paper being prepared for submission during the pre-election period;

Agreed/Not agreed    Agreed/Not agreed

OR

2. officials shorten the current consultation timeframe so that a paper seeking final policy decisions can be considered by Cabinet before 23 June.

Agreed/Not agreed   Agreed/Not agreed

(c) Agree that officials should consult with the original submitters on the content of the OECD’s discussion draft on branch mismatch structures under Action 2 of the BEPS Action Plan.

Agreed/Not agreed   Agreed/Not agreed

(d) Agree that officials should plan to use the election period as an opportunity to consult on draft legislation.

Agreed/Not agreed   Agreed/Not agreed

Matthew Gan
Tax Specialist
The Treasury

Paul Kilford
Policy Manager
Policy and Strategy
Inland Revenue

Steven Joyce
Minister of Finance

Hon Judith Collins
Minister of Revenue

Background

1. On 6 September 2016 the Government released a Discussion Document seeking feedback on proposals to address hybrid mismatch arrangements in line with the recommendations in Action 2 of the OECD BEPS programme (T2016/1319 IR2016/342 refers).

2. This report summarises the major themes of the submissions and our responses. The submissions are generally ordered from the more general and high level to the more specific.

3. Although the expected effect of the hybrid rules will generally be to simplify commercial transactions (because they will remove the incentive to undertake transactions in a more complex tax-motivated fashion), as a technical matter, their interaction with the existing tax legislation raises an unusually large number of issues, some of them very technical. Because we are continuing to consult on those technical issues, we have not dealt with most of them here. However, they will be put before Cabinet in the process of seeking final policy approval.

General submissions

4. Submissions varied significantly in responding to the proposals. Some submitters were generally supportive of New Zealand taking action in line with the OECD hybrids package, subject to various provisos including that it was done in a co-ordinated fashion with other jurisdictions and/or that there should be concessions for hybrid regulatory capital. However, a greater number were in favour of adopting a targeted or phased approach to the OECD hybrids package focused on countering hybrid arrangements that are of most concern to New Zealand. The principal reasons for this were as follows:

  • The rules will increase the effective tax rate on inbound investment which is currently enjoying hybrid tax benefits, they will raise the required return from that investment, and therefore reduce investment in New Zealand.
  • In many cases, these negative effects will arise with no increase in New Zealand’s tax revenue. This is the case where hybrid investment into New Zealand is replaced by debt investment into New Zealand.
  • The rules may change the tax treatment of genuine commercial transactions inappropriately.
  • Our international tax rules are relatively robust and New Zealand is not as exposed as other countries to hybrid mismatches.
  • Abusive transactions can be dealt with effectively with simpler and more targeted rules.
  • New Zealand should not enact hybrid rules before international acceptance of the rules has been evidenced by enactment in other countries.

5. A submission which was made in discussions against adoption of the rules is that their effect will be to replace hybrid arrangements with funding from countries with very low tax rates, or which do not tax foreign source interest. Such arrangements will not be subject to the hybrid rules. Generally these countries are what are commonly referred to as tax havens, but they could also include more established countries with which New Zealand has a double tax treaty, such as Hong Kong and Singapore (which have territorial tax systems).

6. Officials are not convinced by the argument that the rules should not apply to inbound hybrid investment for the following reasons:

  • We consider that in some instances a disallowed hybrid instrument may be replaced with equity, resulting in higher tax payments in New Zealand.
  • Revenue may be raised even when the counterfactual investment is debt as hybrids often have a higher interest rate compared to ordinary debt.
  • Countries that are some of our most significant inward investment sources are also implementing anti-hybrid rules. The advantages of using hybrids to invest in New Zealand will be eliminated for investors from these countries, regardless of our course of action.

7. There is some force in the argument that double non-taxation, or close to it, can be achieved using debt funding through low or no tax countries, or countries with pure territorial tax systems, and enacting anti-hybrid legislation is therefore pointless. However, in most cases there will be an additional cost to routing funding through these countries, because the interest paid will be subject to 15% rather than 10% New Zealand withholding tax. Perhaps more importantly, by pushing companies into using such countries, the negative effect of low tax jurisdictions on corporate tax revenues becomes more visible than it does with hybrids. Residence countries are able to neutralise this form of tax planning using controlled foreign company rules if they wish to do so (as New Zealand already does, and as the OECD recommends). The hybrid rules will not put an end to all tax planning using cross border transactions. But they make useful progress towards that objective.

8. The remainder of this report is dedicated to the more specific submissions received and officials’ initial responses to them.

National sovereignty/loss of coherence concerns

9. A number of submitters were concerned about national sovereignty aspects of the proposals. These concerns were based on the fact that the rules mean that the New Zealand tax treatment of a cross border transaction can vary, depending on how that same transaction is treated in another country. Submissions objected to this, on the basis that it means a loss of sovereignty and that it reduces the coherence of the New Zealand tax system.

10. The first objection is without foundation. New Zealand is free to enact laws implementing the OECD hybrid rules or not as it sees fit, and is equally free to repeal them. No national sovereignty is ceded. New Zealand’s tax system already contains numerous provisions which allow foreign tax systems to affect the amount of New Zealand tax imposed on a person. Examples are the foreign tax credit rules and the rule which taxes dividends derived from foreign direct investment by New Zealanders if the dividends are deductible in another country.

11. The second objection has more substance. However, it is axiomatic that the hybrid rules will have this outcome. For the conceivable future, countries will have different rules for taxing instruments, transactions and entities. These rules will generally exhibit a high degree of coherence domestically. For example, a dividend paid by one domestic resident to another will be taxed as dividend by both parties. However, because the rules are different, they cannot be coherent in a cross border context, without some form of co-ordination such as the hybrid rules. Some payments which New Zealand treats as dividends will be treated as interest by another country, and vice versa. The effect of the hybrid rules is to introduce a different set of rules for certain cross border transactions, which increase global tax coherence by reducing double non-taxation outcomes. There is a loss of domestic consistency (since the same instrument may be taxed differently depending on the tax treatment of the foreign counterparty), but in a cross-border transaction, this is a less important concern than cross-border coherence.

12. Most importantly, the hybrid rules do not affect coherence in a purely domestic context. They will only apply in a cross border context, and only where the outcome of applying the two domestic tax systems involved produces double non-taxation. In a world where more and more business is done across borders, cross-border tax mismatches are likely to become an increasingly significant problem.

Rules will raise the cost of capital in New Zealand, in many cases without raising revenue in New Zealand

13. This submission was made by a number of submitters. It primarily focuses on the effect of the rules on inbound debt/equity hybrids, where the return on an investment in New Zealand is properly deductible in New Zealand, and properly treated as an exempt dividend or otherwise not subject to tax in the investor country. The hybrid rules will deny a deduction for the return in this case.

14. Submissions argued that this would:

  • Make no difference to New Zealand’s tax take. The theory is that all foreign firms wish to minimise their New Zealand tax. Accordingly, they all have the maximum amount of debt they are allowed under the thin capitalisation rules (60% of their gross assets). This being so, the response to the introduction of the hybrid rules will be to replace hybrid debt with ordinary debt, the return on which would generally be taxed in the other jurisdiction, but still deductible in New Zealand.
  • Push up the cost of capital in New Zealand, because it would increase the tax imposed on the return, and therefore decrease the amount investors are prepared to invest in New Zealand.

15. As to the first submission, this theory does not seem to be supported by the facts. The average debt to asset ratio of large foreign-controlled firms in Inland Revenue’s International Questionnaire database in 2015 was 26 percent (with a median of 18 percent).[1] Only 64 firms in the database, or 20 percent of the total, had a debt to asset ratio higher than 50%.

16. It does not appear to be the case that, in response to the introduction of the hybrid rules, all hybrids will be replaced with ordinary debt. Given how foreign-owned firms are currently capitalised, we consider it more likely that some portion of hybrid capital will be replaced by ordinary equity, the return on which is taxed in New Zealand.

17. Moreover, if the Government does not enact anti-hybrid rules, this could signal to the private sector that the Government has a permissive attitude towards hybrids. There is therefore a risk that some firms currently operating in New Zealand will replace part of their equity with hybrid capital because of the available tax advantages. This could have a reasonably large fiscal cost.

18. It is not necessarily the case that foreign firms do wish to minimise their New Zealand tax. The data above, showing that typical debt levels of large foreign-controlled firms are far below the New Zealand-tax minimising level of 60 percent, demonstrates this. For example, some Australian firms may prefer to pay tax in New Zealand instead of Australia. As stated in the recent Financial System Inquiry in Australia,[2] the share price of Australian firms is increasingly being set by non-resident investors, who do not benefit from franking credits, so may prefer paying tax in New Zealand, where the corporate tax rate is lower.

19. As to the second submission, submitters are correct that, in some instances, the total tax impost (i.e. New Zealand tax plus foreign tax) on investors currently using hybrids will increase. This will make New Zealand a less attractive investment location to these investors. We do not think this is a significant concern for several reasons.

20. As discussed above, we consider that if hybrid mismatches are eliminated, some hybrid capital would, be structured as equity. In these cases, the effect of the hybrid is to eliminate New Zealand tax on the investment. Neutralising hybrid mismatches will increase the total tax on investors because New Zealand tax would be payable. We do not consider that this is a problem. This treatment is in line with our general taxation settings, where we do impose a reasonable level of tax on foreign investment here. We think these settings serve New Zealand well.[3]

21. In some cases the alternative to a hybrid investment is debt, where the effect of the hybrid is to eliminate foreign tax on the investment. In this situation submitters’ concern that these changes will push up the cost of capital in New Zealand has more force, as there would be no accompanying increase in New Zealand tax payments.

22. However, in many situations submitters’ contention that these changes will push up the cost of investing in New Zealand is incorrect. Several other countries are also enacting anti-hybrid rules, including two of our largest trading partners (Australia and the UK). Collectively, these countries account for 59 percent of total FDI into New Zealand.[4] Our own enactment of anti-hybrid rules will have no impact on the total tax impost on hybrid capital originating from these countries, as the mismatch will be neutralised regardless through the primary/defensive hybrid rules structure.

23. Even when an investor is from a country that is not enacting anti-hybrid rules, and the counterfactual investment is debt, the enactment of anti-hybrid rules is not an unambiguous loss for New Zealand. Hybrid elements frequently increase the interest rate on a financial instrument, so a switch to ordinary debt may reduce interest deductions here and accordingly increase New Zealand tax payments.

24. Nevertheless it remains the case that, in some instances, the cost of investing in New Zealand will be pushed up because of this reform without any change in New Zealand tax revenues. We consider this to be a less pressing concern for New Zealand than it would be for other countries. While FDI is generally considered highly sensitive to company taxation, we argue in our inbound investment framework that tax is much less likely to play a critical factor in investment decisions into New Zealand. This is because New Zealand is an island nation, far away from the rest of the world. Much FDI here is likely to be associated with the supply of goods and services to the domestic market, which would be difficult to do without establishing a base here.

25. In any event, there are potential indirect benefits to New Zealand from eliminating the inefficiencies that result from hybrid mismatches and the associated double non-taxation. This argument is dealt with in detail in the 2016 joint Treasury/IRD paper New Zealand’s taxation framework for inbound investment. It is worth setting out the key passage here (see p21):

There are more general arguments in favour of joining a multilateral effort to remove arbitrage possibilities (which are at the heart of many BEPS issues). When companies engage in BEPS, the result is that no tax is paid anywhere on a portion of income. This clearly leads to an inefficient allocation of investment internationally as cross-border investments are subsidised relative to domestic investments. Eliminating this misallocation would increase worldwide efficiency, leading to higher worldwide incomes. The best approach for New Zealand may be to co-operate with other countries in eliminating this worldwide inefficiency in the hope of gaining its share of this extra worldwide income.

Double non-taxation reduces company taxes worldwide. While there may be arguments that in certain circumstances the cost falls on other countries, it would be naïve to suggest that the cost never falls on New Zealand. Experience suggests that once taxation is eliminated in the residence country, source country taxation is placed at risk. For example, the BEPS-induced decline in US taxation of US residents’ foreign-sourced income is often cited as a major reason for the increased focus on reducing source-country taxation by US multinationals. In that case, a general move to eliminate BEPS possibilities would make tax collections in all countries, including New Zealand, more secure and less vulnerable to unexpected tax planning.

26. Moreover, quite random reductions in tax, depending upon the opportunism of taxpayers, are likely to distort the allocation of investment in New Zealand and lead to complex arrangements that are themselves a source of inefficiency. Identifying these situations, or designing rules that turn off our anti-hybrid rules in them, would be difficult. Such an approach could also be questioned by our trading partners – the tax advantages conferred by hybrids are, by definition, not intended by either country.

27. There is a broad public concern that BEPS is unfair. Large companies escaping tax while earning substantial profits in a country has been the subject of considerable public controversy. Overall there are strong arguments for considering initiatives in this area. An important priority for the Government is considering rules to address BEPS.

28. Given all of this, we remain of the view that implementing anti-hybrid rules with a general application remains in New Zealand’s best interests.

Rules should be limited to deal with NZ-specific hybrid concerns

29. In favour of this submission, submitters pointed to the fact that the rules are relatively complex and have the potential for overreach. They said that many of the structures considered in the Final Report have not been seen in New Zealand, and therefore do not need to be counteracted.

30. We agree that the rules are complex, which is part of the reason we are conducting a series of workshops on technical aspects of the rules to minimise the risk that they reach further than they should. However, on balance, we do not think a partial approach would serve New Zealand well. The rules are a coherent package. Indications from other countries adopting the rules are that they will adopt all of them, subject only to relatively minor modifications. It will be preferable for New Zealand to do the same. This should reduce the need to make subsequent piecemeal amendments. It will also ensure our rules are internationally comparable. If an element of the rules were deliberately omitted from New Zealand’s response, this might be seen as a tacit blessing of that type of mismatch, inviting undesirable tax planning, with all the attendant risk of disputes and law changes.

31. Lastly, while New Zealand almost certainly has not experienced all of the types of transactions considered in the Final Report, there is no doubt that New Zealand taxpayers have engaged in complex cross border tax planning, and that the structures entered into would have engaged most, if not all, of the proposed rules if they had been in force.

Rules should not be enacted until more widely adopted

32. Some submitters suggested that it would not be sensible for New Zealand to be an early adopter of the hybrid rules, and that we should wait for other countries, in particular Australia, to adopt them first. It was not altogether clear in some cases why early adoption was seen as undesirable. It could be because:

  • If another country already has the rules, their adoption by New Zealand will have no impact on the taxation of hybrid arrangements between New Zealand and that country. If another country does not, then adoption of the rules by New Zealand will be the event that eliminates the tax benefit of such arrangements. This is simply an argument against the adoption of the rules, and is dealt with in the remainder of this report.
  • If another country does not have the rules, it may be a more attractive investment destination than New Zealand, at least for investment from other countries that do not have the rules.
  • Even if another country has the rules, they may not be implemented in a pure and consistent way based on the OECD recommendations and/or other countries will have other features in their overall tax regimes so that they remain internationally attractive to multinational groups. New Zealand will benefit from waiting and seeing how the rules are adopted in larger economies.
  • If New Zealand waits to introduce the rules until they are more globally adopted, businesses will be more familiar with them, and New Zealand will be perceived as less of a special case.

33. The most obvious response to this submission is that in fact, the rules are being widely adopted, and by many countries with which New Zealand has close investment links. Australia, the UK and the countries making up the EU account for approximately 62% of the direct investment into New Zealand.

34. Leaving that aside, officials note that:

  • The hybrid rules work to neutralise mismatches involving the tax base of a country that adopts them regardless of their adoption by any other country.
  • Since Australia is also adopting the rules, there is less downside, from a “favourable destination for investment” perspective for New Zealand from doing so. Adoption of the rules will not make New Zealand a less favourable destination for our largest source of direct investment, nor will it make New Zealand a less favourable investment jurisdiction than Australia.
  • In relation to certain double deduction structures involving Australia,[5] if Australia adopts the rules and New Zealand does not, that might well be to the detriment of the New Zealand tax base. This might also be the case in other situations.
  • There is no evidence that the existence of hybrid mismatches has led to any investment in New Zealand that would be at risk if they were eliminated.
  • While acknowledging it may be safer to see how the detail of the rules is implemented in other jurisdictions, there is an advantage to New Zealand in being in the leading group of adopters, particularly with Australia. New Zealand has the chance to have some influence in how the rules are implemented around the world, we are able to benefit from engaging with other countries who are also actively engaged in developing their rules (particularly the case with Australia) and it may also prove possible to introduce our rules in a more co-ordinated fashion with Australia.

35. We note that since the discussion document was released, the UK hybrid rules have come into force (1 January 2017), the EU has released a binding directive which requires EU members to expand their hybrid rules so that they apply to transactions with non-EU countries (effective 1 January 2020), and Australia remains committed to introducing hybrid rules (effective 1 January 2018 or 6 months after enactment). In all cases the rules as enacted or proposed are broadly those contained in the OECD’s Action 2. No other jurisdictions have proposed implementing the OECD recommendations for hybrids at this stage.

Compliance cost concerns

36. There is no doubt that the need to comply with the hybrid rules will involve some additional cost for business. There will be the initial cost of helping to develop and understanding the rules, and then the cost of ensuring that they are complied with. However, in the vast majority of cases, compliance costs can be minimised by not entering into, or unwinding, hybrid structures, and replacing them with structures that in most cases are commercially much simpler and cheaper, albeit less tax effective.

37. The imported mismatch rule was a particular target of this submission. We appreciate the concerns raised here as the imported mismatch rule applies where the hybrid mismatch does not directly involve New Zealand, and serves as an integrity measure for some of the other OECD recommendations for hybrids. Prima facie it will require corporate groups to identify hybrid mismatches which are not subject to direct counteraction, and then to determine how the benefit of such mismatches should be apportioned between payments that are subject to the imported mismatch rule, which could involve multiple jurisdictions.

38. The cost of compliance with this rule is reduced by the fact that it does not apply to payments made to other countries which have hybrid rules. Accordingly it will not apply to payments by a New Zealand resident to (for example) an Australian one (assuming Australia adopts the rules), and adoption of the rules in New Zealand will ensure that compliance costs in respect of this rule do not arise in respect of payments from countries which have hybrid rules to New Zealand. If New Zealand is within the first wave of adopters, there will be additional compliance costs as it waits for other countries to come on board, particularly so if New Zealand introduces the rules before Australia (although this is not expected on current timelines).

39. We envisage that as a practical matter, the kinds of multinational groups where imported mismatches might conceivably apply will employ skilled tax managers, one of whose tasks will be to review the existence of mismatch arrangements throughout the group. The introduction of hybrid rules by other countries means that this task will be required to be performed regardless of whether or not New Zealand has hybrid rules.

40. The workshops with submitters on technical aspects of the rules are intended to ensure that compliance and administrative costs do not become an undue burden on businesses or Inland Revenue. Imported mismatches will be covered through the upcoming workshops and we will report back on this matter as part of final policy recommendations.

Effect on taxation of foreign branches of New Zealand companies

41. A number of submitters argued that the adoption of recommendation 6, denying the ability of New Zealand companies to use foreign branch losses to reduce New Zealand taxable income, meant that New Zealand should revisit its current system of taxing the business income of foreign branches. Indeed, some submissions argued that an exemption for active income of a foreign branch should be enacted regardless of whether the hybrid rules proceed.

42. This proposal was considered and rejected in an earlier Cabinet paper (T2013/2166 PAS2013/162). If the hybrid rules were to proceed on the basis set out in the preceding paragraph, there would be a good argument in favour of revisiting the proposal.

43. However, the OECD is in the process of modifying its published position on this point. It now recommends that foreign branch losses be non-deductible in the head office country only if the losses are used in the branch country to reduce the tax on income which is not taxed in the head office company. For many ordinary businesses, this will not be the case, and therefore the adoption of the modified recommendation 6 will not affect their ability to use their foreign losses against New Zealand taxable income.

44. Accordingly, we recommend at this stage that the hybrid project proceed without considering the general tax treatment of foreign branch income and losses. However, this issue could be considered in the near future if the Government wished to revisit it, subject to resource constraints.

Branch mismatches

45. On 22 August, the OECD released a Public Discussion Draft titled BEPS Action 2: Branch Mismatch Structures. This document discusses cross border tax mismatches arising in the context of branches. As the title suggests, the OECD sees these mismatches as part of the hybrids project (BEPS Action 2).

46. The branch mismatches in the Discussion Draft are analogous to those considered in the Government’s September discussion document. They involve deductible/non-includible, double deduction, or imported mismatches. The only difference is that these mismatches arise because of differences in countries’ rules for taxing branch income, rather than because of differences in how countries tax entities or instruments.

47. The counteractions proposed in the Discussion Draft are also of the same nature as those proposed in relation to hybrid mismatches.

48. Officials are of the view that these mismatches also need to be considered. We note that the UK’s hybrid rules deal with branch mismatches, and that UK officials have stated that without this, the rules would have been much easier to circumvent.

49. Officials believe it will be much better to consider branch mismatches in the context of the current consultation. Accordingly, we propose to discuss them at our meetings in mid/late March with the Corporate Taxpayer Group and Chartered Accountants Australia and New Zealand. We will also approach the New Zealand Law Society and the New Zealand Bankers' Association and offer them the opportunity to consider the branch mismatch issues. Apart from some individuals who submitted on the Government discussion document, these four groups represent all of the original submitters. We expect submitters will be comfortable with this approach.

De minimis rule

50. Four submissions supported a de minimis rule either generally, or specifically for the imported mismatch rule, to reduce compliance costs. One supported consideration of a general de minimis rule, but was concerned about possible complexity in the rules to show that that the de minimis could be relied upon. The submitter noted the imported mismatch rule (recommendation 8) as a place where a de minimis might be particularly useful. Two suggested a general de minimis so the rules would only be targeted at higher value transactions (e.g. $1m of relevant income/expenditure) or would not apply to smaller taxpayers (e.g. turnover under $80m). The fourth was supportive of a de minimis rule for the imported mismatch rule.

51. Other than where the rules apply to timing mismatches, officials do not support a de minimis at this stage, but this will be discussed further as part of the private sector workshops. The OECD Final Report does not have a de minimis. For many of the rules, e.g. recommendations 3 and 6, it would be very complex to have a de minimis based on transaction size which could not also be abused. The issue of size is partly resolved by observing that the rules only apply to taxpayers entering into more complex cross border transactions. Within that context, even smaller taxpayers can be expected to understand how they are taxed in the countries in which they operate, recognizing that in most cases the rules only operate within control groups or related parties.

52. In relation to the imported mismatch rule, presumably the de minimis would generally be based on the level of a New Zealand company’s interest expense. This could be a workable rule, though if a group has operations in, for example, the UK, the group will have identified possible imported mismatches in any event.

53. Officials will keep this matter under review, with a particular eye to whatever measures are adopted in Australia.

Regulatory capital (banks and insurance companies)

54. A number of submissions addressed the question of whether regulatory capital required to be issued by banks and insurance companies should be subject to the hybrid regime. Capital adequacy regulations may lead, if not inevitably then without apparent effort, to the issue of cross border hybrid arrangements. A significant amount of hybrid capital has been issued by the New Zealand branches of the Australian trading banks with New Zealand operations. This capital is hybrid because it is treated as debt by New Zealand (and is therefore deductible) and as equity by Australia (and can therefore have franking credits attached to it, which can reduce or eliminate an Australian holder’s Australian tax liability on the dividend).

55. Submissions made the following arguments:

  • Deductible/frankable instruments should not be regarded as hybrids at all, because franking credits are a limited resource and represent tax actually paid in Australia.
  • Bank regulatory capital should be excluded from the hybrid rules, given that the legal terms which give rise to its hybridity (subordination to other debt, conversion to equity in the case of distress etc.) are often the result of regulatory requirements.
  • Bank regulatory capital should be excluded from the hybrid rules because it is economically important.
  • Banks do not have a choice as to whether or not to attach franking credits to the return paid on deductible/frankable instruments – attachment of credits is a requirement of Australian law. Accordingly it would not be appropriate to apply the hybrid rules to the tax treatment of that return.
  • Bank regulatory capital should be excluded from the hybrid rules because the effect of including it will be that it is replaced with debt having a higher rate of interest, which would reduce the New Zealand tax base. Currently the hybrid debt has a lower rate of interest, because the third party lenders (mostly Australian individuals and investment entities) are prepared to take a lower cash return given that they receive franking credits on top of the interest.
  • Bank regulatory capital should be grandparented if issued before a certain date, especially because it is often publicly issued and refinancing it would not be straightforward.

56. With the exception of the sixth point, and possibly the fifth point, officials do not believe that these submissions have much force for the following reasons:

  • In relation to deductible/frankable instruments, there is a hybrid mismatch. It is true that the nature of the imputation system means that this mismatch cannot be inexhaustibly tapped. However, so long as a company is not otherwise distributing all of its tax-paid profits (which is relatively unusual), the mismatch can be profited from in a similar way as a deductible/non-assessable mismatch.
  • As to the second and third submission, applying the hybrid rules to bank regulatory capital does not prevent such capital being issued cross border. It simply ensures that it does not enjoy an unintended tax benefit. The banking regulators have no interest in whether regulatory capital gives rise to tax benefits or not, and the removal of such benefits will in no way undermine their work. The result of applying the hybrid rules is simply that regulatory capital instruments are treated the same way as they would if both the issuer country and the investor country had the same tax rules.
  • As to the fourth submission, the Australian requirement to attach franking credits to the distribution is entirely consistent with New Zealand denying a deduction for the distribution. The distribution will then be treated as a payment of a dividend by both Australia and New Zealand. There is more of a difficulty for Australia in determining how it should apply the hybrid rules. It might, for example, be difficult for Australia both to require a franking credit to be attached to the dividend, and deny the shareholder a credit for that imputation credit. However, this is an issue for Australia to determine.
  • As to the fifth submission:
    • When the deductible/frankable instruments were first issued, to some extent they did replace equity, rather than debt, financing. It is possible that if they are cancelled, they will be partly replaced with equity.
    • The banks issuing these instruments are currently operating with more equity than “required” by the thin capitalisation rules, and accordingly the hypothesis that they will always minimise their New Zealand equity is not correct.
    • Because they bear a higher risk, deductible/frankable instruments have a higher funding cost than ordinary debt. This helps counter the rate reduction achieved by the tax arbitrage.
    • Not all deductible/frankable instruments are issued to third parties. There are also structures involving intra-group issuances which support third party issuances. The return payable on some instruments issued in these structures have been sufficiently high to raise transfer pricing concerns.

However, in the event that Australia decides not to deal with the treatment of deductible/frankable regulatory capital, officials would wish to consider more closely the effect on the New Zealand tax base of applying the hybrid rules to it.

57. New Zealand’s stance in this matter will not be relevant if Australia acts, in accordance with hybrids recommendation 2, to tax the return on deductible/frankable instruments as interest (and therefore not frankable). A decision on this has been expected for some time, but has not yet been made. Officials intend to meet with submitters on this point once the Australian decision on regulatory capital has been announced. Officials will also consult with the Reserve Bank of New Zealand before final policy recommendations are made.

58. Officials are sympathetic to the arguments in favour of grandparenting regulatory capital issued before the release of the Government discussion document on 6 September 2016. The finer details of grandparenting are likely to be clearer in our next report on this matter. Officials also hope that any grandparenting will be co-ordinated and consistent with whatever decision is made by Australia.

Effect on New Zealand foreign trusts

59. The discussion document on hybrid mismatch arrangements proposes to apply the hybrid rules to tax the foreign source income of New Zealand foreign trusts (i.e. trusts with a New Zealand trustee but no New Zealand settlor) if that income is not being taxed to:

  • the beneficiary, in the case of beneficiary income;
  • the settlor, in the case of trustee income,

in the beneficiary/settlor’s country of residence, if the non-taxation in the residence country is the result of the trust being a reverse hybrid. We received one submission in favour of this proposal and 7 not in favour.

60. Many of the submissions referred to the fact that the 2016 report of the Government Inquiry into Foreign Trust Disclosure Rules (the Shewan Report) recommended no change to the current trust tax regime. They argued that this recommendation supported not applying hybrids recommendation 5.2 to New Zealand foreign trusts. However:

  • the terms of reference of the Inquiry were directed solely at trust record keeping and disclosure, not trust taxation;
  • the comments which were made in the Shewan Report on trust taxation were not made having any regard to the double non-taxation issue which is central to the hybrid rules. The Shewan Report was considering the appropriateness of the trust tax regime as a matter of the coherence of the New Zealand tax base on a stand-alone basis. It correctly stated that the non-taxation of non-residents on non-New Zealand source income was and remains orthodox international tax policy (paragraph 4.15). Leaving aside the hybrid rules, the Report recognized that it is reasonable for New Zealand not to tax such income when it is derived by the New Zealand resident trustee of a trust with no New Zealand settlor. However, the Shewan Report did not consider the impact of BEPS Action 2 on international tax policy. In particular, it did not consider whether a New Zealand foreign trust is a reverse hybrid (as defined for purposes of the hybrid rules) and whether, if it is, New Zealand tax should be imposed on its income.

61. The Inquiry concluded that the changes to trust disclosure rules and practice which it recommended would deal adequately with the problems identified, including reputational risk. It was not asked to, nor did it, consider the problem of double non-taxation, and accordingly no conclusions were reached or recommendations made that are relevant to this issue.

62. Officials have discussed the preceding paragraphs with John Shewan, who agrees that we have accurately reflected the scope and conclusions of his report.

63. Submissions against our proposal raised a number of other points which, along with our responses, are summarised below.

  • A New Zealand trust is not a reverse hybrid, since it is not tax transparent in New Zealand. That is because the trustee is always taxable on New Zealand source trustee income. This submission is not correct. An entity can be partially tax transparent. The Final Report states (paragraph 140) that a person will be treated as tax transparent in respect of a payment where the reverse hybrid attributes or allocates a payment that it has received to an investor, and the effect of such attribution or allocation is to treat the payment as it would have been treated had it been paid directly to the investor. Clearly then a New Zealand foreign trust is tax transparent with respect to income allocated to beneficiaries. As the Government’s discussion document stated, this is not the case for trustee income. The argument for New Zealand taxation of foreign source income in that situation rests on the New Zealand principle that the residence of the settlor determines whether or not the trust’s foreign source income is subject to New Zealand tax.
  • The current tax treatment of New Zealand foreign trusts is appropriate and should be maintained. We agree that the current tax treatment makes sense on a single jurisdiction basis. However, it can lead to double non-taxation as a result of a hybrid mismatch. If the jurisdiction of the settlor (in the case of trustee income) or the beneficiaries (in the case of beneficiary income) would ordinarily tax the income, but is not doing so because it regards the income as derived by the New Zealand trustee, then there is double non-taxation as a result of differing treatments of the trust. This is an issue which the hybrid rules are trying to address.
  • Determining whether a New Zealand foreign trust and the settlor are in the same “control group” is not possible. Some work may need to be done in this respect, but our current associated person rules will provide a good foundation.
  • The fact that the foreign settlor of a trust is not taxed in another country on foreign source trustee income does not justify New Zealand taxing the New Zealand trustee on that income. The settlor is not taxed because the income is not theirs. Indeed, the settlor may be dead when the income is derived. This submission ignores the fact that the basis for New Zealand not taxing this income is the residence of the settlor. The logical extension of this approach is that the settlor’s residence country should regard the income as that of the settlor.
  • It might be more logical to tax the New Zealand trustee on its foreign source trustee income if it can be determined that the beneficiaries who will receive the income are not taxable on the income. However, in a discretionary trust this is not possible, because it is not known which beneficiary will receive the income. We do not agree that the foreign tax treatment of the settlor should not be taken into account. We do agree though that, if it can be shown that a beneficiary is subject to tax on foreign source trustee income, that would mean there is no hybrid mismatch to counteract in New Zealand.
  • Non-taxation of the settlor or a beneficiary might be a result of the person being resident in a country with no income tax, or with a territorial tax system, rather than because the person is resident in a jurisdiction which treats the income as derived by the trustee. We agree that in this case, double non-taxation is not a result of a hybrid mismatch, and there should be no hybrid counteraction. Applying the rules only to cases where mismatches result from hybridity will be required generally, and this is not a special case.
  • Difficulties would arise from the proposal where the trust holds foreign shares which are taxed under the fair divided rate (FDR) method. The FDR method is unique to New Zealand. Since the settlor residence country will not be taxing on the same basis, the reason that the trust’s income (calculated under New Zealand law) is not taxable in the settlor country is that that country will not see the settlor as having such income. We agree that differences of this kind will create complexity. However, this complexity will be an issue for other aspects of the rules as well, and will need to be dealt with. As a matter of principle, the issue would be resolved by determining, in the year the New Zealand income is derived whether the beneficiary or trustee would be taxed on that income if it also arose in their residence jurisdiction.
  • Compliance costs from this proposal would be substantial and in most cases no tax would be generated. There would no doubt be some compliance costs from this proposal, though in the main they would revolve around the need to communicate information already known by one of the parties or its advisors to the other parties or advisors. An assumption that this sort of “intra-group” communication is possible underpins many of the proposed rules and we do not consider there are any good reasons for treating trusts differently. It is possible that no or little New Zealand tax would be generated. However, the proposals would assist in the general move towards shutting down hybrid-mismatch-driven double non-taxation, the aim of which is to increase global tax revenues.
  • The proposal goes well beyond the ambit of the hybrid proposal, and appears to be advanced to support New Zealand tax applying when none would ever arise apart from the existence of a New Zealand resident trustee. This submission misses the point that the effect of the rules will only be to rectify double non-taxation arising from the existence of the New Zealand resident trustee. For example, if the settlor would not be taxed on the foreign source income if it derived the income directly, there is no suggestion that the hybrid rules would impose New Zealand taxation on that income if derived by a New Zealand tax resident trustee on whom the settlor has settled the income producing property.

64. Since New Zealand trusts can give rise to double taxation due to hybrid mismatches, we intend to continue to develop our proposals in this respect, though we agree that applying New Zealand tax to trustee income on the basis of the existence of a hybrid mismatch will present some challenges.

Interaction with withholding tax

65. A number of submissions were concerned by the proposal in the Discussion Document that where an interest payment is disallowed under the hybrid rules, withholding tax would still be imposed as if that payment were interest. They were concerned that this would constitute double taxation. If the payment were treated as a dividend and fully imputed, it would not be subject to withholding tax.

66. This is a complex issue. It may well be the case, at least with respect to recommendations 1 and 3, that ideally, the withholding tax treatment would match the deductibility status. However, we have assumed that this might not always be possible, particularly if the payer is not aware at the time of making the payment that the hybrid rules deny a deduction. Officials are continuing to review this question.

Transitional

67. The OECD Final Report recommended no grandparenting of existing transactions and an effective date based on the tax year which gives taxpayers sufficient time to restructure their transactions, which will all be either between related parties or within a control group.

68. The Discussion Document proposed to follow this approach, with the effective date being the first balance date after enactment of the legislation.

69. A number of submissions were in favour of grandparenting existing transactions, either without limit, or for 3-5 years. This was on the basis that taxpayers should be entitled to retain tax benefits that existed when transactions were entered into.

70. As set out in paragraph 58, we now propose to consider limited grandparenting for frankable/deductible instruments issued before the release of the Discussion Document. Consistent with the OECD, and the approach in the UK, we do not propose any further grandparenting. The hybrid rules are doing no more than removing tax benefits which were not, in aggregate, intended by either country, and restoring a more “normal” result. The hybrid rules apply generally to transactions between related parties, which can generally be undone with relative ease. If there are any situations where this is not so, as with the frankable/deductibles, they have not emerged in consultation to date. If any emerge, we can consider them on a case by case basis.

71. Many submitters supported New Zealand having the same effective date for the rules as Australia. Others submitted that New Zealand should not enact its rules until after the Australian rules have become effective, in order to give more clarity.

72. Officials are sympathetic to the submissions for co-ordination on this point with Australia, but continue to believe at this stage that the core proposal (effective for balance dates after enactment) is a reasonable one.

Consultation process

73. A number of submissions sought further consultation on the proposals. Officials are sympathetic to these submissions, and are currently engaged in a consultation programme that is scheduled to take place between now and June. As mentioned above, we anticipate this process will also include consultation on the OECD’s discussion draft on branch mismatches. This timeline could be altered to allow for final Cabinet decisions to be made before the pre-election period commences on 23 June. However, this would inevitably reduce the scope and/or depth of the consultation. The private sector may also be disappointed by such a change. We welcome your views on the timing of this process.

74. Some submitters also sought the opportunity to review draft legislation. Draft legislation was released for comment in the UK, and Australia is likely to do the same. Officials currently consider that, if final policy decisions are made before the election (irrespective of whether this is just before or after the pre-election period commences), consulting on draft legislation over the election period would be a useful exercise. Given the inevitable complexity of legislation on these issues, consultation on draft legislation would likely result in a smoother select committee process when final legislation is introduced to Parliament. If you indicate you are comfortable with consultation on draft legislation, we will include a request to that effect in the Cabinet paper seeking final policy decisions.

Appendix 1: List of submitters to Government discussion document Addressing Hybrid Mismatch Arrangements

Andrea Black
ANZ Bank New Zealand Limited
ASB Bank Limited
Bank of New Zealand
Baucher Consulting Limited
Chapman Tripp
Chartered Accountants Australia and New Zealand (CA ANZ)
Corporate Taxpayers Group (CTG)
Deloitte
Ernst & Young Limited
Fisher & Paykel Healthcare Limited
Insurance Australia Group Limited (IAG)
JJL Hoogenboom
KPMG
New Zealand Bankers’ Association (NZBA)
New Zealand Law Society
Olivershaw Limited
PricewaterhouseCoopers
Russell McVeagh
Westpac New Zealand Limited

 

1 Based on the International Questionnaire for the 2015 income year, which has data on all foreign-controlled firms (excluding banks) with turnover of more than $80m – 314 firms in total.

2 Financial System Inquiry Final Report (2014), available from http://fsi.gov.au/files/2014/12/FSI_Final_Report_Consolidated20141210.pdf

3 The reasons for this are discussed in the joint IRD/Treasury paper New Zealand’s taxation framework for inbound investment (June 2016).

4 Based on 2016 data on the stock of direct investment by country, from Statistics New Zealand.

5 Double deduction Australian limited partnership structures.