That the general anti-avoidance rule (GAAR) contain an objects clause requiring that its application be consistent with, and support, the structure and underlying policy reflected in objects clauses in other parts of the income tax law.
The Review's proposed structural reforms have generally embodied the principle that business transactions with similar economic substance should be taxed in the same manner. Those structural reforms, by significantly reducing existing disparities and flaws in the tax law, will reduce the reward from undertaking activities purely for tax purposes.
Nevertheless, where opportunities for tax avoidance remain the powers already available in the general anti-avoidance rule (Part IVA of the 1936 Act) should generally be the measures used against them -- rather than specific anti-avoidance measures.
The objects clause of the GAAR should clarify that the general anti-avoidance powers are to be exercised in a manner consistent with, and supportive of, the tax policy principles embodied in other provisions in the tax law. It should also make clear that the GAAR will not apply to the mere use in a straight-forward and ordinary manner of structural features of the law to best advantage -- for example, the use of an election or an entity structure (as opposed to the use of such features as part of a wider tax avoidance arrangement) does not in itself bear the stamp of avoidance.
Such a statement of policy should confirm the circumstances in which the GAAR could be applied and reduce the perception that valid business practices could unintentionally be subject to the application of the GAAR.
The Review considered arguments for and against including `the principal effect' test in the GAAR. None of the submissions to the Review that addressed the issue supported the principal effect test because of its potential wide scope to disturb business practices that have a genuine commercial purpose. On balance, the Review concluded there did not appear to be a strong need for the principal effect test, albeit that the test is included in the GST law. The existing dominant purpose test provides sufficient scope to deal with tax avoidance schemes.
Structural reform the primary mechanism
(a) That where specific tax avoidance is being driven by structural flaws:
Preference for general over specific anti-avoidance rules
(b) That where a non-structural response to tax avoidance is required:
Board of Taxation to review anti-avoidance framework annually
(c) That an annual review of the application of the policy framework and processes relating to the GAAR and specific anti-avoidance rules be:
In A Platform for Consultation (Chapter 24), the Review concluded that consistent application of policy, legislative and administrative design principles (see Schedule 1 of the Charter of Business Taxation in Section 1) is the preferable means of ensuring tax system integrity -- without resort to complex specific measures to deal with tax avoidance.
Tax avoidance may be characterised as a mis-use or abuse of the law rather than a disregard for it. It is often driven by the exploitation of structural loopholes in the law to achieve tax outcomes that were not intended by the Parliament but also includes manipulation of the law and a focus on form and legal effect rather than substance. The way things are done in order to take advantage of structural loopholes, or dress up or characterise something to satisfy form but not substance can also stamp an arrangement as avoidance. Tax avoidance represents a serious threat to the integrity of the tax system and to the revenue. It is also a form of subsidy from those paying their fair share of tax according to the intention of the law to those shirking their similar obligations.
Where tax avoidance is made possible by structural flaws in the law, sound structural reforms, undertaken promptly, will be able to deal completely satisfactorily with it.
Ideally, within 12 months of identification of a particular tax avoidance arrangement which relies on a structural flaw, that underlying flaw would be addressed. Such prompt action would ensure that the tax system remains structurally sound and the GAAR is not required to be used indefinitely as a substitute for reform.
The Review also considered the use of the GAAR in a period of major structural reform. For example, during the transitional period, significant taxpayer responses are likely to some of the recommendations in this report. The reductions to the company and capital gains tax rates are key examples. The Review endorses the existing practice of not employing the GAAR where certain taxpayer or market behaviour is an acceptable outcome of the tax law's structure.
The future approach to tax avoidance will be based around a clear articulation of the underlying policy of a restructured tax law, supported by the GAAR as modified by Recommendations 6.1 to 6.5, for arrangements structured to frustrate that policy intent. Only where it is not possible to address tax avoidance through structural reform or by the GAAR, should specific anti-avoidance provisions be considered.
Tax avoidance is an ongoing issue, reflecting in part the structure and compromises inherent in constructing a politically sustainable tax system. Dealing effectively and consistently with emerging avoidance requires a well-founded appraisal process. Often the most direct and durable approach will be through structural reform. This is also the course which should be adopted wherever possible because it will contribute most to maintaining the integrity of the tax system.
The following questions (based on the list in A Platform for Consultation, page 522) provide guidance as a basis for developing criteria to support a sound framework for combating tax avoidance while guarding against specific anti-avoidance provisions that cannot be justified after thorough cost-benefit analysis.
What is the problem?
Is the problem caused by a structural flaw in the law?
Can the structural flaw be addressed directly?
What are the revenue implications of the problem?
Can the GAAR be relied upon to deal with the issue?
What are the costs of compliance and the greater complexity that attach to specific anti-avoidance provisions that might be introduced, or to the possible application of the GAAR?
Are there second round effects that need to be considered? One dimension of this question is the effects of market adjustments to the underlying structural flaw. Taxpayer response to the specific anti-avoidance provisions themselves is another.
Ensuring that the integrity of this approach is maintained will also be one of the significant monitoring responsibilities of the Board of Taxation.
An annual review of the application of the GAAR, including its use in public and private rulings, in terms of the policy framework set out in this recommendation will be conducted by the Board of Taxation. The Board's review will be published in its annual report. Clearly, any advice that clarifies the circumstances in which the GAAR has been applied and outlines broad circumstances where it would apply improves certainty for taxpayers. At the same time, the Review believes that such advice must not extend to `mapping the minefield' for taxpayers.
Structural flaws giving rise to the application of the GAAR, as canvassed in that annual report, will be included in the forward work program for attention. Legislation to effect structural reform will need to be given priority in the legislative process (see Recommendation 2.4).
More generally, the Review sees a role for the Board of Taxation in monitoring the policy guiding the implementation of anti-avoidance provisions and advising whether any amendments to the GAAR are needed. To that end, the Board, with the authority of the Treasurer, may consult with taxpayers on appropriate responses to tax avoidance.
That the definition of a tax benefit be expanded to cover any reduction or deferral of tax payable, including through tax rebates and credits or losses.
A tax benefit constitutes a reduction in, or deferral of, tax payable. Practical application of that concept should not depend on the method used by the taxpayer to obtain the tax benefit. The current method-based definition, originally confined to practices involving the reduction of assessable income or increasing allowable deductions, was expanded in an ad hoc manner over recent years by inserting provisions dealing with foreign tax credits, capital losses and franking credit trading. These add-ons can be deleted, because the expanded definition of a tax benefit renders them redundant.
Restrictions on franking credit trading (section 177EA) -- one of the recent add-ons -- have also employed a wider purpose test for taxpayers than the more narrow dominant purpose test applicable in the rest of Part IVA. Singling out franking credit trading for particular anti-avoidance action, in conjunction with recently enacted specific anti-avoidance provisions, was partly influenced by arrangements abusing the section 46 inter-corporate dividend rebate. Accompanying the Review's recommendation that section 46 be replaced (Recommendation 11.4), section 177EA should also be repealed, as supported by focus group participants.
The revised tax benefit definition would apply to `treaty shopping'. Such arrangements involve structuring transactions through treaty countries in order to escape Australian tax. An example would be re-routing royalty payments through a series of countries in order to reduce withholding tax liability. The OECD has recommended that member countries develop robust anti-avoidance provisions to combat such arrangements. The US enacted specific anti-treaty shopping laws in 1997.
That operation of the existing reasonable hypothesis test (in section 177C) be improved by ensuring the counterfactual to a tax avoidance scheme reflects the commercial substance of the arrangement.
Currently, in order to demonstrate the existence of a tax avoidance scheme, the Commissioner of Taxation is required to construct a reasonable alternative transaction or counterfactual which does not give rise to the tax benefit. In some tax avoidance cases promoters of the scheme have argued that the reasonable alternative to the scheme may be that the taxpayer would not have done anything. The recommendation will confirm that this is not the case. For example, if the sale of property had an attached tax benefit, the alternative transaction would be constructed on the basis that the sale of property, without the tax benefit, would have taken place.
That the Commissioner of Taxation be allowed to issue a single determination in respect of a scheme where more than one person can obtain a tax benefit through participation in the scheme.
At present, individual determinations have to be made for every participating member of a scheme. A more practical approach will be for the Commissioner of Taxation to issue a determination declaring the application of the GAAR to a particular scheme where more than one person can obtain a tax benefit through participation in that scheme (for example, an investment project encompassed within a single prospectus).
That a wide range of specific anti-avoidance provisions be repealed consequent on implementation of a number of recommended structural reforms.
The Review's recommendations in relation to taxing unfranked distributions passing between resident entities, the profits first rule, the slice approach, and an effective way to distinguish debt from contributed capital will eliminate the need for most of the specific anti-avoidance measures discussed in Chapter 24 of A Platform for Consultation.
These specific provisions include:
capital streaming via bonus shares and in general (sections 45 and 45A of the 1936 Act);
dividend substitution (section 45B of the 1936 Act);
capitalisation of profits (Division 7B of Part IIIAA of the 1936 Act and the Income Tax (Untainting Tax) Act 1998);
substitution of redeemable preference shares (section 46D of the 1936 Act); and
entity losses (section 170-175 of the 1997 Act).
An additional large number of specific anti-avoidance provisions will also become unnecessary on account of other structural changes and by making the GAAR more robust. Replacement of these specific anti-avoidance provisions with sound structural measures will constitute a significant simplification of tax law.
The extent to which the need for some other specific provisions, which may be viewed as anti-avoidance provisions, is lessened or eliminated requires further evaluation in light of the recommended structural reforms. These provisions include:
tax-free distribution of pre-acquisition profits (sections 46A, 160APHA and 177E of the 1936 Act);
reduction in cost base of shares (to reduce potential capital losses) in respect of distributions to controlling entities of pre-acquisition taxed profits (sub-section 110-55(7) of the 1997 Act) -- an issue relevant to distributions as dividends or in the form of share buy-backs; and
the income injection test (Schedule 2F of the 1936 Act and section 175-10 of the 1997 Act).
That the dividend streaming and franking credit trading rules be amended as follows:
(i) the `45-day' rule be replaced by a `15-day' rule;
(ii) the exemption for small transactions be raised from $2,000 to $5,000; and
(iii) the rules applying to trusts be simplified and clarified for taxpayers in the light of the Review's other recommendations -- such as taxing trusts as companies and simplifying administration of the franking account.
In relation to the specific provisions relating to dividend streaming and franking credit trading, the components of this recommendation do not reflect a thorough assessment of those provisions against the questions set out under Recommendation 6.2. Instead they represent an initial paring back of the undue breadth of those provisions.
The 45-day rule requires that shareholders hold their shares, at risk, for at least 45 days to be eligible to receive franking benefits from dividends paid on the shares.
Commercial transactions often require relatively short periods of owning a share -- that is, shorter than the mandated 45 days. Where the share pays a franked dividend during an ownership period less than the mandated 45 days, the attached franking credit is denied to the resident taxpayer. Furthermore, even if the share were held for at least 45 days, the franking credit is denied if the resident taxpayer has eliminated 70 per cent or more of the ownership risk through other financial transactions during that period.
Modern capital markets offer investors a variety of ways to repackage ownership rights and risks to suit particular investor profiles. The franking credit trading restrictions could deter such transactions, to the detriment of investors, and do not assist in the development of Australia as a global financial centre. Where the revenue cost is not excessive, it is clearly desirable to lessen the adverse impact of these rules.
The 45-day rule was adopted on the basis that the United States operated with a similar provision in relation to its analogue of the inter-corporate dividend rebate -- a different situation to franking credit trading in Australia. Moreover, the US requirement does not provide the only benchmark. Other countries have operated with shorter ownership period rules. For example, a UK rule prevented the transfer of tax credits from dividends paid on shares purchased or sold within a month.
Reducing the ownership period to around 15 days should be possible without significant revenue cost. The tax avoidance schemes uncovered by the ATO, in the period prior to the rules taking effect, have generally involved an ownership period of somewhat less than 15 days (the dominant purpose being to obtain the tax benefit rather than owning the share for commercial reasons). The factors relevant to deciding on an appropriate holding period include the volatility of share prices, the price of reducing risk and the value of the franking benefit.
Commercial transactions with a very short time frame (a matter of a few days) would still suffer a denial of franking credits but the 15-day ownership period rule will undoubtedly provide relief for other commercial transactions.
Franking credit trading is more profitable for large trades. Currently, in recognition of this, an exemption threshold of $2,000 in franking rebates is provided in the franking credit trading rules.
That exemption could be increased, say, to $5,000. The maximum fully franked dividend that could be claimed without regard to the 15-day rule would be $8,900 (at a 36 per cent company tax rate). Assuming these dividends represent a yield of 4 per cent, increasing the threshold would amount to exempting a share portfolio of approximately $220,000.
This modification has the practical effect of reducing the reach of these rules for a great many small investors. Since these taxpayers are unlikely to engage in franking credit trades (especially with the deterrent of the potential application of the GAAR) the revenue loss should not be great.
Where shares are held indirectly through a trust, the beneficiary must hold the beneficial interest, at risk, for at least 45 days to be eligible to receive franking benefits from dividends paid on the shares. The `at risk' requirement is problematic for interests which are inherently riskless (for example, the interest of a discretionary trust beneficiary). Special measures, which add to the complexity of the rule and impose compliance costs, are required for family trusts to ensure family members are not denied franking benefits inappropriately.
There is clearly scope to modify the rules pertaining to trust interests to reduce their complexity and reduce compliance costs.
In determining whether shares are held at risk, offsetting positions relating to the shares need to be taken into account. Sometimes it is not clear whether an offsetting position relates to a particular share, and this can cause uncertainty. The relevant provisions governing offsetting positions allow for the making of regulations to provide further guidance in relation to these and other issues.
Also, the measure used to determine risk -- delta -- may be difficult to calculate at times. For example, when a share's delta or measure of risk is not observable or readily obtainable in the market place, no guidance has been provided to taxpayers on how this measure should be computed. Without such knowledge taxpayers are placed in a position of not knowing whether they have held the share at sufficient risk in order to qualify for franking credits. This guidance should be provided as soon as possible.
Recommendation 11.6 to simplify the franking account will provide a further opportunity to rationalise the anti-dividend streaming rules. The taxing of trusts as companies could also offer some scope for rationalising these rules.
Income recognition principle
(a) That, subject to the exception in paragraphs (b) and (c) in relation to financial distress, gains on the extinguishment of business liabilities be included in taxable income.
Debt forgiveness qualifying for financial distress arrangements
(b) That, to qualify for debt forgiveness arrangements, the decision to forgive:
Tax treatment of qualifying extinguishment gains
(c) That gains on business liabilities forgiven because the debtor is in financial distress:
Matching of limited recourse debt to `realisation' assets
(d) That subject to paragraphs (b) and (c), gains on limited recourse debts that finance assets that are subject to realisation treatment and capital loss quarantining be taxed as capital gains.
A liability may be extinguished in a number of ways - including forgiveness, repurchase, defeasance or early redemption. Irrespective of the circumstances in which a liability is extinguished or the reasons a gain has arisen, if the liability is extinguished for less than the amount owed, the difference represents an economic gain to the issuer (as noted in Chapter 6 of A Platform for Consultation at page 179). Such extinguishment gains -- by relieving the debtor of an economic burden -- should be taxed in a consistent manner. In the creditor's hands, the corresponding extinguishment loss similarly requires appropriate tax recognition.
Extinguishment gains and losses pose two key challenges for business tax policy:
on the one hand, normal commercial practice between unrelated parties transacting at arm's length may see debt forgiven where the debtor is in genuine financial distress -- with debt forgiveness offering some likelihood that the debtor may be able to trade out of those difficulties;
on the other hand, debt forgiveness transactions between related parties not transacting at arm's length -- as, for example, between two subsidiaries of a wholly owned company group -- may offer the prospect for recognising losses artificially created by debt forgiveness whenever the corresponding gain to the debtor is not taxed at the same time.
Provision of tax relief on the gains of a debtor in genuine financial distress has to be balanced against the need for structural integrity in ensuring that losses created by artificial debt forgiveness do not receive tax recognition for the creditor unless the corresponding gain to the debtor is also brought to tax.
Neither side of the debt forgiveness coin -- the case for relieving financial distress against the need for enhanced structural integrity -- can be ignored in determining appropriate arrangements for taxing extinguishment gains. As a general principle, the Review proposes that gains on the extinguishment of business liabilities be included in taxable income (thereby nullifying the tax advantages of artificial loss creation), with the existing debt forgiveness concessions recognising genuine financial distress being retained as an appropriate safety net.
While extinguishment gains should be taxed in a consistent manner, under existing law their treatment is uncertain and unsystematic. In particular, no specific provision or general income tax principle establishes that extinguishment gains arising from changes in market conditions are subject to tax. Nor is there a generic provision or income tax principle that reconciles the tax accounts with the expenditure that the taxpayer actually bears. Instead, there is a patchwork of provisions dealing with extinguishment of liabilities:
Existing commercial debt forgiveness (CDF) provisions are designed to address extinguishment gains arising out of debts forgiven as a result of changes in the debtor's creditworthiness. They do not extend to gains arising out of changes in market variables.
Division 7A treats the forgiveness of debts owed by shareholders or associates to private companies as non-frankable dividends.
Proposed Division 243 provides for the clawing back of capital allowance deductions that are financed by limited recourse debts that are not repaid in full.
Current value shifting provisions applying to wholly owned groups do not address the potential to create artificial losses from debt forgiveness.
Introduction of a generalised regime for the taxation of extinguishment gains will remove a fundamental weakness in the existing tax law, enhance certainty, result in greater structural integrity and allow the removal of a range of specific provisions. Structural weakness in the current regime has contributed significantly to the need for interim tax avoidance measures pending structural reform in this and other areas. These interim measures include those directed at lease assignments (Recommendation 10.13) and that dealing with capital losses created by the forgiveness of debt between wholly owned subsidiaries of a company group (Recommendation 6.19).
A generalised regime dealing with extinguishment gains will address arrangements which avoid current balancing charge provisions; have features such as structured non-payment of debts because the lender achieves its required rate of return through tax benefits in the early years of an arrangement; or involve disposal of assets not for cash but for the benefit of not having to repay a liability, a benefit currently not recognised for tax purposes.
Having accepted the general principle that extinguishment gains should be taxed, the question then arises of how to bring extinguishment gains to tax. Two broad approaches are possible. Extinguishment gains could be taxed by inclusion in taxable income or by application against current or future tax deductions.
Taxing extinguishment gains by inclusion in taxable income is the most robust approach. First, it brings the economic gains to account when they are realised. This limits the ability of taxpayers to structure tax deferral or tax mitigation around the extinguishment of a liability. Second, it produces symmetrical treatment between debtors and creditors. Whenever a debtor makes a gain on the extinguishment of a liability, a creditor makes a corresponding loss. Creditors can generally deduct their losses.
Inclusion in taxable income is also broadly in line with the approach to taxing extinguishment gains adopted in New Zealand, the United Kingdom and the United States.
The disadvantage of including extinguishment gains in taxable income is that it can cause some debtors cash flow difficulties. The economic gain is not matched by a corresponding cash inflow.
The alternative approach is to apply extinguishment gains against current or future tax deductions or other benefits. This is less likely to cause debtors cash flow difficulties. Under this approach, tax deductions that would otherwise be claimed by the debtor (for example, arising from tax losses or depreciation allowances) are reduced to the extent that a liability is extinguished.
As discussed in Chapter 6 of A Platform for Consultation (pages 180-181), there are two possible ways of applying extinguishment gains against current or future tax deductions.
One method is to apply extinguishment gains to reduce future tax deductions in accordance with a set of detailed ordering rules. This is adopted in the existing commercial debt forgiveness provisions (the CDF method).
The other method is to claw back write-off allowances or reduce the tax values of assets that can be linked to an extinguished liability (the expenditure recoupment method).
Both the CDF and expenditure recoupment methods can result in significant tax deferral. Taxation of the gain is generally dependent on the timing of deductions. This means gains can be deferred, possibly indefinitely, with consequential inefficiencies. For example, application of a gain to reduce the tax value of an asset would give a debtor an incentive to retain that asset.
Further, the expenditure recoupment method is reliant on complex tracing provisions that would be hard to apply and open to manipulation. Trying to link extinguishment gains to particular expenditures is difficult because of the fungibility of money.
On balance, inclusion in taxable income stands out as the soundest approach -- particularly in enhancing structural integrity -- and the Review considers that it should be the general rule for the taxation of extinguishment gains. As a major exception, however, the Review also recognises that such a rule could have an adverse impact on debtors that are in financial distress.
Commercial decision making can involve creditors entering into arrangements with debtors to settle liabilities for payment of less than the full amount owing. Typically these arrangements are designed to enhance the long term viability of a debtor by putting it in a position to trade out of its financial difficulties.
In these circumstances, the triggering of an immediate tax liability in the hands of the debtor on the extinguishment of the liability could frustrate the negotiation of settlements altogether. There is, therefore, a case for deferring the taxation of extinguishment gains that arise when a creditor makes a bona fide commercial decision to forgive a liability based on the debtor's inability to pay.
In these cases, the CDF method of applying extinguishment gains to reduce future tax deductions represents an appropriate approach, provided that unapplied gains are carried forward from year to year. An exception to this would be in the case of bankruptcy. To ensure a fresh start, unapplied extinguishment gains will be disregarded on discharge from bankruptcy.
Determining the scope of the financial distress exception is problematic because what is `financial distress' may depend on subjective commercial factors. To counter possible abuse, the concession will be carefully targeted only to apply where a decision to forgive is made on a non-tax commercial basis and where the debtor is in genuine financial distress. To qualify for concessional treatment, forgiveness would need to be by formal, executed agreement between non-related parties dealing on an arm's length basis.
Moreover, the exception should be available only where there is direct nexus between a creditor's decision to forgive a liability and the debtor's incapacity to pay. To illustrate, this means that the exception should not be available to a debtor, otherwise in good financial health, who is able to extinguish a liability for less than full value because the creditor's recourse is limited to an asset that has happened to fall in value.
Where assets financed by extinguished liabilities are subject to capital loss quarantining, an issue arises whether the extinguishment gain should be treated as a capital gain for tax purposes.
Take the example of an individual who borrows to buy shares. Assume the shares subsequently fall in value and the debt is forgiven. Under the general rule, the gain on the extinguishment of the liability would be included in taxable income. Conversely, the loss on the shares would be quarantined for offset against gains on assets receiving capital gains and loss-quarantining treatment, or collectibles and other assets held by the individual wholly or purely for private purposes (see Recommendation 4.10(d)). Without a specific character matching rule, the extinguishment gain could not itself be offset against the capital loss.
As noted, because of the fungibility of debt, a blanket rule requiring the matching of extinguishment gains or debt to losses on associated assets would be very difficult to administer. Tracing is not an issue, however, where a limited recourse debt is used to finance the acquisition of property (to which the lender's recourse is confined). In such cases there is a clear nexus between the borrowing and the underlying property. Matching will therefore be undertaken by treating extinguishment gains as capital gains to the extent that the loss on assets financed on a limited recourse basis is subject to capital loss quarantining.
A separate issue is whether gains arising on business liabilities extinguished for reasons of natural love and affection or by testamentary disposition should be taxed.
On one view, the costs associated with these liabilities are deductible irrespective of the source of the finance. Taken in that context, gains arising on all business liabilities should be taxed.
The contrary argument is based on an analogy with the gifting provisions. A forgiveness in such circumstances is comparable to a gift and unlikely to give rise to an allowable loss in the hands of the creditor. It would not therefore be appropriate to tax the corresponding gain in the hands of the debtor. On balance this is considered to be the better view.
Hence, where a liability is forgiven for reasons of natural love and affection or by testamentary disposition, the debtor will be treated as having repaid the creditor the amount of the liability that is outstanding. On this approach forgiveness will not give rise to a gain in the hands of the debtor, nor will it give rise to a corresponding loss in the hands of the creditor.
Repair continuity of ownership test
(a) That the defects in the continuity of ownership test currently applying to companies' tax losses be removed so that, for losses claimed after the date of announcement, the test is satisfied only if a company's majority underlying ownership is maintained throughout the period from the time the tax loss was incurred to the time the loss is claimed.
Retain same business test
(b) That the same business test be retained, subject to the removal of inter-entity loss multiplication.
In Chapter 28 of A Platform for Consultation options are provided for dealing with loss and gain duplication. Illustrations of loss duplication are given in Examples 28.1 to 28.3 in that chapter. Broadly, duplication arises because losses and gains on assets are reflected in the value of membership interests in an entity that holds those assets -- while the tax value of the membership interests remains unchanged. Losses and gains on the sale of membership interests in an entity are duplicated when the entity is also allowed the losses or taxed on the gains arising from disposal of the assets. In entity chains there can be multiple layers of duplication.
Where a company has carry-forward tax losses, the continuity of ownership test would prevent substantial duplication subject to the same business test, by denying the loss to the company where there has been a change in the company's majority underlying ownership. Broadly, the existing continuity of ownership test is satisfied if the majority ownership of a company in a loss year is the same as in the year in which the loss is sought to be used to reduce taxable income.
The current test has two defects. First, it does not require continuous majority shareholding. This defect could allow a loss to be completely duplicated by a shareholder selling its 100 per cent equity interest to crystallise a tax loss. The shareholder could then restore its majority ownership status in the company by reacquiring its shares when the company seeks to claim the duplicate loss. Secondly, there is no requirement to test for proportionate changes in shareholding among a continuing group. This defect could allow a realised loss to be almost completely duplicated by an equity owner, say, selling down its interest from 99 per cent to 1 per cent. Table 28.1 on page 599 of A Platform for Consultation illustrates this flaw -- as well as the remedy reflected in the recommendation. Recommendation 6.9(a) will address both these flaws.
Current tracing rules will continue to protect listed public companies and their wholly-owned subsidiaries from having to monitor their shareholdings on a daily basis. Broadly, these entities are required to test their shareholdings only annually -- unless there is an abnormal trade in their shares.
The same business test allows a company whose majority underlying ownership has changed to carry forward its tax losses where the company continues to carry on the same business -- even after the shareholders of the company sell the company to other taxpayers to realise the loss reflected in the value of the company's shares. In these circumstances, however, loss duplication may be temporary where the losses are applied to reduce subsequent profits earned in the company. The duplication would be reversed if the profits freed from tax by the losses are distributed or if the company is sold. The Review considers that the problems arising from the potentially temporary duplication of losses do not justify the adverse impact on continuing individual shareholders of denying the loss carry forward in cases where businesses satisfy the same business test.
The possibility of multiplication of the loss needs to be addressed, however, where entities are interposed between individual shareholders and a loss entity (an entity with realised or unrealised tax losses). Losses on the membership interests in the interposed entities will be denied to the extent of the tax losses in the loss entity. That is illustrated in Figure 6.1. This will prevent multiplication of losses up an entity chain. The tax value of any inter-entity membership interests in a loss entity will be reduced by the loss entity's realised and unrealised tax losses (net of unrealised gains) when there is a change in the loss entity's majority underlying ownership. The reduction in tax value will be made in respect of direct and indirect interests in the loss entity. The requirement to reduce tax values may be negated to the extent that it can be demonstrated that the loss asset has not impacted on the market value of the interests in the entity.
Figure 6.1 Removing inter-entity losses
Under the new entity tax regime, the application of the amended continuity of ownership test and the same business test will extend to all other entities subject to that regime (see Section 13). Section 15 contains a discussion of how the duplicating of losses will be prevented within consolidated groups through the design of the consolidation regime.
That, where there is a change in an entity's majority underlying ownership after the date of announcement, losses subsequently realised on assets held at the time of ownership change be made subject to the same business test:
(i) to the extent of the net unrealised loss of these assets relating to the period before the change; and
(ii) subject to Recommendation 6.9(b) concerning the removal of inter-entity loss multiplication.
Both the current continuity of ownership test and the same business test apply only to realised losses. These rules may be circumvented where entities are sold with unrealised losses. The recommended approach will align the treatment of unrealised losses with realised losses.
The approach is a modification of the Canadian-based model presented in A Platform for Consultation (pages 600-601). The modification is consistent with Recommendation 6.9(b) to retain the same business test for realised losses. Both models presented in A Platform for Consultation (pages 600-601) could have been modified to align the treatment of unrealised losses with realised losses. The modified Canadian-based model will be relatively easier to comply with and administer even though assets will be required to be valued at the time of a change in majority underlying ownership.
Its operation is illustrated in Example 6.1. Focus group members generally supported the Canadian-based model as presented in A Platform for Consultation. The United Kingdom-based model is more arbitrary and imprecise, and would require more extensive record keeping.
Removal of grouping provisions to mitigate loss cascading
(a) That loss transfer and asset rollover concessions currently applying to transfers between entities be removed when consolidated tax treatment for wholly owned entity groups (see Section 15) is implemented.
Disposal of loss assets within majority owned groups
(b) That losses not be realised as a result of the disposal of `loss assets' by one entity to another in the same majority owned group (and tax values of the assets remain unchanged).
Loss cascading occurs where losses are duplicated through a chain of companies so that multiple losses are allowed to the same group. Loss cascading is facilitated by the existence of company grouping provisions (relating to loss transfers and asset rollovers) and the recognition of intra-group membership interests as assets for capital gains tax purposes. Consolidation will prevent cascading. While consolidated entities will be allowed to transfer losses and assets within the group, intra-group membership interests will generally no longer be recognised.
Currently, losses may also be cascaded where `loss assets' (assets carrying an unrealised loss), including membership in an entity with losses or loss assets, are disposed of within a majority-owned group. This will be addressed by ensuring that the sale of a loss asset by one entity to another in a majority owned group does not give rise to a realised loss for tax purposes. The tax value of the asset would be `rolled over' with no change to the acquiring entity.
The measure will have the same effect as the current compulsory rollover for capital gains tax purposes for loss assets transferred within a wholly owned group.
Replacement of existing provisions
(a) That the general value shifting rules referred to in Recommendations 6.13 to 6.16:
De minimis exception
(b) That a de minimis exception be included which balances integrity considerations and the containment of compliance costs.
Other adjustments to be made as necessary
(c) That the tax value adjustments referred to in Recommendations 6.14 to 6.16 be accompanied by such other adjustments as are required to implement general value shifting rules under the new legislative framework for calculating taxable income.
Broadly, `value shifting' describes transactions and other arrangements which result in the value of an asset being reduced with a consequential increase in the value of another asset. The two assets involved may be owned by the same taxpayer or by different taxpayers. Assets for these purposes include shares in companies and interests in trusts.
Broadly, most gains and losses are calculated as the difference between an asset's `cost base' (its `tax value' under the proposed legislation) and its market value when the asset is realised. Value shifting arrangements distort the relationship between the asset's market value and tax value. When assets are realised, inappropriate gains and losses are calculated.
For assets where gains are taxed and losses allowed on a realisation basis, value shifting presents a particular problem to the integrity of the tax system. It is possible to shift value out of assets (to bring forward losses) and defer gains on assets to which the value is shifted (by choosing not to realise those assets).
Share value shifting (variation of share rights and issues of shares at `undervalue' -- that is, less than market value) and `asset stripping' within 100 per cent commonly-owned and grouped companies have been addressed in the existing tax law.
The main deficiency with the existing law is that it deals only with some types of value shifting. The following are examples.
Share value shifting is covered, but not value shifting on interests in trusts.
Asset stripping rules are limited to 100 per cent owned companies: trusts are not dealt with, nor are less than 100 per cent owned, but controlled, entities which are capable of implementing arrangements for shifting significant values.
Asset stripping rules are limited to asset transfers and creations and then only where these occur at undervalue. Value shifts which do not involve the transfer or creation of property (for example, from debt forgiveness) are not addressed. Nor are transfers or creations for more than market value.
The current law is also very complex and detailed, with resulting high compliance costs. There is also no general de minimis exclusion. So, if a right is created intra-group for one dollar less than its market value, the asset stripping provisions potentially apply. A further problem with the existing rules is that they are too broadly drafted in parts, leading to unintended outcomes. For example, in kind dividends may be treated inappropriately as asset strips under the current law.
As deficiencies in the rules are identified, they are patched, resulting in a steady stream of complex amendments. This is all because the current law is ad hoc without a solid base upon which to set value shifting rules.
Consultation revealed a keenness for improvements to the existing law with an emphasis on workable legislation and reducing uncertainty. More complex, realistic examples in the legislation or extrinsic material would assist in this regard.
General value shifting rules, as outlined in Chapter 29 of A Platform for Consultation, will deliver significant integrity benefits to the new business tax system. Equity and efficiency will be enhanced as taxpayers are taxed more consistently on transfers of value, whether these occur by way of conventional realisation (for example, sale) or by value shifting. Generalised rules will also provide more consistent treatment for comparable value shifts across entities and transactions. The new provisions will also obviate the need for a continuing stream of anti-avoidance amendments as new value shifting transactions are detected.
A comprehensive de minimis exception is needed which balances integrity considerations and containment of compliance costs. Consultation revealed a preference for a more realistic exception than exists in the current law.
The design of the general value shifting rules will need to take into account other structural reform measures recommended by the Review including general debt forgiveness provisions, a general arm's length principle, the broad definition of entity distributions and reforms to the tax treatment of rights -- as well as the approach to specifying tax values of assets and liabilities in the core rules for determining taxable income (Recommendation 6.12(c)).
That the general value shifting rules be applied to:
(i) interests in, and assets of, entities (and their associates) that are controlled by other entities (and their associates);
(ii) assets of entities (and their associates) that control other entities; and
(iii) non-controlling interests in entities, but only where the holders of those minority interests actively participate in, or otherwise facilitate, value shifting arrangements.
Control is an appropriate threshold for the application of the general value shifting measures. A `wholly owned' threshold would leave significant value shifting transactions unaddressed whereas a threshold lower than control would impose unreasonably high compliance costs. Consultation to date has indicated support for a control threshold.
Ordinarily, value shifting measures would not impact on holders of non-controlling interests. It is appropriate to regard benefits obtained by such persons as windfalls. However, if such persons actively participate in, or otherwise facilitate, value shifting arrangements, the general value shifting rules will apply to them.
That the following general value shifting rules be applied to direct value shifts to the extent not addressed by other reform measures (for example, the tax treatment of rights):
(i) where value is shifted between the assets of different taxpayers or from post-CGT to pre-CGT assets of the same taxpayer -- gains be recognised;
(ii) where value is shifted between post-CGT assets of the same taxpayer -- tax value adjustments operate to neutralise the effect of such value shifts; and
(iii) losses not be crystallised by value shifting.
A direct value shift is where the arrangement takes value directly out of an asset. For example, this may occur if share rights are varied between classes of shares, rights attaching to trust units or interests are varied, or other rights are created out of an existing asset. (To the extent that new rules for the taxation of rights -- see Section 10 -- address the shifting of value out of underlying assets, the general value shifting rules will not be needed.)
Recognition of gains is appropriate where value is directly shifted between assets held by different taxpayers, or from post-CGT to pre-CGT assets of the same taxpayer. In the first case, the shift is equivalent in economic terms to a transfer of value by change of ownership. In the second case, any gain transferred cannot be recaptured on the realisation of the pre-CGT asset.
Where value is shifted within a post-CGT portfolio of assets owned by the same taxpayer, it is not appropriate for a gain to be crystallised for tax purposes. There has been no change of ownership of the transferred value. Tax value adjustments to the affected assets are sufficient to neutralise the value shift in this case.
For assets on a realisation basis, revenue costs are likely to be involved if the law allows losses to be crystallised on value shifting arrangements. Taxpayers could enter value shifting arrangements to create potential losses to be realised immediately, while deferring realisation of gains on assets to which the value is shifted. For this reason the recommendation is not to recognise such losses.
That direct value shifts from pre-CGT to post-CGT assets be recognised under the general value shifting rules.
Taxpayers should receive recognition for value shifted out of their pre-CGT assets into post-CGT assets. Although it would not be feasible to maintain pre-CGT status for the transferred value, the tax value of the post-CGT asset could be increased by the (pre-CGT) market value shifted.
That gains or losses not be crystallised on indirect value shifts, with tax value adjustments being made at the time of the value shift (using the `loss-focused' approach to implement these adjustments).
An indirect value shift is one where an arrangement affects the value of an entity and, as a result, the values of interests in the entity are indirectly affected. For example, if an entity transfers an asset worth a million dollars to an associate in exchange for an asset worth two million dollars, not only is the value of the transferor entity increased, but so too is the value of interests (for example, shares or trust interests) held directly or indirectly in the transferor.
That gains or losses be crystallised for tax purposes by the indirect effect of value shifting is not appropriate. In many cases, such treatment could lead to multiple recognition of gains and losses -- when, for example, the loss created is at the end of a chain of entities. Gains and losses can be recognised when the assets indirectly affected by the value shift are actually sold, so long as tax value adjustments are made at the time the value shift occurs to ensure the correct later calculation of the gain or loss.
A tax value adjustment regime is preferred over other methods for recognising the impact of an indirect value shift for the following reasons:
it can be done at the time of the value shift;
it is an observable response in a self-assessment environment;
it removes complications if other transactions occur before realisation; and
people are familiar with the approach because it is used under the current law.
A `loss-focused' approach for implementing the tax value adjustments involves making downward adjustments to tax values only where the value shift reduces the market value of entity interests below their tax values. Compensating upward adjustments are made to the tax values of entity interests whose market value is increased by the value shift, only where downward adjustments have been required on the other side of the transaction. The upward adjustment will have regard to the downward adjustment. Other value shifts at the indirect level are ignored, which helps to contain compliance costs. A `loss-focused' approach achieves the right balance between integrity objectives and containment of compliance costs.
Illustrations of the `loss-focused' approach to tax value adjustments for indirect value shifting are given in Example B.1 on page 628 of A Platform for Consultation.
Consultation to date has also revealed support for both tax value adjustments and the `loss-focused' approach to implementing those.
Implement structural rule for arm's-length valuation
(a) That a high level rule be introduced to ensure arm's-length value is used in calculating the taxable income of taxpayers who do not deal at arm's length.
Repeal specific arm's-length tests
(b) That those specific arm's-length tests in numerous provisions in the current law having the same effect be repealed.
Where taxpayers do not deal at arm's length, the current law substitutes a market value for the value ascribed by the parties if the ascribed value is different from the market value.
Such modifications under the current law are often made on a transaction-by-transaction basis (that is, a different provision addresses each transaction type). This necessitates numerous provisions scattered across the current law. Each modification is often expressed in different words and yields slightly different results depending on the transferred property.
The result of these numerous provisions, each with its own language, is a confusing state of affairs for non-arm's-length transactions. A large number of gaps arise where it is not clear whether the general principle should apply or whether taxpayers should qualify under an exception to the general principle. This outcome reflects the fact that most of the provisions were inserted into the tax law as ad hoc responses to particular situations. See A Strong Foundation, pages 33-34, for examples.
A general rule that substitutes an arm's-length value in non-arm's-length dealings represents a fundamental design principle that will achieve consistency across the tax law in calculating taxable income of taxpayers. It will also sweep away over one hundred provisions -- of itself representing significant tax law simplification.
Even where only one of the existing arm's-length rules applies, asset values can still be manipulated in order to distort the calculation of capital gains and losses. The value shifting rules (discussed at Recommendations 6.12 to 6.16) seek to prevent manipulation of asset values and represent a unified structural solution in the capital gains area. The general arm's-length principle will lend further support to value shifting reforms. An important design issue in the development of the new core rules for specifying taxable income is the integration of tax values of assets and liabilities, the general debt forgiveness provisions, the general value shifting arrangements and the general arm's-length principle -- as well as the broad definition of distributions from entities.
The general arm's-length rule will apply to:
related and unrelated parties who do not deal on an arm's length basis;
situations where two related parties transact through an unrelated party involving a non-arm's-length value;
international transactions taxation issues, recognising the need to remove unnecessary overlap with the transfer pricing rules (see A Platform for Consultation, pages 705-706); and
transactions involving debt forgiveness.
To enhance certainty for taxpayers and the ATO, the general arm's-length rule should contain guidance for determining an arm's-length value. For example, practical meaning needs to be given to what is an `arm's-length value' and how it is to be ascertained -- by reference to a market value or, where it does not exist, to a commercial value. Such guidance as currently exists for the specific provisions will be subsumed by guidance accompanying the general arm's-length principle. In this area, too, there are opportunities to simplify and clarify advice to taxpayers.
A single arm's-length rule will restate the current law's intention in a clear form. The advantages of this approach include:
a single rule replaces numerous provisions; and
the rule addresses gifts between related parties, transfers of assets for non-arm's-length consideration between unrelated parties, forgiveness of debt and avoids specific definitions of related parties (for example, `associates', `related parties').
A wide range of exemptions is currently available under the various provisions that substitute an arm's-length value. These exemptions include, for example, transfers of some assets such as trading stock and depreciable property to new partners, transfers of business assets more generally to beneficiaries of an estate, transfers of capital assets in cases of marital breakdown, involuntary disposals of assets, and transfers of assets upon incorporation or upon various types of entity reorganisation.
Consistent with regular reviews of tax expenditures, these exemptions should be reviewed regularly to determine whether they meet their original purpose and are cost effective.
Intra-group transfer of revenue losses not to allow loss duplication
(a) That where a revenue loss has been transferred between members of a wholly owned company group, crystallisation of a capital loss be prevented by:
Interim application of loss duplication measure
(b) That the loss duplication measure of paragraph (a):
As discussed under Recommendation 6.9, loss duplication arises when a single economic loss is recognised for tax purposes more than once. This can occur because losses realised in an entity are reflected in the value of interests in that entity.
Paragraph 28.7 of A Platform for Consultation highlighted a form of loss duplication called `loss cascading'. This occurs where losses are artificially duplicated through a chain of companies. Loss cascading is facilitated by the existence of loss transfer and asset rollover concessions applying to wholly owned groups. Some forms of loss cascading are prevented under the existing law but others are not.
An example of where loss cascading is not prevented is where a group company, which has incurred an economic (or actual) revenue loss for a year of income, transfers the loss to another group company under the revenue loss transfer provisions of the 1997 Act. Revenue losses include losses on assets outside the capital gains tax regime (for example, losses on assets held by traders or on depreciable assets). Because the effect of this loss will also be reflected in the value of direct and indirect interests (for example, shares) held in the group company, the revenue loss can be duplicated by the sale of the direct or indirect interests held in that group company. This is illustrated in Example 6.2.
The net capital loss transfer provisions of the 1997 Act prevent loss duplication where a net capital loss is transferred by requiring cost base reductions to certain interests held in the loss company. In contrast, the revenue loss transfer provisions of the 1997 Act do not contain this requirement.
This measure will apply to revenue losses transferred from 22 February 1999, consistent with the Treasurer's press release of that date.
Intra-group debt forgiveness not to permit artificial loss creation
(a) That where there is a forgiveness of a debt owed between commonly owned companies, the CGT effects of any related value shifting be negated by:
Interim application of debt forgiveness value shifting measure
(b) That the debt forgiveness value shifting measure of paragraph (a):
Capital gains tax (CGT) value shifting refers to arrangements which shift value out of assets, often to other assets. `Assets' can include shares in companies. Value shifting is a problem because it is possible to shift value out of assets (to bring forward losses) and defer gains on assets to which the value is shifted (by choosing not to realise those assets).
Paragraph 29.14 of A Platform for Consultation highlighted several deficiencies in the existing law dealing with value shifting. One of the deficiencies is contained in Division 138 of the 1997 Act, which deals with value shifting by asset stripping. Broadly, Division 138 applies to value shifts which occur between companies under common ownership, where the value shift arises by the transfer or creation of an asset for less than the asset's market value (or in some cases, cost base).
Division 138 does not apply where a debt, which arose from a loan from one commonly owned company to another, is forgiven and the debtor company does not pay any consideration for the forgiveness. In this situation, there is no transfer or creation of an asset. The transaction nevertheless shifts value out of interests in the creditor company, potentially allowing an artificial loss to be claimed on the disposal of those interests. The loss is artificial because no economic loss would be suffered as a result of the arrangement. This is illustrated in Example 6.3.
As for the preceding measure, this measure would apply to debts forgiven from 22 February 1999.
That all entities subject to the entity tax system be classified as either `closely held' or `widely held' for taxation purposes.
Entities will need to be classified as either closely held or widely held under the new entity tax system as some provisions apply only to widely held entities whilst other rules apply only to closely held entities.
For example, only widely held entities:
may be collective investment vehicles; and
may provide a certain level of benefits for members in the form of discounted goods or services that are excluded from the definition of `distribution'.
Provisions relevant to closely held entities include those addressing non-commercial loans involving a closely held entity and its members.
At present a number of different systems determine, for taxation purposes, the extent of the public control of an entity. For example, in the case of companies the tax law applies a `public' and `private' dichotomy. However, this is not the same distinction as exists under the Corporations Law. This distinction also cannot be easily applied to trusts. The trust loss provisions of the taxation law use concepts such as `very widely held', `widely held' and `closely held'.
A single, uniform approach for determining the extent of public control will increase both consistency and simplicity. A closely/widely held distinction for classifying entities covered by the entity tax system will:
be the simplest way of providing a single (consistent) method for classifying the nature of an entity's ownership; and
provide a comprehensive code which does not rely on other statutory rules, such as the Corporations Law.
That a `widely held' entity subject to entity taxation be defined as an entity where:
(i) the membership interests (for example, units or shares) are held by not fewer than 300 persons; and
(ii) 20 or fewer individuals do not hold, directly or indirectly, 75 per cent or more of:
The definition of `widely held' is broadly based on the standard definition used in the trust loss provisions. Taxpayers are familiar with this definition, which is easy to apply to trusts.
The additional requirement that the membership interests in the entity are held by not fewer than 300 persons is important in achieving a genuinely broad-based membership.
That a closely held entity subject to entity taxation be defined as an entity which is not widely held.
In order to prevent either a possible overlap between the definition of widely and closely held entities, or a gap between those definitions, closely held entity will be defined as any entity which is not widely held.
That a trust be treated as closely held if under the documents establishing the trust there is a discretion as to which member is entitled to receive all or some of the income, or capital, of the trust.
This recommendation builds on the current rule under the trust loss provisions, where a discretionary trust cannot be classified as widely held. Its rationale is that, once there is a discretion as to the application of income or capital, it is not possible to determine who really controls that income or capital. Also, because the objects of a discretionary trust do not have an interest in either the profits or capital of the trust, it cannot be established whether 20 or fewer individuals, directly or indirectly, hold fixed entitlements to at least 75 per cent of the income or capital of the trust, or 75 per cent of the voting interests of the trust.
That an ongoing process be implemented to periodically and systematically review all tax preferences to determine whether the objectives for their introduction are:
(i) still current; and
(ii) most appropriately delivered through the taxation system.
Tax policy design principle 10, proposed in A Strong Foundation (page xxi), suggested that `Tax incentives should only be provided following a formal assessment of their net impact on the national taxation objectives and only where assessed to be an essential or superior form of government intervention.' This principle is attached to the proposed Charter of Business Taxation (see Recommendation 1.3) and should apply equally to all existing as well as proposed new tax incentives.
The Government is committed to reviewing certain tax concessions arising from special legislation - such as the Pooled Development Funds program - on a regular basis. In this report, the Review has also recommended future reviews in relation to its scrip-for-scrip and venture capital proposals. However, such processes do not cover the operation of non-program tax preferences that arise from provisions in the tax law where those provisions are not the subject of a separate administrative process.
Three classes of preferential treatment will remain:
those delivered through the income tax base;
those delivered through the entity structure; and
those delivered through classes of income being taxed preferentially.
In some cases, preferential taxation treatment can have adverse economic efficiency consequences. Special treatment can also compromise the structure and integrity of the law. Without proper evaluation, the true cost to the nation of these provisions cannot be ascertained.
The processes for the evaluation and review of tax expenditures will be facilitated through the collection of appropriate information from taxpayers -- the subject of Recommendation 3.10.