Chapter 4 - Current problems with the rules and proposed solutions
Comments applying to all the proposals in this paper
Costs and benefits
4.1 Every proposal in this issues paper will, in a minority of cases, reduce the amount of interest deduction that may be taken. In that way, it will also reduce returns for some foreign investors. Some investors might pull out of New Zealand. An important question is whether or not the potential withdrawal of investment is outweighed by some additional tax revenue.
4.2 The McLeod tax review (“Tax Review 2001”) considered the taxation of foreign investment into New Zealand. The review found that some reasonable level of tax could be imposed on foreign investors without unduly affecting incentives to invest. This is in New Zealand’s best interests.
4.3 However, there is some level at which additional taxes will cause more harm than good. We are therefore interested in your views about the costs of the proposals for taxpayers and investors.
Application and transition
4.4 If implemented, we propose the changes would all be applied from the first income year beginning after enactment of the legislation.
4.5 We expect that some investors will change their mix of debt and equity funding accordingly.
4.6 We have attempted to design the proposals so that genuinely external debt funding, which could be difficult to alter, will not usually be affected. This is likely to reduce the effect of changes to the mix.
4.7 Nevertheless, there could be constraints on the ability of companies to substitute equity for related-party debt. We are interested in submissions on this point.
Questions we would like you to consider
|
Problem 1: Single non-resident requirement
4.8 The inbound thin capitalisation rules apply to a resident company only if the company is controlled by a single non-resident (section FE 2(1)(c)).
4.9 The case of a single controlling investor is the most obvious case in which a shareholder can arbitrarily determine the level of debt in the company, unless prevented by a thin capitalisation rule. Subject to any non-tax constraints that might be imposed by, for example, an unrelated lender, a controlling shareholder can classify their investment as either debt or as equity, or any mixture of the two.
4.10 However, there are other cases in which more than one shareholder can explicitly agree between themselves to set the level of debt in a company.
4.11 The risks of New Zealand-sourced income escaping New Zealand tax seem very much the same in these cases as in the single non-resident shareholder case, and this suggests widening the scope of the existing rules.
4.12 At the other extreme from the case of a single non-resident controller, a single shareholder among many shareholders of a listed company will be unable to control the level of debt in the company. It is also very unlikely that the shareholders would explicitly agree between themselves to manipulate the level of debt unless this process was initiated by the company itself, in which case the stapled-stock rule in section FA 2B might well apply.
Proposal: Single non-resident requirement
4.13 We propose that the inbound thin capitalisation rules be widened to include companies in which non-residents who are not necessarily associated persons according to the tax definition, but who act together, hold an interest of 50% or more.
4.14 The intention of the change would be to make companies subject to the rules when shareholders can collectively act in the same way as an individual controlling shareholder to determine the level of debt the company will hold.
4.15 What it means to “act together” would not be defined exhaustively in legislation.
4.16 However, acting together will include at least:
- explicitly co-operating with each other through a written or tacit shareholder agreement;
- being effectively co-ordinated by a person or group of people, such as a private equity manager or managers.
Example: James, Henry and Judith each hold one third of the shares in NZ Co and they have a shareholder agreement which determines what they may do with their holdings and regulates certain other actions. The three non-residents are “acting together”. Example: An investment promoter contacts ten people and convinces them to form a consortium to buy out NZ Co. The promoter arranges for a partnership to be legally documented and the ten people contribute their capital, which is used to purchase all of NZ Co’s shares. The promoter effectively administers the partnership and directs its activities. The ten investors are “acting together”. Example: X Co, which manages investments on behalf of foreign investment partnerships, arranges for two partnerships (not legally associated) to incorporate companies. The newly incorporated companies each purchase 49% of the shares in NZ Co. X Co holds the remaining 1%. The companies are “acting together”. |
4.17 We would not expect portfolio shareholders of a public company to be acting together and would look to clarify that in any legislation.
4.18 Widening the application of the thin capitalisation rules in this way raises questions about the definition of the “worldwide group” of a company. When there is a single non-resident controller, the worldwide group is the group of the controller. When there is no single controller, it is not clear whose worldwide assets and liabilities should be included in the worldwide group, or to what extent those assets and liabilities should be included. If a bank was one of the parties acting together, there could be further complications because registered banks are subject to separate thin capitalisation rules.
4.19 Therefore, and to avoid significantly increased complexity of the rules, we propose that when there is no single non-resident controller, the worldwide group for the purposes of the 110% safe harbour will be the New Zealand group. On its own, this would negate the extension of the rules to people acting together, because the New Zealand group’s debt-to-asset ratio would always be less than 110% of the worldwide group’s ratio. Further measures to overcome this problem are outlined as part of other proposed changes in this paper.
4.20 An alternative to widening the rules to cover groups acting together would be to widen the rules to cover all companies in which non-residents hold an interest of 50% or more. Using this approach, defining “acting together” would be unnecessary as acting together would not be a consideration. However, such a test could bring in shareholders that are very unlikely to be able to influence the level of debt held by the company (publicly listed companies would need to be explicitly excluded).
4.21 Another alternative approach would be to prescribe specific and exhaustive criteria which would indicate when people would be considered to be acting together. For example, the thin capitalisation rules might apply whenever there was a shareholder agreement. Such an approach is not favoured because of the risk that the specific criteria would be easily circumvented unless there were complex criteria to prevent it.
Questions we would like you to consider
|
Problem 2: Worldwide group is NZ group and worldwide debt is shareholder debt
4.22 Interest is not denied under the thin capitalisation rules if the debt-to-asset ratio of the New Zealand group of a company or trustee is less than 110% of the debt-to-asset ratio of the corresponding worldwide group (section FE 6).
4.23 In some cases the worldwide group is almost the same as the New Zealand group, and so the 110% safe harbour is always available. This allows the non-resident investor to use as much debt, including shareholder debt, as they wish. This is inappropriate.
Example: Jan incorporates SPV Co in the Cayman Islands, capitalising it with $90 of shareholder debt and $11 of equity. SPV Co buys NZ Co, capitalising it (using its own funds) with $90 of debt and $10 of equity. The worldwide debt-to-asset ratio is 90.1%. The New Zealand debt-to-asset ratio is 90%, so no interest deductions will be denied. |
4.24 The 110% safe harbour contemplates a foreign parent with substantial foreign operations and third-party borrowing, in which case the comparison to a worldwide ratio is more meaningful.
Proposal: Worldwide group is NZ group or worldwide debt is shareholder debt
4.25 We propose that when determining the worldwide ratio in the inbound rules, debt will not be counted if it is linked to shareholders of group entities.
Example: Jan (see previous example) cannot include shareholder debt when calculating the worldwide debt-to-asset ratio. The worldwide debt-to-asset ratio is therefore 0%. Example: Aus Co owns NZ Co. Aus Co borrows $100 on commercial terms from an unrelated bank. Aus Co has $150 of assets, which is the sole security for the loan. NZ Co has $75 of assets and $52 of shareholder debt. The worldwide debt-to-asset ratio is 67%. The New Zealand debt-to-asset ratio is 69%, so the 110% safe harbour applies. |
4.26 The intention of the proposal is that the worldwide group test will be used only when it is meaningful – that is, when New Zealand debt levels are being compared to debt levels of a multinational organisation with genuinely external funding.
4.27 Where funding is not genuinely external, and so could effectively be a substitute for equity, there is a reasonable case for using thin capitalisation rules to alter the mix of debt and equity funding. This prevents foreign investors from unduly reducing the effective New Zealand tax rate.
4.28 In contrast, when debt is from truly unrelated parties the shareholder does not benefit from a lower effective tax rate and there is not such a strong case for preventing interest deductions.
4.29 Determining the particular conditions for linking a debt to shareholders is a complex matter. However, our preliminary views are that, we might exclude debt from the worldwide ratio in the following cases:
- the debt is owed to a person having an income interest in a group entity; or
- the debt is owed to a person that has received funds directly or indirectly from a group entity, a person with an income interest in a group entity, or an associated person, on the condition or expectation of both parties that some or all of those funds would be used to provide the debt.
4.30 The second condition is to prevent back-to-back loans.
4.31 It might also make sense to exclude a debt if security for the debt or a guarantee of repayment comes from outside the worldwide group. These features might indicate that debt has been artificially increased to a level that would not be sustainable if the worldwide group was standing alone. Our instinct is that such an exclusion would be necessary to make the proposal effective, but we are seeking your feedback about cases in which this would not be appropriate.
Example: Cayman Co and Bahama Co each own 40% of NZ Co. They also each own 47% of Tiger Co, which is not part of the worldwide group for thin capitalisation purposes. Tiger Co has $100 of assets. NZ Co has $20 worth of assets. NZ Co borrows $20 from a bank. The loan is secured over NZ Co’s asset and Tiger Co’s assets. The loan may not be included in the worldwide group’s debt for the purposes of the debt-to-asset ratio. |
4.32 Because it is difficult to anticipate all the ways in which shareholder debt might be transformed into apparently external debt, a specific anti-avoidance rule might be required in addition to the listed exclusions.
4.33 The proposal we are making is an additional restriction on the availability of interest deductions. All the existing restrictions continue to apply, including the inclusion of all debt – from related or unrelated sources – in the New Zealand group calculations.
4.34 Overall, the proposal appears to be superior to alternatives we have considered. A real difficulty with the current 110% safe harbour is that it appears to be suitable only for large multinational groups with considerable holdings outside New Zealand and ultimately public shareholdings.
4.35 One option would have been to prohibit the use of the 110% worldwide group test if less than 50% of the worldwide group’s assets were outside New Zealand or the New Zealand group was not controlled by a single non-resident. This would have been a simpler option to administer. However, it would also have arbitrarily denied interest deductions in rare cases of genuinely high levels of external gearing. External borrowing does not reduce the effective New Zealand tax rate for a foreign equity investor, so there is not a strong theoretical case for forcing a reduction of gearing in these cases.
4.36 Abroad, other countries have adopted an “arm’s-length” test, which would allow debt deductions in these sorts of cases as long as it could be demonstrated that an unrelated (non-shareholding) party would be prepared to advance the same amount of debt as has been actually taken on by the domestic operation. In practice, it is difficult to decide what an arm’s-length level of debt is, and we do not favour such an approach. We are also concerned that, in practice, an arm’s-length test is ineffective in preventing excessive interest deductions. We also observe that a 60% debt-to-asset ratio is unusually high for normal multinational businesses.[1}
4.37 The proposal we are making will increase compliance costs for some taxpayers. However, the number of people affected is likely to be small. In practice, and particularly since the introduction of specialised thin capitalisation rules for registered banks, the 110% test is relied upon in few cases.
Questions we would like you to consider
|
Problem 3: Trusts as main shareholder (non-resident with effective control)
4.38 The thin capitalisation rules apply to a trustee of a non-complying trust settled by a non-resident if the non-resident has made more than 50% of the total settlements on the trust (subsection FE 2(1)(d)).
4.39 They do not apply to other resident trustees. This causes problems in practice.
4.40 For example, suppose a New Zealand business is owned by a resident trustee of a trust, and that the trust has been settled by a resident subsidiary of a foreign company. If the trustee has complied with New Zealand tax obligations, it is a complying trust. The settlor is a special purpose vehicle, has few assets and no interest expenses. The trustee borrows most of its capital directly from the foreign company, but is not subject to the thin capitalisation rules so has unrestricted interest deductions.
4.41 On the face of it, the rules should apply to trustees more broadly than they currently do.
Poposal: Trusts as main shareholder (non-resident with effective control)
4.42 We propose that the thin capitalisation rules will apply to a resident trustee if 50% or more of settlements made on the trust have been made by a non-resident or a group of non-residents acting together, or by an entity that is subject to the inbound thin capitalisation rules (the rules already apply to any non-resident trustee).
4.43 The intention of the proposal is to prevent a trustee from borrowing on behalf of a group to avoid the application of the thin capitalisation rules.
4.44 It is noted that in the case of a foreign trust earning only foreign-sourced income, there is unlikely to be any interest deduction in the first place, and so the application of the thin capitalisation rules would be of no consequence.
4.45 We expect this proposal to have minimal effect on legitimate business structures.
Questions we would like you to consider
|
Problem 4: Problems with individuals or trustees holding NZ groups
4.46 Section FE 3(2) excludes, from the New Zealand group of a resident natural person or trustee, an excess debt outbound company or a member of the New Zealand group of an excess debt outbound company. This means that a natural person or trustee may be considered separately from companies, including CFCs, they own, and not be required to consolidate those interests.
4.47 This is in contrast to section FE 3(1), which ensures that a natural person or trustee must include all associated residents, even companies, in their New Zealand group, and must consolidate the debts and assets of those entities with those of the natural person or trustee (see also section FE 14(2)).
4.48 Section FE 3(1) applies to inbound investors, while section FE 3(2) applies to outbound investors, but there is no obvious basis for distinguishing between the two when it comes to consolidation of companies and non-companies.
Proposal: Problems with individuals or trustees holding NZ groups
4.49 We propose that the New Zealand group of an individual or trustee under section FE 3(2)(a) should no longer exclude excess debt outbound companies or those included in a New Zealand group of an excess debt outbound company.
4.50 The intention of the proposal is to require consolidation of an individual’s or trustee’s holdings with holdings of underlying entities, just as is required under section FE 3(1)(a).
Questions we would like you to consider
|
Problem 5: Capitalised interest
4.51 Recall that the debt-to-asset ratio of a group determines whether or not the thin capitalisation rules will limit interest deductions (section FE 6).
4.52 Asset values are determined according to generally accepted accounting practice. In at least the case of International Financial Reporting Standards, this means that asset values must sometimes include capitalised interest costs (NZ IAS 23).
4.53 Where a tax deduction has been taken for these interest costs, which is commonly the case, they should in theory be excluded from the asset value for thin capitalisation purposes.
Proposal: Capitalised interest
4.54 We propose that capitalised interest expenses should be deducted from asset values, to the extent the expenses have been deducted for New Zealand tax purposes.
4.55 Where a New Zealand tax deduction has been allowed for interest that is capitalised for accounting purposes, it is not appropriate that this be included in the New Zealand group assets.
4.56 Some compliance costs will be imposed by this proposal through having to separate out capitalised interest expenses. However, in previous consultation on capitalised interest expenses – in the context of the financial arrangement rules – we have been advised that it is a straightforward exercise to identify capitalised interest costs. We therefore believe compliance costs should be minimal.
Questions we would like you to consider
|
Problem 6: Asset “uplift”
4.57 As noted in the previous section, the debt-to-asset ratio of a group determines whether or not the thin capitalisation rules will limit interest deductions (section FE 6). Asset values are determined according to generally accepted accounting practice.
4.58 For particular assets, including much intangible property, this means assets will be recorded at cost (less amortisation or impairment). Revaluation of such assets, including internally generated goodwill, is not permitted because a reliable value cannot be determined. In contrast, when money is paid to an unrelated party for an intangible asset, including goodwill, the amount paid is likely to be a more reliable indication of actual value.
4.59 The restriction on revaluations is being circumvented by some groups who report increased asset values following internal reorganisations. In these cases, existing subsidiaries are sold to other group companies for market value, which includes the value of internally generated intangibles. It is questionable whether or not this is permitted under generally accepted accounting practice. In any case, it raises concerns that artificially high debt deductions are being allowed because of unreliable (high) asset valuations.
4.60 There may be a case for restricting asset values in cases like these.
4.61 The concerns are likely to be lessened in cases in which the internal organisation coincides with an external acquisition – for instance, when an entire multinational group is purchased by an unrelated party which sets up a new holding company structure in New Zealand. In this case there has been an external purchase and purchased goodwill or other intangible assets are recognised at their fair value. Allocation of that goodwill between entities within the group could still be a concern, however, depending on the details of accounting for the acquisition, and how any resulting asset uplift has been distributed around the worldwide group.
Proposal: Asset uplift
4.62 We propose that when the total asset value of a New Zealand or worldwide group increases as the result of the sale of assets between associated persons, the increase will be ignored for thin capitalisation purposes.
4.63 The intention of the proposal is that it will not be possible to revalue assets upwards by means of an internal reorganisation, unless a straightforward revaluation of that asset would be permitted under generally accepted accounting practice. At least for tax purposes, it should not be possible to circumvent normal restrictions on revaluations by buying and selling entities or assets within a commonly controlled group.
4.64 An exception might be made if the sale of assets took place as part of the sale of the entire group to a previously non-associated party, and the increased asset value reflected the fair value of the assets to that party, as determined under generally accepted accounting practice. In this case, the sale to an external party has a cost which can be used as a reliable basis for determining underlying asset values. Whether or not such an exception is practical will depend on the ability to write a sensible rule for determining when the sale of assets is linked to the sale of the entire group.
Questions we would like you to consider
|
1 We looked at a sample of approximately 80 foreign-controlled New Zealand groups who were surveyed by Inland Revenue to obtain financial data. We then obtained worldwide group debt-to-asset ratio from the Orbis database for these companies. Of 64 companies for which data was available, only 2 had a worldwide ratio exceeding 60%.