Income tax: can Part IVA of the Income Tax Assessment Act 1936 apply to an asymmetric swap scheme?

Yes, Part IVA of the Income Tax Assessment Act 1936 (ITAA 1936) [1] can apply to an asymmetric swap scheme.

This final determination will apply both before and after its date of issue. However, the Determination will not apply to taxpayers to the extent that it conflicts with the terms of settlement of a dispute agreed to before the date of issue of the Determination (see paragraphs 75 to 77 of Taxation Ruling TR 2006/10).

An asymmetric swap scheme commonly consists of two swap transactions entered into between an Australian resident company and an unrelated non-resident counterparty.

The non-resident counterparty is commonly an international financial institution that develops, trades in, and markets exposures to proprietary benchmark indexes in equities, commodities, and currencies.

The Australian resident company (for example, a financial institution) uses a business unit that is subject to a lower effective tax rate, due to the concessional treatment afforded to income and deductions, such as an offshore banking unit (OBU) [2] , prior to the 2023-24 income year [2A] , to enter into a short swap with the non-resident counterparty over an equity, commodity or currency benchmark index.

The Australian resident company uses its business unit that is subject to the normal company tax rate, such as the domestic banking unit (DBU), to enter into a long swap with the non-resident counterparty over the same equity, commodity or currency benchmark index.

A subsidiary company [3] (SPV) may be inserted between the Australian resident company and the non-resident counterparty. The SPV enters into the long swap with the non-resident counterparty. The SPV, in turn, then enters into a back to back long swap with the Australian resident company.

The long swap and the short swap are not symmetric because the Equity Notional Amount (ENA) of each swap is calculated by reference to the different tax rates applicable to the Australian resident company (DBU) and the business unit (OBU), that is, the mathematical relationship is usually fixed as follows: ENA of the short swap = ENA of the long swap x 1 - high tax rate 1 - low tax rate

The Australian resident company (that is, the DBU and OBU) will have a net long exposure, whereas the non-resident counterparty will have an identical net short exposure. The formula used to determine the ENAs is designed to ensure that the Australian resident company is not exposed to index fluctuations after tax. On a pre tax basis the Australian resident company may make profits or losses depending on the index fluctuation. On an after tax basis the Australian resident company does not make any profit or loss. [4] The impact of including other features in the structure, such as the drift adjustment and the fee, are described in paragraphs 23 to 41 of this Determination.

Prior to entering into the asymmetric swap scheme, the parties agree upon the features of the swaps, including: • The selection of a particular equity, commodity or currency benchmark index and/or its sub-indices, the performance of which is transferred or traded through the swaps. The index appears to be designed to reflect a market neutral investment strategy. The terms of the index, the sub-index selections, and the index calculations may be set out in a term sheet annexed to the swap confirmations. • Any capacity to change the benchmark index or its components over the life of the swaps. • The ENA of the long swap and the ENA of the short swap, in accordance with the formula in paragraph 8 of this Determination. • The size of the leverage factor (LF). The LF increases the effect of movements in the underlying index on the payments made under the swap. • The amount of the 'drift adjustment' percentage which is commonly based upon the historical performance of the benchmark index, for example, over a preceding period of time the index rose on average by 5%. The parties agree to impose the drift adjustment percentage at the start of the swap term. If a LF is used, it increases the effect of the drift on the payments made under the swap. • The term of the swaps. • Any caps, floors or collars restricting the range within which the swaps may trade, and limiting the upside and downside exposure to gains and losses arising from movements in the index. • Early termination rights, suspension rights, and the absence of break costs.

The drift adjustment, which is determined by reference to the historical performance of the index, may be calculated and agreed by the parties directly or may be the subject of actuarial or financial analysis by a third party prior to agreement by the parties.

The parties then execute: (i) an International Standard Derivatives Association (ISDA) master agreement, (ii) the long and short swap confirmation agreements pursuant to the ISDA master agreement, (iii) term sheets for the swaps, and (iv) any other required or agreed related contracts and documentation, such as settlement agreements and netting agreements.

Under the swap agreements there is no initial exchange of the principal amounts of the swaps, although the non-resident counterparty may provide a cash collateral deposit to the Australian resident company as security for the non-resident counterparty's maximum net exposure under the swaps.

The non-resident counterparty makes periodic (for example, quarterly) payments over the term of the short swap to the OBU. The non-resident counterparty also makes periodic payments over the term of the long swap to the DBU. The parties describe these payments as a fee to induce the OBU and DBU to enter the short swap and the long swap respectively. The fee is stated to be a percentage of the ENA of each swap. The fee is paid only during the tenure of the swap contracts.

The drift adjustment purports to have the effect that at the time the swaps are entered into, there is a neutral expectation of a profit or loss arising in either the long or short swap.

At the conclusion of the asymmetric swap scheme due to expiry of the agreed term or early termination, a single net payment is made between the Australian resident company and the non-resident counterparty, based upon the movement of the index after application of the drift adjustment to the long and short swaps.

The transaction structure involves establishing the steps detailed in paragraphs 3 to 8 of this Determination.

The effect of the structure is highlighted in the attached spreadsheet (Annexure 1). [5] This spreadsheet sets out the cash flows arising from the long swap and the short swap for a range of both positive and negative index movements. The leverage factor is incorporated into the cash flow figures. It should be noted that this spreadsheet does not incorporate the drift adjustment [6] or fees paid. [7]

The spreadsheet shows: • Any net profit before tax on the swaps is matched by the net tax paid such that the net profit after tax of the Australian resident company is always zero; • Any net loss before tax on the swaps is matched by the net tax value of losses such that the net profit after tax of the Australian resident company is always zero.

Significantly, the structure establishes an effective tax rate of 100% in relation to the asymmetric swaps for the Australian resident company. Annexure 1 shows that if the index: • Increases [8] - then the profit [9] derived on the transactions is matched in full by the tax payable; [10] or • Decreases [11] - then the quantum of the loss [12] on the transactions is matched with the tax value of losses of the same amount. [13]

Moreover, a critical feature of the structure is that it provides the Australian resident company with an after tax built-in hedge by ensuring the tax outcome of the structure eliminates any profit or loss that can be made on the index. [14]

The asymmetric swaps are similar to equity swaps in that they transfer the economic return on an index. The major differences are: • The asymmetric swaps do not have a fixed or floating interest rate leg adjusted by adding or subtracting a margin from the interest rate payments; • The asymmetric swaps are over an index the components of which are determined by the non-residents trading models; • Equity swaps do not incorporate a drift adjustment; • Fees are not normally paid in equity swaps.

Besides the movement in the index, the two important additional features that are overlaid on the transaction structure are the drift adjustment and the fees.

The drift adjustment is calculated by reference to the historical performance of the benchmark index and is purported to remove a directional bias in the index. The non-resident counterparty provides the Australian resident company with the data (either actual and/or back tested) that reflects the past performance of the index over a designated period of time. The Australian resident company then determines, either in-house or through an actuary, the appropriate drift percentage.

The annual drift adjustment is determined by multiplying the ENA of the swap by the leverage factor and the drift percentage. The cash flow resulting from the daily index movement is adjusted by the daily drift amount. Moreover, the cash flow resulting from the index movement over the term of the swap is adjusted by the annual drift adjustment (being the aggregate of all daily drift adjustments). For example, on the long swap if the index rises by for example, 3%, absent the drift adjustment, a cash flow of 3% of the leveraged ENA would be paid by the non-resident counterparty to the Australian resident company, but when the drift adjustment is applied, for example, 5%, then 2% (index ? of 3% less drift of 5% = minus 2%) of the leveraged ENA would be paid by the Australian resident company to the non-resident counterparty (that is, in the opposite direction).

Although the drift is referred to as an 'adjustment', economically it is a payment obligation [15] being a fixed and determined amount taken on by the relevant party at the start of the swaps and adjusted over the term of the swaps and included in the calculation of the final net payment at the termination of the swaps. For the long swap, it represents a movement of funds from the Australian resident company to the non-resident counterparty. For the short swap, it represents a movement of funds from the non-resident counterparty to the Australian resident company. On a net basis, the drift adjustment represents a payment by the Australian resident company to the non-resident counterparty.

It has been suggested that a purpose of the drift adjustment is to ensure that at the time that the swaps are entered into, there is a neutral expectation of the value of the benchmark index rising or falling over the term of the swaps by removing any initial bias for profits to accrue in either the Australian resident company's DBU or its OBU.

However, past performance provides no assurance that it will be repeated in the future. At day 1 of the swap the shares [16] that comprise the index will have a market price. That market price will reflect the historical data concerning price changes and the markets' view, based on information available at that time, as to future expectations as to how the stock will perform. The past performance of the shares is already embedded in their market price.

The Australian resident company is taking on the risk of the actual performance of the equity index and should be entitled to the full return on the index. Payment of the drift effectively removes the expected return whilst maintaining the same level of risk for the Australian resident company.

A Drift factor is not used in conventional equity swaps.

The effect of the drift adjustment on the transaction structure is shown in Annexure 2. Like the index movement, the drift effect is mirrored in the long and short swaps. Accordingly, the structure's effective tax rate of 100% is maintained. The major changes arising from the drift adjustment are: • Any net loss before tax is increased by the net drift effect, and the net tax value of the losses is similarly increased such that the net profit after tax is always zero. • Any net profit before tax is decreased by the net drift effect, and the net tax paid is similarly decreased such that the net profit after tax is always zero.

The payment of the drift represents a constant sum. [17] The net drift [18] is always a negative figure. The Australian resident company pays the net drift over the term of the swaps and because of the design features of the transaction structure is fully compensated for this payment by the tax outcome. Irrespective of whether the swaps are in profit or loss, the Australian resident company is fully compensated for the payment in the form of the deductions allowed and/or the tax value of the losses the payment creates.

In an asymmetric swap scheme, the non-resident counterparty makes a payment to the Australian resident company (DBU and OBU) on both the long swap and the short swap. The payment is calculated as the aggregate of a negotiated percentage of the ENA of the long swap and short swap. It has been stated that the reason the non-resident counterparty is willing to make a payment on both of the swaps is that the counterparty is seeking to hedge a long exposure it has in its proprietary trading business. To obtain the hedge, the non-resident counterparty is prepared to make a payment on both of the swaps to obtain the net short position from the combination of the two asymmetric swaps.

Commercially, the fee would be expected to be a percentage calculated by reference to the net short position obtained by the non-resident counterparty, that is, the ENA of the long swap less the ENA of the short swap, not a negotiated percentage of the ENA of both the long and short swaps. Accordingly, the fee is not commercially priced. In a commercial swap transaction it would be expected that if a fee was paid in one direction on the long swap by one party, then a fee would be expected to be paid in the opposite direction by the other party on the short swap.

Notwithstanding the non-resident counterparty's stated purpose in paying the fee, the Australian resident company will not enter the combined swaps unless it receives a payment (however described) on both swaps. Although it may make profits or losses on a pre-tax basis, the Australian resident company has no after tax exposure and always receives a constant profit after tax. In the absence of the fee payment there is no reason for the Australian resident company to enter into the asymmetric swap scheme at all.

The drift and the fee are connected. Under the asymmetric swap scheme the non-resident counterparty's receipt from the transaction is always the drift less the fee (less any transaction costs). That amount will not change irrespective of any movement in the equity index.

Similarly, the Australian resident company's return from the scheme will always equal the after tax fee income received. On an after tax basis the effect of any movement in the index will be sheltered by the structure.

The quantum of the fee is always less than the drift. Where profits arise the fee is preserved after tax but every $1 of profit above the after tax fee amount is paid, in full, to the revenue authority; where losses arise the fee is also preserved after tax due to the tax value of the losses.

The structure sets up two swaps that mirror each other (albeit with asymmetric ENAs). Movements of the index are mirrored in the long and the short swaps; the drift adjustment is also mirrored in the long and the short swaps. If the fee payments were mirrored in the long and short swaps, in all cases the net profit after tax figure (NPAT) would be nil. Accordingly, there would be no incentive for the Australian resident company to enter the arrangement. However, because the fees are paid in one direction on both swaps, the Australian resident company receives an incentive to enter the scheme and pay the drift adjustment. [19]

The effect of the payment of the fees is set out in Annexure 3. The main features are: • The net profit after tax figure is a constant sum [20] irrespective of the direction of the index movement. • In the scenario in which the Index - Drift + Fee = 0, [21] the Australian resident company has a net profit before tax of zero but has a net profit after tax of the constant sum. • In the scenario in which the Index - Drift + Fee = NPAT, [22] the Australian resident company has a net profit before tax of the constant sum, it has a tax liability of zero but has a net profit after tax of the constant sum. • The effective tax rate varies from 100% dependent upon whether the swaps are in profit or loss. [23] • The Australian resident company obtains the advantage detailed at paragraph 20 of this Determination in relation to the index movement, and at paragraph 32 of this Determination in relation to the drift payment, as well as, a constant after tax fee.

There are a number of features of the asymmetric swap scheme that suggest that the pricing does not conform to accepted commercial methodologies and practices. These are: • Ordinarily, the receiver on the long swap would receive all upside from increases in the index at current market prices and be exposed to the risk of losses from any decrease in the index. However, in the asymmetric swap scheme the drift adjustment alters this outcome. • An equity swap will have two legs, the equity index return leg and the fixed or floating interest rate leg adjusted by adding or subtracting a margin from the interest rate payments. The interest rate legs are at market prices as they will be determined by current interest rates at the time the swaps are entered into. No interest rate leg is present in the asymmetric swap scheme. • The risk of providing the hedge to the non-resident counterparty is not assumed by the Australian resident company, it transfers that risk to the revenue authority through the structuring of the transaction. • The caps, floors and collars detract from the swap as the party does not get the full upside or downside movement, these features would usually be separately priced at market rates and paid for by the party requesting their inclusion (the Australian resident company), however there is no indication that any of the participants carried out that process and itemized such costs. • A party seeking a hedge would normally be expected to pay an arms length amount based upon the actual exposure it is seeking to hedge, that is, ENA of the long swap less ENA of the short swap. [24] The decision to base the payment amount on the aggregate of each ENA [25] is not tied to the actual exposure sought to be hedged. These factors indicate that the asymmetric swaps are not commercially priced, and that the long swap and the short swap would not be entered into on a stand alone basis.

Part IVA can only apply to the asymmetric swap scheme if it has its intended effect under Division 9A of Part III (Division 9A). Division 9A provides for the taxation treatment of OBUs. However, this final Determination does not analyse the operation of Division 9A. For present purposes it is assumed (but not ruled) that the scheme is effective.

The application of Part IVA necessarily requires consideration of the particular facts of each case. For structured financial transactions, like the asymmetric swap scheme, there are common features in the transactions that are present for all participants who execute the arrangement. [26] This analysis examines those common features and comes to a conclusion as to whether the provisions of Part IVA can apply. This analysis is undertaken with the proviso that an individual transaction may have circumstances that impact upon the ultimate conclusion as to whether Part IVA applies.

For Part IVA to apply there must be a scheme as defined in section 177A; a taxpayer must have obtained a tax benefit in connection with the scheme under section 177C; and it would be concluded that at least one person who entered into or carried out the scheme did so for the sole or dominant purpose of obtaining a tax benefit having regard to the factors in section 177D.

The definition of scheme in subsection 177A(1) is drafted in wide terms to mean, relevantly, 'any agreement, arrangement, understanding, promise or undertaking ...', and 'any scheme, plan, proposal, action, course of action or course of conduct'. For the purposes of section 177A, the asymmetric swap scheme includes all of the steps taken to effect the implementation and carrying out of long and short swaps, involving the Australian resident company's DBU and OBU and the non-resident counterparty, including: (a) any promotion or canvassing or marketing of the asymmetric swap schemes, and any other preliminary steps leading up to the swaps being entered into; (b) the entering into or carrying out of the steps specified at paragraphs 5 to 7 and 10 to 16 of this Determination, in particular the entering into of the long and short swaps and the agreement as to the amount of the drift factor and the fees; (c) the termination or early termination of the swaps; and (d) the incurring of any net loss in relation to the long and short swaps by the Australian resident company.

The relevant tax benefit will be that identified by paragraph 177C(1)(b) but may, in certain circumstances, extend to a tax benefit identified under paragraph 177C(1)(a). [27] A taxpayer obtains a tax benefit in connection with a scheme for the purposes of paragraph 177C(1)(b) if a deduction is allowable to the taxpayer where the whole or part of that deduction would not have been allowable, or might reasonably be expected not to be allowable, to the taxpayer in relation to the year of income if the scheme had not been entered into or carried out. The application of the test involves a prediction as to events that would have taken place if the relevant scheme had not been entered into or carried out and the prediction must be sufficiently reliable for it to be regarded as reasonable. Thus the enquiry directed by Part IVA requires comparison between the scheme in question and an alternative postulate. A particular application of the definition provision of 'tax benefit' in subsection 177C(1) involves consideration of the particular materials answering the various categories in paragraph 177D(b). To draw a conclusion of purpose from the eight matters identified by paragraph 177D(b) will require consideration of what other possibilities existed. [28]

Having regard to the features of the scheme and the transaction structure as outlined in paragraphs 17 to 21 of this Determination, the drift adjustment is commercially unjustifiable. A simple approach might be to identify the drift adjustment as the tax benefit because it is a fixed and determined amount that is agreed upfront. Cancellation of the tax benefit would simply involve excising the drift adjustment from the transaction. Depending on how the index moves, the drift adjustment will either decrease the Australian resident company's assessable income or increase its allowable deductions and the taxpayer will obtain a tax benefit within the meaning of paragraphs 177C(1)(a) or 177C(1)(b) respectively.

However, an asymmetric swap transaction is a specific structured transaction that is not part of, and does not consist of some of the steps in, any broader commercial transaction. A reasonable person might expect that if the scheme had not been entered into or carried out, no similar transaction, or perhaps any transaction, would have been entered into or carried out. In particular, one would not expect the parties to enter into an asymmetric swap which lacked the drift adjustment. Without the inclusion of the drift on the asymmetric swap scheme, the payment of the fees at non-commercial rates to the Australian resident company on both swaps would not be possible as set out in paragraphs 33 to 40 of this Determination; and the non-resident counterparty could not have obtained the net short exposure to the index at the non-arms length price obtained in the asymmetric swap scheme as set out at paragraph 41 of this Determination. The scheme would therefore lack its raison d'être.

As the Australian resident company is hedged on an after tax basis, the asymmetric swap scheme can have no effect or outcome other than obtaining a tax benefit, (or in some rare cases a tax detriment), and it is reasonable to assume nothing would have happened if the asymmetric swap scheme had not been entered into or carried out. [29]

Where the drift exceeds the index, a deduction is allowable to a taxpayer as a result of the scheme and it follows that the taxpayer will have obtained a tax benefit in connection with the scheme pursuant to paragraph 177C(1)(b) being the whole amount of the deduction claimed.

However, it is necessary to observe that there is a possibility (albeit remote) that the index movement may equal or exceed the drift percentage over the term of the swaps. In such a case, although strictly speaking Part IVA may still apply to the scheme (because it may have been the purpose of a participant in the scheme to obtain a tax benefit in connection with it, though none was obtained) there will be nothing to cancel under section 177F. However, this circumstance is not, in itself, determinative of whether or not the requisite purpose exists to obtain a tax benefit.

Where a taxpayer has obtained a tax benefit in connection with a scheme under paragraph 177D(a), the eight factors in paragraph 177D(b) must be considered to reach a conclusion as to whether at least one of the participants entered into or carried out the scheme with the purpose of enabling a taxpayer to obtain a tax benefit in connection with the scheme. If after a consideration of these eight factors, which are posited as objective facts, [30] it would be concluded that the person or one of the persons, who entered into or carried out the scheme did so for the purpose of enabling the relevant taxpayer to obtain a tax benefit in connection with the scheme, Part IVA will apply.

A reference to purpose refers to the 'dominant' purpose (subsection 177A(5)), being that purpose which was the ruling, prevailing or most influential purpose. [31] In the following analysis it is the purpose of the Australian resident company that is primarily examined. A conclusion drawn from the eight factors in paragraph 177D(b) requires comparison between the scheme and an alternative postulate or counterfactual scenario. [32]

A scheme may fail to achieve the purposes of its participants, and thus, as noted in paragraph 51 of this Determination, the mere existence of the possibility, at inception, that a tax benefit might not be obtained by anyone will not preclude an adverse conclusion as to purpose under section 177D. Where the possibility is remote, as here, it is of little weight in disestablishing the existence of a prevailing purpose of obtaining a tax benefit. Where that remote eventuality appeared likely to happen, should the parties choose to terminate the scheme, the action (itself a scheme) of terminating the original scheme would point to the existence of a prevailing purpose of obtaining a tax benefit, under paragraph 177C(1)(a).

There are a number of objective factors that support the view that the scheme may be designed to avoid an outcome where the index movement is greater than the drift: (i) the market neutral investment strategy; (ii) the resident company agreed (via the drift adjustment) to pay its expected return to the non-resident counterparty over the term of the swaps; (iii) the resident company had an incentive to agree to a larger drift adjustment because the fee was calculated as a percentage of the drift; (iv) in some cases, the cherry-picking of the historical data periods to select the drift percentage; (v) there is no commercial basis to impose the drift (that is, the transaction structure needs to be in place for the drift to be paid); (vi) the lack of certain commercial features, for example, pricing, the inclusion of non-commercial features such as the drift; and the absence of break costs, that impact on the normal profit making dictates of business dealings; (vii) the indifference of the resident party to make any profit above the constant sum; and (viii) the ability to terminate the transaction at one days notice with no break costs (in particular, where the non-resident counterparty is out of the money).

On an objective analysis of the facts, the dominant purpose of the resident company is to obtain a fixed return (which is sourced from the agreed drift adjustment). The resident company is indifferent to movements in the index because any profit or loss on the index is eliminated after tax and therefore could not be viewed as a dominant purpose of the resident bank in entering into the transaction. In particular, the resident company is indifferent to making any profit above the fixed return. [33]

The scheme is designed to ensure that the risk and exposure of the hedge is effectively passed from the swap parties to the revenue authority. This was the critical component of the scheme for the resident party. The resident party would not enter the swaps if this feature was not in place. [34] From a purpose perspective this was the most crucial feature of the scheme.

Paragraphs 59 to 74 of this Determination consider the eight categories in paragraph 177D(b).

Manner involves consideration of the ways, methods, and/or procedures by which, the particular scheme was carried out in comparison to the counterfactual, to enable contrivance and artificiality such as additional steps or complications to be identified. [35] This structured financial transaction was created by a foreign investment bank and is tailored for a specific market, namely, companies resident in tax jurisdictions in which the opportunity exists to utilise different tax rates within the jurisdiction. The parties agreed from the outset that a drift adjustment would be imposed (a feature of these swaps that does not appear in any other commercial swap arrangement) and that fees would be paid to the Australian resident company on both the long swap and the short swap. The parties also either decided upon or negotiated the variables relevant to the effective operation of the scheme, for example, ENA amounts, drift adjustment, the quantum of fees and the leverage factor. These variables are linked by mathematical formulae that predetermine the results of the transaction.

A swap for the 'net' long position would achieve the same commercial result for the parties but eliminate the tax benefits obtained from the structure. The bifurcation of a single swap into a long swap (DBU) and a short swap (OBU) through the application of the mathematical formula set out in paragraph 8 of this Determination has an air of artificiality and contrivance.

The separation of the swap contracts is pivotal to the structuring of this swap transaction, that is, the tax attributes of the structure are lost if the contract activities are placed wholly within either the OBU or DBU.

The long and short swap contracts are closely linked and dependent upon one another, that is, the contracts are conflating. This is based upon the contracts: • being entered into on the same date, for the same term, over the same benchmark index and sub-indices; • were negotiated together; • are identical mirror positions but for the ENA's and the Australian resident company receiving fees on both swaps; • if one swap terminates the other swap must terminate; and • both contracts need to be on foot for the Australian resident company to: (i) obtain the after tax hedge; and (ii) to transfer the risk associated with that hedge from the parties to the revenue authority.

This factor indicates a purpose of obtaining a tax benefit evident from the complex structuring involved in this financial arrangement and the creation of artificial and contrived tax aspects for example, a 100% effective tax rate and transfer of hedging risks and exposures to the revenue authority.

This factor is directed to examining the scheme for discrepancies between the form of the scheme and its substance, particularly commercial or economic substance. The scheme takes the form of complex interrelated swap arrangements that use a benchmark index the performance of which is transferred or traded through the swap. The swaps are designed to provide a hedge to the non-resident counterparty. The form is overly complex as it could be achieved through a single swap. From a pre-tax standpoint, the ramifications of entering into the scheme are commercially real, if the index rises the participant will be exposed to an economic profit, if the index falls the participant will be exposed to an economic loss that is, the Profit and Loss of the company is affected by movements in the index. To that extent form equates to substance.

However, from a 'substance' standpoint, the index is irrelevant to the scheme. The participants are not exposed to the movement of the index on an after tax basis. In reality the scheme imposes a feature - the drift adjustment - that is not present in any other commercial swap arrangement. The practical effect of which is to mis-price the swap by sending to the non-resident counterparty funds that, in part, are returned to the resident company in the form of the fees on the two swaps. Any cash flow outgoings by the Australian resident company (for example, index movement and drift adjustment) are fully compensated due to the financial structuring of the scheme; that is, the tax value of expenditures (normally 30% of each $1.00 of outlay) is inflated to parity (that is, 100% of each $1.00 outlay).

At inception the parties have agreed on all of the variables. The after tax position is then preordained, irrespective of the index. The after tax position is always the same, whether the index increases by 1% or 8% or falls by 1% or 8%, that is, the non-resident counterparty always receives the drift less fees on the swaps; the resident company always receives the fees on the swaps. The scenarios detailed at dot points 2 and 3 of paragraph 40 of this Determination shows the disconnection between the form and substance of the scheme. Further, the scheme inflates the tax value of any losses obtained from the transaction. These features indicate a dominant purpose on the part of the Australian resident company of obtaining a tax benefit.

The timing as to when the scheme was carried out is a neutral indicator as to purpose. The length of the swap arrangement varies for individual taxpayers (1-5 years), an equity swap would not normally be on foot for an extended timeframe (>1-2 years). The ability of participants to terminate the arrangement, often at one days notice, with no break costs is not a normal commercial feature of swap deals.

The result in relation to the operation of the Act that would be achieved by the scheme is that certain amounts of income are not included in the Australian resident company's assessable income by being moved from it's DBU to it's OBU, [36] or deductions are allowable that would not have been allowable. [37] Moreover, the tax value of expenditures (normally a 30% credit for every $1 outlay) is inflated. This indicates a purpose of obtaining a tax benefit on the part of the Australian resident company.

The financial position of the Australian resident company is changed in that the payment of the drift in concert with the fees received, either reduces assessable income in circumstances where the index increases, or inflates deductions (in the DBU) in circumstances where the index falls. Moreover, the available tax value of losses is artificially inflated by the imposition of the drift in concert with the fees paid. The underlying structure provides the Australian resident company a risk free hedge (although that risk and the resulting financial exposure are passed on to the revenue authority).

The non-resident counterparty's financial position is enhanced as it receives: • a fixed amount of funds (the drift); and • the short position hedge, At the cost of paying: • a fee.

The SPV receives a fee for facilitating the long swap between the resident and non-resident counterparty.

The Australian revenue bears the risk of the hedge provided by the domestic party to the non-resident counterparty.

There are no other consequences.

The parties are unrelated but deal with each other on non arm's length terms. In the course of their banking business the parties may have had ongoing dealings with each other in global banking markets. This factor points to normal commercial practice. The parties may also have the relationship of selling and buying, respectively, tax based financial products or other tax avoidance schemes. That circumstance will tend against accepting that normal commercial practice is present. However, the circumstance that the asymmetric swaps are not priced on an arms length basis and in accordance with commercial pricing methodologies will tend to indicate that the parties are colluding to achieve a purpose ulterior to the normal commercial purposes for which swaps are entered into. This feature indicates a 'tax driven' purpose.

Accordingly, it should be concluded that one or more of the persons who entered into or carried out the scheme did so for the dominant purpose of enabling the Australian resident company to obtain a tax benefit in connection with the scheme.

This analysis has proceeded on the basis that the benchmark index exists, that it is actively traded and that there is no inherent bias in the scheme, through either the structure, or the constitution of the index. If such evidence came to light then the Part IVA analysis would be strengthened.

It follows that Part IVA will apply to the scheme with the result that the power to cancel tax benefits obtained in connection with it is enlivened. The Commissioner is empowered to cancel the whole or part of the tax benefit. In general the Commissioner's approach as to whether to cancel the whole or part of the tax benefit obtained by the taxpayer is to reconstruct the tax position that might have been expected if the scheme had not been entered into so as to reinstate fully the revenue lost through tax avoidance. In some circumstances when an allowable deduction tax benefit represents an actual outlay of the taxpayer's own money having some, if not a prevailing, income-producing purpose, and the remaining part does not represent an economic outlay, the Commissioner will determine to cancel only the remaining part of a tax benefit. (In effect, the real tax benefit is taken to be so much of the deduction that is obtained without economic outlay.) On the other hand, a transaction without any purpose except that of obtaining a tax benefit will generally involve a strict application of Part IVA, even though it involves the expenditure of real moneys, and it might be expected that tax benefits will be cancelled in their entirety; the schemes here in consideration fall in this category. However the actual decision in regard to a particular scheme will depend on that scheme.