3. Potential solutions

3.1 Two high-level options have been identified for exempting foreign-sourced income derived by a non-resident through a PIE:

• A PIE with resident and non-resident investors and only foreign-sourced income. The non-resident investors would have a zero percent portfolio investor rate (PIR) for all income. Resident investors would have standard PIRs.
• A look-through global investment option that would allow a PIE to have both resident and non-resident investors and New Zealand and foreign-sourced income. The PIE would apply a different tax rate to different types of income derived by non-resident investors and it would therefore be necessary to track income from different sources, apportion expenditure to that income and allocate it to investors as if each stream were the only income derived by the PIE.

3.2 These options are discussed in more detail below.

Option 1: PIE with resident and non-resident investors and foreign-sourced income

3.3 This option would allow resident and non-resident investors to invest in a PIE that derived only foreign-sourced income, with perhaps some allowance for a minimum threshold of investment in New Zealand equity and debt. The non-resident investors would have a zero percent PIR for all income of the PIE, including the minimum income threshold from New Zealand sources discussed below.

3.4 Income that relates to offshore investments and that is allocated to a non-resident would be deemed to be foreign-sourced income. A CFC and non-resident trustee of a trust settled by a New Zealand resident would be treated as residents for the purposes of applying the exemption.

3.5 In exploring the type and level of New Zealand-sourced income that could be allowed under a minimum threshold exemption, officials have compared the treatment a non-resident investor would receive if the investment were made directly with that received if the investment were made through the PIE. If the non-resident would be more favourably treated when investing through the PIE, we propose a restriction on that income. For example:

• Interests in land. A non-resident investing directly in an interest in New Zealand land pays New Zealand tax on any lease payments. A non-resident would be more favourably treated if they had a zero rate in a PIE with New Zealand-sourced income derived from an interest in land. Officials therefore propose that a PIE that holds a direct interest in land in New Zealand not be eligible for this treatment.
• New Zealand equity. If a non-resident invests directly in shares in a New Zealand company, and receives a fully imputed dividend, the dividend will generally have borne New Zealand tax of 30%. Where a dividend is not fully imputed, the non-resident investor will be subject to NRWT at 15% or 30%, depending on the double tax agreement that applies.
  A non-resident would be more favourably treated if they had a zero rate in a PIE which received non-fully imputed dividends. Officials therefore suggest that investments in New Zealand equities be excluded. However, it may be that a minimum holding would be acceptable if this was required to reduce administrative costs or to allow existing “global equity” or “balanced managed” type funds to participate.
  Any minimum threshold could be based either on the value of assets or on income. If asset-based, the PIE could hold no more than, say, 5 percent of the value of all its assets in New Zealand equities which distribute unimputed dividends. If an income-based test is applied, income from other entities that is neither fully imputed dividends nor excluded income would not be allowed to exceed 5 percent of all income derived by the PIE in a year. The income-based test is more sensitive in targeting investments that do not give rise to a concern but it is more administratively complex.
  Officials are interested in feedback on these alternatives and the possible level of a minimum threshold exemption.
• NZ financial arrangements. Interest paid to non-residents investing in New Zealand debt instruments is subject to an approved issuer levy (AIL) of 2 percent or NRWT. A restriction on investments in New Zealand financial arrangements by a PIE with zero-rated non-resident investors is therefore required. Officials propose a minimum level of New Zealand-sourced income from debt for administrative and liquidity purposes.
  The Government is looking at the feasibility of introducing an exemption from AIL and NRWT for New Zealand bonds issued to non-residents. The restriction on investments in New Zealand financial arrangements would be reviewed if the policy in relation to AIL changed.
  Financial arrangements involving hedging instruments, as defined by NZ IAS 39, would not affect the amount of New Zealand assets that the PIE can hold. The policy behind this is to allow for a PIE to hedge against foreign-exchange risk in New Zealand, as this hedging would otherwise occur offshore, with no benefit to New Zealand.

3.6 Provisions would be required to deal with a breach of the requirement to have only foreign-sourced income (outside the allowed minimum levels).

Option 2: Look-through global investment option

3.7 This option would allow a PIE to have both resident and non-resident investors and New Zealand and foreign-sourced income. There would be no restriction on the New Zealand-sourced income but the PIE would apply a different PIR to different types of income derived by non-resident investors, as follows:

3.8 Income that relates to offshore investments and that is allocated to a non-resident would be deemed to be foreign-sourced income. A CFC and non-resident trustee of a trust settled by a New Zealand resident would be treated as residents for the purposes of applying the exemption.

3.9 It would be necessary to track income from different sources, apportion expenditure to that income and allocate it to investors as if each stream were the only income derived by the PIE. However, the PIE may establish its own rules (restrictions on the types of investors or investments) to reduce the number of permutations. The rules could also allow the use of minimum rather than actual PIRs. For example, a PIE which invested in New Zealand equity could apply a PIR of 15% (in relation to income of a non-resident of a treaty country) rather than separating income taxable at 15% from that taxable at 0%.

3.10 Despite the fact that there will be no tax liability in relation to foreign-sourced income under this option, the PIE would still be required to calculate and attribute net income, losses and tax credits to investors as if there were a tax liability. This is required to enable New Zealand to meet its exchange of information obligations under double tax agreements, to enable Inland Revenue to verify that the income of the PIE is from a foreign-source, and to cater for those taxpayers who have been incorrectly treated as non-resident.

Proposed approach and submission points

3.11 Each option discussed above has a number of advantages and disadvantages.

3.12 Option one is a more flexible vehicle and is still relatively simple. Existing funds are more likely to fit within the criteria and an individual moving from non-resident to resident status can remain with the fund. However, this option has the disadvantage that ensuring the de minimis rules are not breached involves some compliance and administrative costs.

3.13 Option two should maximise marketing opportunities for the funds management industry, allowing non-residents to invest through a PIE in New Zealand assets at the right tax rate and receive the benefit of no New Zealand tax on disposal of New Zealand shares. For example, an Australian investing in New Zealand debt through the PIE, who currently is taxed at 30%, would be taxed at 2%.

3.14 Option two also provides maximum flexibility to use existing funds, allowing fund managers to tailor funds to accommodate the degree of complexity they can handle. There is also no requirement to monitor minimum thresholds under this option.

3.15 Option two does, however, have significant disadvantages. There is administrative complexity for the PIE in tracing different sources of income and apportioning expenditure. The legislation is likely to be complex and this option also has the highest administrative costs for Inland Revenue. Finally, the marketing advantage may be overstated as it is unlikely that the tax paid by the PIE is creditable in a foreign jurisdiction. As a result, there may be a disincentive to invest in New Zealand assets through a PIE.

3.16 Officials are interested in feedback on the perceived benefits and compliance costs of the options outlined. In particular:

• Whether the funds management industry has a preferred option.
• Whether and why a minimum level of New Zealand-sourced income is useful under option one.
• The appropriate level for the minimum threshold exemption under option one.
• Whether the minimum threshold level should be an income or asset test.
• Whether the risks outlined in pages 4–7 to exempt foreign-sourced income derived by non-residents through a PIE are real risks, and the perceived likelihood of them occurring.
• The likelihood of the proposed benefits outlined in pages 3–4 occurring.