“CASTLES ON A CLOUD” – DEDUCTIBILITY OF FEASIBILITY EXPENDITURE
Vivian Cheng – Chapman Tripp, Wellington
Geoff Clews – Barrister, Auckland
Doubts raced through my mind as I considered the feasibility of enforcing a law which the majority of honest citizens didn’t seem to want – Eliot Ness
“Feasibility expenditure” is a term generally used to describe expenditure incurred to determine the practicality of a new proposal. Typical examples include expenditure incurred on engineering surveys, environmental studies, market research, and obtaining relevant professional advice.
The deductibility of feasibility expenditure is an important issue for any business that incurs costs in investigating or exploring whether the acquisition or development of new assets is practical or possible. Most businesses look to grow. Growth will usually require increased production (or the equivalent). Increased production will require increased capacity achieved in one way or another. The heart of the feasibility expenditure issue is whether and to what extent the cost of evaluating options for business growth is deductible. It is an issue that affects every business aspiring to grow.
The issue has become topical in recent times in light of the Court of Appeal’s decision in Commissioner of Inland Revenue v TrustPower Limited  NZCA 253, delivered on 19 June 2015. The approach taken by the Court of Appeal is departs from the accepted practice in New Zealand and from the approach endorsed by the Commissioner’s 2008 Interpretation Statement IS 08/02: Deductibility of Feasibility Expenditure (“the Interpretation Statement”).
In this paper we set out the legislative framework and the accepted approach to deductibility of feasibility expenditure prior to the TrustPower case. We then examine the High Court and Court of Appeal decisions in light of those in an attempt to make some sense of the decisions and to assess their implications.
There are no specific rules in the Income Tax Act 2007 (“the Act”) for feasibility expenditure. Deductibility is therefore determined in accordance with the general permission in s DA 1 and subject to the general limitations in s DA 2.
The general permission in s DA 1 allows a taxpayer a deduction for expenditure or loss incurred in deriving assessable or excluded income, or incurred in the course of carrying on a business for the purpose of deriving such income. The general permission is, however, overridden by the general limitations in s DA 2 to the extent they apply. Of particular relevance to feasibility expenditure is the capital limitation in s DA 2(1), which denies a deduction to the extent that the expenditure or loss in question is “of a capital nature”.
The analysis of deducibility of feasibility expenditure is therefore a two-step process. First, is there a sufficient nexus between the expenditure and the taxpayer’s income or income-earning business? Secondly, if a sufficient nexus exists, is the expenditure of a capital nature so as to be non-deductible? At least that two-step process has been acknowledged and followed prior to the Court of Appeal’s decision in TrustPower.
The Interpretation Statement has formed the basis of accepted practice for deductibility of feasibility expenditure in New Zealand since it was issued on 4 June 2008. It confirms the two-step approach to deductibility of feasibility expenditure and draws on relevant case law from New Zealand, Australia and Canada.
General permission and the nexus with income
In line with the legislative framework, the Interpretation Statement states that the first requirement for obtaining a deduction for feasibility expenditure is that the expenditure must be incurred in the course of the taxpayer’s normal business operations. The expenditure must arise as an ordinary incident of the business or income-earning activity. The correct characterisation of the nature of the relevant business is therefore vital to resolving whether a sufficient nexus exists between the expenditure and the taxpayer’s business/income-earning activity.
The Interpretation Statement’s discussion on whether a sufficient nexus exists focuses on whether the expenditure is preliminary or preparatory to the commencement of a business or, in the case of a taxpayer already in business, related to a new business or income-earning activity. All of the examples given by the Interpretation Statement of circumstances in which the requisite nexus is not established involve expenditure incurred in investigating the feasibility of a completely new business, or a new income-earning activity of an existing business that is fundamentally different from the existing income-earning activity (eg a logging and saw-milling company investigating the feasibility of producing garden tools for sale).
It is implicit in the Interpretation Statement that expenditure incurred by a taxpayer in an existing business will prima facie satisfy the nexus requirement unless the expenditure relates to a new income-earning activity. This is consistent with case law and accepted practice, though there is at least one instance of commentary doubting that.
Capital limitation and the “commitment” test
If the nexus requirement is satisfied, the Interpretation Statement states that consideration must be given to whether the expenditure incurred is capital or revenue in nature. Again, this approach is in line with the legislative framework.
The Interpretation Statement acknowledges that expenditure incurred to evaluate a proposed course of action may be revenue in nature notwithstanding that the proposal being evaluated, if implemented, would give rise to a capital asset. The Commissioner considered that the principles applied in Canadian cases such as Bowater Power Co Ltd v Minister of National Revenue  CTC 818 are consistent with New Zealand and Australian authorities. The taxpayer in Bowater was in the business of generating and selling electrical power and energy. The Canadian Federal Court (Trial Division) found that expenditure incurred by that taxpayer on surveys/engineering studies relating to the development of additional power and the location of physical plant for a power station was revenue in nature.
The Interpretation Statement accepts that feasibility expenditure incurred principally for the purpose of placing a taxpayer in a position to make an informed decision about the acquisition of a capital asset will not generally be expenditure incurred in relation to that particular asset. That is presumably because the inquiry towards which the expenditure is applied might just as well lead to a decision not to acquire the asset but to adopt some alternative stratagem. Furthermore, if the feasibility expenditure is incurred principally for evaluating one or more proposals, it is unlikely the expenditure will relate to the business structure sufficiently to be capital in nature.
The Interpretation Statement states the point at which expenditure incurred alters from being related to the evaluation/exploration of options (ie revenue) to being related to the development or acquisition of an identifiable capital asset (ie capital) is when a commitment or decision has been made to proceed with the acquisition or development of the capital asset. Once a decision has been made to proceed, any expenditure incurred after that time will be treated as related to the capital asset and will not be deductible. This is the case irrespective of whether the expenditure is successful or not.
Referring to Justice Wild’s judgment in Milburn NZ Limited v Commissioner of Inland Revenue (2001) 20 NZTC 17,017 (HC), the Commissioner states at paragraph 126 of the Interpretation Statement:
in the capital versus revenue context in relation to an existing business a distinction may be drawn between amounts expended on initial investigations to determine possible prospects and amounts expended once a decision to proceed with any prospect in particular has been made. Once a decision has been made to proceed with the acquisition or development of a capital asset, it seems that expenditure is considered to be incurred on the business structure rather than the income-earning process. It is also noteworthy that Wild J compared the three sites in question with the other 48 prospects that were investigated. This suggests it is relevant to consider the frequency with which expenditure of the type under consideration is incurred in the ordinary course of the taxpayer’s business
The Interpretation Statement goes on to say that a taxpayer will have made a commitment or a decision to proceed when preliminary work has been completed with indications that the development or acquisition of an asset is technically and financially viable, and the facts indicate that the development or acquisition is proceeding. Although the point at which that occurs is a question of fact and degree, in our experience most taxpayers regard as deductible feasibility expenditure what they incur in considering the possibility of capital investment in the period up until their board or management decide to commit to the investment.
The Interpretation Statement also highlights the relevance of the feasibility expenditure in question being of a type that is incurred frequently as part of the constant demands on the business. This is reflected in the passage quoted above as well as the examples given in the Interpretation Statement of when feasibility expenditure can be deducted immediately. One of the examples involves feasibility expenditure incurred by a quarry company that continually investigates potential quarry sites for the purpose of gaining reward from any site ultimately developed. Another example involves a quarry company that investigates 20 potential sites for development with the board approving the development of two of the sites (the Commissioner considering in that case that the expenditure incurred on the 18 sites that were not selected for development was on revenue account).
The Interpretation Statement is not universally applauded. In an article earlier this year John Prebble and Hamish McIntosh criticised it on a number of grounds. These included the inadequacy of the argument that, prior to commitment, feasibility expenditure is capable of revenue account treatment and that the Interpretation Statement conflates the relevant questions related to pre-commencement expenditure (non-deductible in any event) with pre-commitment expenditure.
They pose the proposition that it would seem odd if expenditure to improve an existing asset is capital, but expenditure to decide whether to acquire or construct the asset at all is revenue. They respond to the proposition that the latter might just as well lead to other business responses being preferred to the acquisition or construction of an asset by saying that there is no doctrinal justification for considering that such expenditure is on revenue account. This seems to suggest that, even if undertaken in the context of an existing business and with a frequency that would suggest revenue account activity, expenditure that might lead to an asset acquisition or construction is never on revenue account. This view was published just prior to the Court of Appeal’s hearing on TrustPower.
The facts of the case are well-known. TrustPower is an electricity retailer. The electricity it sells is either generated by it or purchased on the wholesale market. It generates (by hydro or wind) about half of the electricity it sells.
In order to maintain flexibility as to whether it generates or buys electricity in the future, TrustPower regularly sought and obtained resource consents for potential generation projects. This “development pipeline” was continually evolving and comprised at least 200 projects at varying stages of feasibility assessment at any one time.
The development pipeline is intended to provide TrustPower with information as to the viability, feasibility, and costs of building new generation capacity, so that it can decide at any one time whether to build a new generation project or buy electricity on the market. The options available to the company for its expansion lay each side of the capital/revenue divide: build more capacity or buy more “stock”. The projects were often reassessed against each other and alongside the option of continuing to purchase electricity. There were always more projects in the pipeline than TrustPower had the financial and resource capability to construct. There was no assurance that any particular project would proceed. TrustPower did not own (or at most owned only a portion of) the property relating to the projects.
Expenditure at issue
TrustPower incurred $17.7 million of expenditure in the 2006 to 2008 tax years in taking preliminary steps and then applying for and obtaining various resource consents in relation to four possible hydro and wind generation projects in the development pipeline. The resource consents involved were land use consents, water permits and discharge permits. Some of the land use consents were granted for an unlimited term, but other consents were for fixed periods of generally 10, 15, or 35 years.
The expenditure at issue was incurred as part of the first step in TrustPower’s three-step feasibility analysis process. It included expenditure incurred on obtaining extensive project viability assessments (including assessments of environmental effects) before TrustPower made a firm decision to proceed with the resource consent applications, which formed part of the applications once the decision was made to proceed.
TrustPower’s characterisation of its feasibility analysis process, which was accepted by the High Court on the basis of unopposed evidence from the company, is as follows:
(1) Step 1 - Identify potential generation sites and conduct a high-level evaluation of the feasibility of the site (applying for resource consents is the third and final stage of this step);
(2) Step 2 - Carry out more detailed work such as evaluating what can be constructed, developing civil engineering and electromechanical designs, calling for tenders from manufacturers and analysing the economic feasibility of the project more carefully; and
(3) Step 3 - Prepare a business case for consideration by the Board.
The evidence from TrustPower was that until it has resource consent in relation to a particular project, it does not know what it can construct and what limitations will be placed on the scheme operations and therefore cannot determine whether the project is viable when compared with other possible options. The resource consents reflect only in part what is a preliminary design concept and TrustPower’s evidence is that at the point of making the relevant application the engineering design would only be about 10% complete. Resource consents are therefore necessary before a decision can be made to proceed to the second step of the feasibility analysis process.
History of the dispute
The Commissioner issued notices of proposed adjustments in 2010 advising she would disallow the deductions claimed by TrustPower in the 2006, 2007 and 2008 tax years.
The dispute was referred to the Adjudication Unit, which considered the issue for the 2007 and 2008 years (reassessments for the 2006 year were issued in September 2011 before the Adjudication Unit could give its decision). The Adjudication Unit decided that the expenditure was not capital in nature because TrustPower had not committed (in the sense of a formal documented decision by the Board) to proceed with the acquisition or development of any of the proposed projects in the relevant years. The Adjudication Unit went on to consider whether the resource consents were capital assets and concluded that the consents could be on capital or revenue account depending on the facts, but would in most cases be capital. The Adjudication Unit was unable to determine which consents were of a capital nature or to allocate costs and returned the dispute to the Service Delivery Group.
The Commissioner subsequently issued reassessments to TrustPower in March 2012, disallowing the deductions claimed in the 2007 and 2008 years. The total amount at issue over the three tax years is approximately $10.6 million of tax and $4.3 million of interest.
TrustPower issued proceedings in the High Court, which was heard on 12-30 August 2013. Justice Andrews delivered judgment on 12 November 2013 in favour of TrustPower. The Commissioner appealed. The Court of Appeal hearing took place on 23-25 March 2015 and judgment of the Court was delivered by Justice White on 19 June 2015.
TrustPower submitted that the expenditure was incurred in the ordinary course of its business of deriving income from the generation and sale of electricity and therefore on revenue account. The expenditure brought about the opportunity to make a continuing assessment of whether to build capacity or to buy electricity.
TrustPower argued that the resource consents were not stand-alone assets. They were obtained as part of the overall feasibility process for assessing possible generation projects and should not be viewed separately. The expenditure was therefore deductible feasibility expenditure indistinguishable from recurring operational expenditure incurred to source electricity for resale.
TrustPower submitted that, even if the resource consents were stand-alone assets, they were held on revenue account. If, contrary to its submissions, the resource consents were found to be capital assets, TrustPower said that it did not commit to applying for the consents until shortly before the applications were filed, with the result that expenditure incurred up until that point was deductible feasibility expenditure.
On appeal, TrustPower challenged the Commissioner’s reliance on the provisions of the Act relating to “depreciable intangible property”, on the basis that they are not applicable when the general permission is relied on for deduction. TrustPower also objected to the Commissioner’s characterisation of the expenditure as being of a capital nature and her identification of the date when TrustPower committed to acquiring the resource consents, on the grounds that they were contrary to the factual findings made by Justice Andrews in the High Court.
The Commissioner’s position
The Commissioner submitted that the resource consents were “depreciable intangible property” for tax purposes and therefore capital assets. Expenditure incurred on them was on capital account and non-deductible from the time TrustPower committed to applying for the consents.
The Commissioner argued that, in any event, the expenditure was of a capital nature and subject to the capital limitation. The resource consents for the projects were “long life consents” providing an enduring benefit of “strategic optionality” to TrustPower (ie the ability to proceed immediately with development of the project within the consent period, to defer construction within that period, to block or exclude other generators from undertaking competing activities, or to sell the resource consents if the project ceases to be feasible or if the price is right).
The Commissioner considered that TrustPower committed to applying for the resource consents (ie the expenditure became on capital account) when TrustPower decided to engage consultants to produce assessments of environmental effects, which were in a format suitable for inclusion in a consent application. The Commissioner classified $6.6 million of the $17.7 million disputed expenditure as expenditure incurred as part of the “consenting” process rather than a part of the “feasibility” process.
On appeal, the Commissioner also argued that the High Court erred in not treating the resource consents as stand-alone assets.
Justice Andrews in the High Court approached the case in terms of four principal issues:
(1) Are the resource consents assets on a stand-alone basis separate from the projects to which they relate? If not, then the expenditure incurred in obtaining them is deductible feasibility expenditure because TrustPower had not committed to the projects.
(2) If the resource consents are stand-alone assets, are they capital or revenue assets? If they are capital assets, then the expenditure incurred in obtaining them is not deductible. If they are revenue assets, the expenditure is deductible.
(3) If the resource consents are capital assets, was the Commissioner wrong in her conclusions as to the dates from which TrustPower was committed to applying for the resource consents?
(4) If the Commissioner was correct as to the dates from which TrustPower committed to applying for the resource consents, did she wrongly allocate expenditure as between feasibility expenditure and expenditure incurred in obtaining the resource consents?
Resource consents were not stand-alone assets
On the first issue, Justice Andrews found that the resource consents acquired by TrustPower were not stand-alone assets. This conclusion was sufficient for the High Court to find in favour of TrustPower. The Court accepted that if the resource consents were not stand-alone assets then expenditure incurred in obtaining them is deductible feasibility expenditure.
Justice Andrews rejected the Commissioner’s attempt to use the statutory scheme of the Act as support for the proposition that the resource consents are capital assets. Her Honour cautioned against drawing inferences from the wording of one section of the Act as to the interpretation and application of another section. In any event, Justice Andrews found that the depreciation provisions were not applicable because TrustPower did not use the resource consents, or have them available for use, during the relevant years. Furthermore, Her Honour found that the land use consents were significant in each of the four projects and were not “depreciable intangible property” listed in Schedule 17 of the Income Tax Act 2004.
The key reason that Justice Andrews concluded the resource consents were not stand-alone assets appears to be the finding of fact that they were of little or uncertain value. Justice Andrews accepted TrustPower’s evidence that the development pipeline is only one part of its business development, that having resource consents is only one of many factors that must be considered in deciding whether to proceed to construction, and that there is very little competitive advantage in holding resource consents. Although the consents were of value to TrustPower, Her Honour accepted that the value only exists as part of a package of rights and may be tenuous at best. Justice Andrews concluded that the resource consents were part and parcel of the projects and it would be artificial from a practical and business point of view to regard them as separate assets in their own right. Accordingly, the expenditure incurred by TrustPower in obtaining the resource consents must be treated in the same manner as the projects to which they relate.
Justice Andrews noted the Adjudication Unit’s decision that expenditure relating to the projects was feasibility expenditure and not capital in nature, and held that the same conclusion should apply in relation to expenditure incurred in obtaining the resource consents. It follows that TrustPower’s expenditure was feasibility expenditure deductible pursuant to the general permission. In so concluding, Justice Andrews accepted that the obtaining of resource consents was a crucial and integral part of TrustPower’s business and part of its feasibility analysis process.
If stand-alone assets, resource consents were revenue assets
Although the conclusion that the resource consents were not stand-alone assets was sufficient for the High Court to find in favour of TrustPower, the Court went on to consider whether the consents, if they were stand-alone assets, were capital or revenue assets.
Justice Andrews applied the indicia identified by the Privy Council in BP Australia Ltd v Commissioner of Taxation for the Commonwealth of Australia  AC 224. Her Honour concluded that from a business and practical point of view the resource consents, if they were stand-alone assets, were revenue assets.
Influential to this conclusion was the factual finding that TrustPower’s purpose in incurring the expenditure was to assess the feasibility of projects in the development pipeline (in particular, to define the parameters of possible projects and to enable an assessment of possible projects against TrustPower’s other options for sourcing electricity), rather than to obtain the resource consents. Justice Andrews accepted that the resource consent applications were part of, and one of the current necessities of, TrustPower’s regular feasibility process. The Court considered that the expenditure was indiscriminate as part of TrustPower’s general business operations expenses.
Despite accepting TrustPower’s submission that there is inherent uncertainty surrounding the terms and duration of the resource consents and whether projects will be progressed further along the development pipeline, Justice Andrews considered that the resource consents did provide TrustPower with an enduring benefit. However, Her Honour found that, in the absence of a commitment to proceed with the underlying projects, the resource consents were not the means by which TrustPower generate income. Therefore, even if the resource consents were assets, they could not be regarded as part of TrustPower’s business structure.
Date from which expenditure would be classified as capital if resource consents were capital assets
At the High Court, it was common ground between the parties that, if the resource consents were capital assets, the classification of expenditure as capital or revenue depended on whether it was incurred as part of the feasibility process or as part of the consenting process. It was also accepted that the distinction between the two is based on time at which TrustPower committed to applying for the resource consents.
Justice Andrews accepted TrustPower’s evidence that, although it instructed consultants to prepare reports in a form suitable for inclusion in an application, it was not committed to lodging an application until such time as it had received and considered all relevant information. Her Honour found that the expenditure incurred by TrustPower in the period up to around the time applications were lodged was revenue expenditure and deductible under the general permission. Justice Andrews therefore did not consider the last issue as to the classification of costs incurred after the commitment date.
The Court of Appeal unanimously overturned the High Court’s decision. In reaching its decision, the Court approached the case quite differently from Justice Andrews.
The Court dismissed upfront the Commissioner’s argument that resource consents were depreciable intangible property and therefore capital assets. The Court found there was nothing in the scheme of the Act to suggest that the depreciation regime was intended to override the general provisions relating to the income/capital distinction. To the contrary, the Court held that expenditure that is deductible under the general permission is excluded from the depreciation regime.
The Court therefore considered that it must first determine the character of the expenditure under the general principles relating to the capital/revenue distinction.
Capital/revenue distinction – Expenditure is capital under general principles
The Court of Appeal concluded “with little difficulty” that the expenditure at issue was incurred on capital account under general principles.
After canvassing the established case law on the capital/revenue distinction, the Court observed that the general principles from Hallstroms Pty Ltd v Federal Commissioner of Taxation (1946) 72 CLR 634 (HCA) and Commissioner of Taxes v Nchonga Consolidated Copper Mines  AC 948 (PC) remain the best guide for distinguishing between income and capital and may be sufficient without resort to the BP Australia factors (at ). Drawing on those cases, the Court framed the inquiry as one of whether the expenditure was intended to achieve, from a practical or business point of view, the creation, acquisition, or enlargement of TrustPower’s business structure.
In concluding that the expenditure was capital (and therefore non-deductible), the Court of Appeal’s primary objection to TrustPower’s claim for deductions appears to be that the object of the expenditure was capital. The Court considered that TrustPower’s development pipeline was the means by which it determined the viability, feasibility, and costs of building new generation capacity, and any new generation capacity involved an extension or expansion of its business structure. The Court stated at  that:
All of the “feasibility expenditure” related to possible future capital projects. It was not incurred by Trustpower in deriving income from its existing business. As Lockhart J recognised in Ampol, preliminary expenses incurred in the establishment, development or extension of a capital item such as a mine will ordinarily be in the nature of capital expenditure.
Notwithstanding the High Court’s conclusion that the resource consent had no intrinsic value, the Court considered that the resource consents gave TrustPower valuable rights and options in relation to its long term programme of future capital works. These enabled TrustPower to postpone decisions for the terms of the consents until it was in its interests to construct new generation plant. The Court also found that the consents conferred a first mover advantage and rejected TrustPower’s argument that the resource consents do not “block” competitors.
The Court accepted Justice Andrew’s finding that the resource consents were not stand-alone assets but considered that to be irrelevant in light of the Court’s finding that the expenditure was for the purpose of extending or expanding TrustPower’s existing business and was therefore on capital account.
The Court also referred to and placed some significance on English and New Zealand cases which held that expenditure incurred in obtaining resource consents and permissions were on capital account. The Court considered that it was irrelevant whether the taxpayer has committed to the project (at ):
By obtaining resource consents, TrustPower invested unequivocally in capacity, whether or not it was committed at that time to proceed with the build. The investment was inherently capital in nature.
The Court considered that “[d]etermined objectively, there was a sufficient connection between the expenditure and capital” (at ). The Court then went on to find that the requisite nexus between the incurring of the expenditure and the deriving of the income was not established (at ):
[T]he disputed expenditure was not incurred “in carrying on” TrustPower’s business or in earning the income of the existing business or in performing the income-earning operations of the existing business. TrustPower’s profit-making enterprise is the generation and retailing of electricity, not the development of its pipeline of possible new projects or the investigations of, and applications for, resource consents for those projects. Possible future projects in its development pipeline are for the purpose of extending, expanding or altering its business structure in the future not part of the carrying on of TrustPower’s ordinary business activities or the taking of steps within that framework, being the generation and retailing of electricity. In terms of s DA 1 the requisite nexus between the incurring of the expenditure and the deriving of the income is not established.
This finding marks a fundamental shift from the approach in the High Court. It accords to TrustPower’s expenditure on resource consents no connection at all with the carrying on of its business. Divorcing the pipeline of possible new projects from the business of earning income from generating and retailing electricity meant that the pipeline was being treated as having only one outcome – new projects and increased capacity. TrustPower’s lynch pin concept was called “optionality”. Its expenditure was seen in the High Court as being incurred to give it business choices related to the sourcing of electricity for sale – the so-called “build or buy” choice. The Court of Appeal declined to follow that approach, saying effectively the consents could only ever have a business outcome in building new capacity and so their costs were capital.
Capital/revenue distinction – Application of the BP Australia factors
Although the Court concluded that the expenditure was capital after applying the general principles, the Court went on to consider the application of the BP Australia factors even though it was not strictly necessary. Unsurprisingly, the Court was satisfied that those factors support the conclusion that it had already reached.
The Court disagreed with Justice Andrews that characterising the expenditure as feasibility expenditure meant that it was of the same character as TrustPower’s other operating costs. Notwithstanding Justice Andrews’ factual findings that the purpose for the expenditure was not solely or principally to obtain resource consents, the Court considered that the expenditure related to possible future capital projects and was not incurred in earning income from TrustPower’s existing business.
The Court agreed with Justice Andrews’ analysis that although the expenditure may be regarded as recurrent rather than once and for all, it provided TrustPower with an enduring benefit once the consents were obtained.
Unlike Justice Andrews, the Court considered the expenditure to have been incurred on TrustPower’s business structure. It stated its reasoning at  as follows:
The fact that the resource consents cannot themselves generate any electricity or create any income for Trustpower does not mean that they should be found to be “revenue assets” or that expenditure on them should be on revenue account. As we have held, the resource consents were acquired by Trustpower in order to move the four specific projects along its development pipeline of capital projects. Obtaining the resource consents was a critical step or integral component of the development of the four specific capital projects. The expenditure on the resource consents for that purpose was therefore clearly on capital account. The fact that no electricity was generated or income created from the expenditure simply confirms that it was not on revenue account.
The last sentence of this excerpt from the judgment is telling. The Court saw no connection between the resource consent expenditure and the fact that the company might well decide to acquire electricity, sell it and derive income from sources other than its own increased capacity. The fact that the expenditure might well lead to that outcome was apparently insufficient connection with the income-earning process for a deduction to be sustained. It suggests a very narrow concept of a business being carried on for the purpose of deriving income.
In concluding that the BP Australia factors supported its conclusion, the Court noted that even if the resource consents are viewed as stand-alone assets, the BP Australia factors do not mean that the expenditure was incurred on revenue account.
Time of commitment
It was not necessary for the Court to determine when TrustPower committed to applying for the consents, given its decision that all of the disputed expenditure was on capital account irrespective of whether Justice Andrews was correct in holding that the resource consents were not stand-alone assets.
Although it was common ground between the parties that if the consents were capital assets the deductibility of expenditure depended on whether it was incurred before TrustPower committed to applying for the consents, the Court considered the commitment test to be a pragmatic rather than legal approach to the issue, stating at  that:
In our view the Judge’s findings about commitment date are academic. If the Commissioner chooses to adopt a date rather than sufficient connection as the test that is a matter for the Commissioner.
The Court went on to say at  that:
the fact that the Commissioner has selected a date of commitment by Trustpower does not mean that the Commissioner was in error in doing so. It was open to the Commissioner to adopt a pragmatic approach to this issue.
The Court noted that commitment is not necessarily when the relevant Board or management decision is made and that the question of commitment may be considered in hind sight (at ):
In our view the fact that Trustpower’s Board may not have made a final decision to apply for the particular resource consent did not mean that when the disputed expenditure was incurred it was not sufficiently connected to a capital purpose. For tax purposes that question may be appropriately determined in hindsight because taxpayers invariably file their returns after the event. In this context, “commitment” in relation to any given payment simply means that the payment is sufficiently connected to the capital purpose of obtaining a resource consent.
The Court of Appeal remitted back to the High Court for determination the issue of how to allocate particular expenditure as capital rather that revenue after the dates Trustpower committed to applying for the resource consents.
Approach to factual issues
The High Court and the Court of Appeal had fundamentally different interpretations of the purpose/objective of the expenditure incurred by TrustPower, which led to the difference in outcomes.
At the heart of the problem is the Court of Appeal’s linkage of the resource consents to the capital projects to which they relate. The Court of Appeal considered that the expenditure could only ever be regarded as connected with a project that is destined to increase/improve TrustPower’s generation capacity. Even though the project might be a distant “castle on a cloud” it was treated as inevitable, at least for the purpose of characterising the expenditure on the consents. Unlike Justice Andrews in the High Court, the Court of Appeal did not see the expenditure as being part of a regular process which may or may not lead to an increase/improvement in generation capacity. Optionality was not a concept that resonated with the Court, even though the evidence at first instance had all supported it as being the company’s object in undertaking the expenditure.
The Court of Appeal’s characterisation of the expenditure as being inevitably related to a capital project, rather than a process, appears to us to be contrary to the High Court’s findings of fact that:
The Court of Appeal approached the case on the premise that expenditure on resource consents could only have an outcome in terms of the project to which they relate, which carries with it an implicit assumption that TrustPower would go ahead with the projects. The High Court, with the benefit of having heard and examined the evidence of both sides, did not identify anything to support that premise/assumption.
The Court of Appeal also departed from the High Court in finding that the resource consents had standalone value.
The High Court considered that the resource consents could only be valuable as part of a complete package of rights. Justice Andrews accepted TrustPower’s evidence that the resource consents did not “block” competitors and, in the absence of a commitment by TrustPower to proceed to construction, there was very little competitive advantage in holding resource consents. The High Court also accepted TrustPower’s submission that the presence of a project in the development pipeline did not mean that the project would ultimately result in one of the benefits put forward by the Commissioner’s expert witness (at ).
Notwithstanding the High Court’s factual findings, the Court of Appeal preferred the evidence led by the Commissioner in holding that the resource consents had standalone value.
There is an overarching issue here as to whether and to what extent the Court of Appeal is able to depart from the factual findings of the High Court. It is not clear how the factual findings of Justice Andrews at first instance could be overturned or simply ignored by the Court of Appeal when the Court of Appeal did not have the benefit of hearing or examining the evidence first hand. In the absence of substantive countervailing evidence, it is surprising that the Court of Appeal was able to make findings that are contrary to TrustPower’s evidence.
We can only speculate as to what might have caused the Court of Appeal to depart so obviously from the High Court’s view. The Commissioner seems to have persuaded the Court of Appeal that inferences could be drawn from documentary evidence that should outweigh the company’s evidence, and that substantial concessions had been made by company witnesses under cross-examination. It is questionable whether that ought to have displaced the findings at first instance.
Was there something inherently different about TrustPower’s circumstances that made it difficult for the Court of Appeal to treat the expenditure as feasibility expenditure? One possibility is that, notwithstanding the fact that there were approximately 200 other projects in the pipeline, the Court of Appeal saw the application for resource consents as evidencing a greater level of commitment to the four particular projects to which the consents relate. This might have led the Court to form the view that the expenditure was sufficiently connected with a capital outcome so as to be fixed with a capital nature.
Approach to the deductibility question
The Court of Appeal's finding that TrustPower’s expenditure did not meet the nexus to income test in s DA 1 was not argued in the High Court and did not appear to form part of the Commissioner’s case on appeal. Indeed the Crown specifically agreed that the expenditure on resource consents met the requirements of the general permission and that the capital limitation was all that was in issue. Accordingly, there is a real question as to whether the argument that section DA1 was not satisfied was available for the Court of Appeal to consider.
Putting that to one side, in our view the approach the Court of Appeal has taken to the analysis of the general permission and the general limitations is wrong. The Court of Appeal has effectively applied the nexus to income test in the reverse in holding that the general permission is not satisfied because “there was a sufficient connection between the expenditure and capital” (at ). That is not the test for deductibility. The general limitations do not exclude an item of expenditure from falling within the general permission. Rather, the general limitations override the general permission. This is a fundamental and accepted principle. The difference is between carving out from business-related expenditure an item for which a deduction is subsequently denied when it would otherwise be allowed, and simply never making it to the allowance of a deduction in the first place.
There is no question at all in our minds that the expenditure incurred by TrustPower satisfied the general permission. It was incurred as part of TrustPower’s ordinary business operations. The business of an electricity retailer requires that it investigates how it would source the electricity that it sells. TrustPower’s situation is distinguishable from cases that involve feasibility expenditure incurred by taxpayers to determine whether or not to enter into a new business or income-earning activity prior to commencement of that business or activity (see Case S39 (1995) 17 NZTC 7,264, referred to in the Interpretation Statement at paragraphs 45 to 49).
The idea that expenditure which might relate to some future expansion of business if a decision is taken to proceed is capital, even when it is incurred well short of any decision being made, is in our view incorrect. It ignores the possibility that the process in which the expenditure is incurred may just as much lead to a decision not to expand by building capacity at all. In the meantime, the expenditure is still related to the business on a DA 1 basis and it is not yet connected with a capital expansion so as to be ruled out under the capital limitation. As the Commissioner stated in the Interpretation Statement at paragraph 187, “commitment requires a decision to proceed, in contrast to a taxpayer continuing to weigh up whether or not to proceed.”
The Court of Appeal’s approach simply assumes that TrustPower’s expenditure had no purpose other than an expansion of its business or generation capacity: "A consent to build a wind farm means that you are going to build a wind farm". TrustPower’s argument is more subtle: "A consent to build a wind farm helps us decide if we will build it or buy electricity instead." One approach assigns the expenditure inevitably to a capital outlay and the other foresees the analysis of business choices requiring that capital options be weighed and assessed against non-capital options. By ruling the expenditure out of deduction under the general permission in s DA 1, the Court of Appeal has denied the possibility that the weighing of business choices, when one might involve capital expansion, is undertaken in the course of conducting a business for the derivation of income. The Court’s construct of the objective for which the pipeline existed ignored that comparative assessment of options and attributed the expenditure only to the possibility of expanding generation capacity, whether or not that was actually ever undertaken. The “castle on the cloud” was treated as if actually built.
So what led the Court of Appeal to so drastically depart from an approach that was as orthodox and as accepted as the approach to the general permission and capital limitation? Perhaps it was the Court’s acceptance (at ) that the characterisation of the expenditure as revenue or capital needed to be considered first “because if the expenditure is on revenue account it will be deductible under s DA 1” that led it astray? Whatever the reason, this is one aspect of the case with which we wholeheartedly disagree and hope would be reconsidered by the Supreme Court.
Weight given to authorities
It is somewhat surprising to us that the Court of Appeal relied on cases such as FCT v Ampol Exploration Ltd 86 ATC 4,859 and Milburn, which are arguably distinguishable from TrustPower, but its judgement did not analyse cases such as Bowater.
As mentioned above, the Canadian Federal Court (Trial Division) in Bowater allowed a taxpayer in the business of generating and selling electrical power and energy deductions for expenditure incurred on surveys/engineering studies relating to the development of additional power and the location of physical plant for a power station. The Court held that expenditure incurred in merely considering whether or not to create a capital asset may be revenue in nature, stating at paragraph 73 that:
While the hydroelectric development, once it becomes a business or commercial realty [sic] is a capital asset of the business giving rise to it, whatever reasonable means were taken to find out whether it should be created or not may still result from the current operations of the business as part of the every day concern of its officers in conducting the operations of the company in a business-like way.
Bowater is factually similar to TrustPower and was considered by the Commissioner to be consistent with NZ and Australian authorities, but was not mentioned in the Court of Appeal’s judgment. Instead, the Court of Appeal relied on Ampol and Milburn.
The Court of Appeal cited Ampol for the proposition that preliminary expenditure incurred in the development or extension of a capital item will ordinarily be capital in nature. The taxpayer in Ampol had no interest from which an income-producing asset could arise, but the expenditure incurred was perceived as commercially sound method of carrying on exploration business and as part of ordinary business activities because Ampol expected to derive a fee as a result.
In Milburn, resource consent application expenditure was held to be non-deductible primarily because Justice Wild found that the taxpayer had committed itself to the development of the relevant quarries before it applied for the resource consents (at 17,023):
These six factors, certainly in combination, indicate to me that the taxpayers, having investigated or evaluated the three sites, had made business decisions to expend money in developing the sites for commercial production. The first step, or one of the first steps, to that end was to apply for the necessary consents.
The expenditure in Milburn was therefore held to be capital because the taxpayer had made a commitment to develop the quarries, not because the resource consents were inherently capital themselves. That was accepted in the Interpretation Statement, which considered Milburn to be consistent with the principle that feasibility expenditure that forms part of normal business operations is deductible even if it would give rise to a capital asset.
Appeal to the Supreme Court
TrustPower has sought leave to appeal to the Supreme Court. Assuming that leave is granted, in our view there is a risk that, even if the Supreme Court accepts that TrustPower outlaid the expenditure for an object which included the “build or buy” comparative costing process (which might suggest revenue account treatment), the Court may well conclude that the consents were valuable assets in and of themselves, capable of being turned to capital account at TrustPower’s choice, such that the costs of acquiring them should nevertheless be non-deductible. Against that, of course, are the first instance Judge’s findings of fact that the consents could not be regarded as standalone assets of any value.
Although that would not be a good outcome for TrustPower, we hope that in considering the issue the Supreme Court would at least reinstate the orthodox approach to deductibility and the “commitment” test, which would enable other taxpayers with different factual circumstances to deduct feasibility expenditure that does not give rise to an identifiable standalone capital asset, or that is incurred prior to commitment to proceed with the acquisition or development of a capital asset.
The TrustPower case is directly relevant to any business that regularly incurs expenditure on resource consents. It will also have more general application to any taxpayer that incurs expenditure in assessing the feasibility of possible future capital projects or investments. On the Court of Appeal’s analysis, any expenditure that can be seen to relate to a capital asset would be non-deductible, even if the expenditure is incurred prior to a decision being made to acquire/develop the asset and even if that decision might never be made.
The Court of Appeal’s approach does not sit well with commercial realities. For most businesses, growth and development is incremental and will not usually take the form of a discrete new plant or capital project. Most business development activities have as their ultimate objective the improvement of profitability and by and large most businesses grow by augmenting their existing productive capacity over time. If the Court of Appeal’s analysis is taken to its logical conclusion, expenditure to assess how to achieve business growth would arguably have a sufficient link to capital to render them non-deductible in their entirety if one of the options in play is the expansion of capacity. This would mean, for example, that salary costs of employees involved in potential capital projects would have to be allocated to those projects and treated as non-deductible (see Christchurch Press Co Limited v Commissioner of Inland Revenue (1993) 15 NZTC 10,206). That does not seem to us to be the correct result.
If it stands, the Court of Appeal’s approach is likely to result in a significant increase in non-deductible non-depreciable “black hole” expenditure. This appears inconsistent with current policy settings and recent legislative initiatives to minimise the incidence of such expenditure, for example in relation to R&D costs. Black hole expenditure will arise if a taxpayer does not proceed with a potential capital project or the project is unsuccessful and does not result in a depreciable asset. In TrustPower, the Court of Appeal suggested that the company has a choice to capitalise and depreciate the expenditure (on the basis that the consents were “available for use” and hence depreciable once they were granted, even if they were not being used and would not be used until TrustPower decided to use them). However, that ignores the fact that for taxpayers whose deductions have been called into question by the Court of Appeal’s decision, it is not up to them to choose to capitalise and depreciate the deducted expenditure. They must rely on the Commissioner to exercise her discretion to reopen a tax return under s 113 of the Tax Administration Act 1994 and there is no guarantee that she would do so (see for example Westpac Securities NZ Limited v Commissioner of Inland Revenue (2014) 26 NZTC 21,118).
It is difficult to reconcile the Court of Appeal’s analysis of feasibility expenditure with the accepted practice set out in the Interpretation Statement. The Court of Appeal’s approach appears to leave no room for the commitment test. However, the Court suggests that the Commissioner may nonetheless continue to apply the commitment test as a “pragmatic approach”, but “that is a matter for the Commissioner”. The Court’s comments on this aspect of the case leave taxpayers in an undesirable position of considerable uncertainty as to whether they should apply the new Court of Appeal approach or continue with the prior approach in the hope that the Commissioner would respect it.
Inland Revenue should issue some guidance on its approach in light of the Court of Appeal decision. We understand that Inland Revenue has no current plans to withdraw the Interpretation Statement. If that is correct, then the status of the Interpretation Statement needs to be made clear to taxpayers. In our view, Inland Revenue should continue to apply the approach set out in the Interpretation Statement until the Supreme Court’s judgment is available (if leave to appeal is granted) or at the very least to tax returns filed prior to the Court of Appeal judgment (if leave to appeal is not granted).
Our title refers to the popular song from Les Miserables, “Castle on a Cloud”. It is used to capture the essence of the Court of Appeal’s approach as we see it: Something in the far distance, a possibility for future construction, future expansion of a business, is treated as if it is a present reality. The Court’s approach is effectively to say that if expenditure relates to a project, it is inevitably expenditure on the project. In taking that approach it rejects the argument that other options may emerge which supercede or are preferred over building the “castle”.
The approach reflects a restrictive view of business operations and unless readdressed by our final Court or by Parliament, raises the risk of a revenue brake on business development and improvement which the country can scarcely afford. If this decision remains the last word, Inland Revenue may well be led to express its own version of the Elliot Ness “feasibility” quote with which this paper opens.