More ruminations on valuation issues.pdf (88 kb)
4 December 2017
The market value of an asset is a pervasive feature of tax law, and when it is in dispute it is almost always problematic. The value of a particular asset is a question of fact but the proper way to find that value is a matter governed by law. In several recent cases the Courts have concluded the valuation process was wrongly executed, undermining the revenue authority’s tax assessment, and in some cases prompting changes to law. This Tax Brief examines what the recent decision in Placer Dome  WASCA 165 adds to the law governing the process of valuation for duty and income tax.
The law behind valuation does not come before Australian Courts often but three recent cases (Resource Capital Fund III, AP Energy and Placer Dome) have all involved examining some of the general principles of valuing assets for tax purposes. The legal problems can crop up in different ways:
- sometimes the issue is to identify precisely the asset to be valued and its articulation with related assets: a line of stamp duty cases examines what is the relevant asset to be valued when a business, including land and goodwill, is being sold. Is the goodwill an independent asset (the sale of which will usually not attract duty these days) or does the goodwill add to the value of the land (the sale of which will definitely attract duty)? A similar issue is the articulation between selling a mining tenement and the information derived from exploration about where the minerals have been found (or not found);
- sometimes the issue is tracing: the revenue liability attaching to a share transaction may require tracing through multiple tiers of subsidiary entities to find and then value underlying assets, a process that presents its own difficulties; and
- sometimes the question is about methodology: whether the best evidence of the current value is from historical records (examining prices paid and received in similar transactions), from an analysis of current costs (what it would cost to re-construct the asset if it were to be built today), or from predicting the future (estimating the current price by discounting to a present value the future expected cash flows over some period at some interest rate).
Placer Dome raised all these questions and more.
The case arose out of the acquisition of Placer Dome by Barrick Gold (Barrick) in 2006 for about A$15bn. Placer Dome had gold mining operations in WA at the time. The takeover exposed Placer Dome to WA duty (under land-rich provisions) if four conditions were met, the only contentious condition being whether, at the time of the takeover, land (anywhere in the world) owned by Placer Dome represented 60% or more of the value of its property (some property being excluded for this purpose). If so, Placer Dome became liable to duty in WA on the value of land and chattels located in WA.
It was agreed that the total value of all relevant types of property owned by Placer Dome was A$12.8bn so the land component had to be worth at least A$7.6bn for duty to be triggered. Both of the taxpayer’s experts valued the land at around A$5.5bn; both of the Commissioner’s experts valued the land above A$7.6bn – one at around A$8.5bn and the other at about A$12bn!
The WA Revenue Office concluded that land represented more than 60% of the value of Placer Dome’s assets and issued an assessment for A$54m in duty based on the value of land and chattels in WA (valued at about A$1bn). The WA Administrative Appeal Tribunal upheld the assessment and the taxpayer appealed to the Court of Appeal of the WA Supreme Court. The taxpayer won the battle – the valuation was unsatisfactory – but the war is not yet over – the Court remitted the matter to the Tribunal to reconsider the taxpayer’s objection, imposing a series of detailed constraints on how this was to be done.
The judgment contains several important observations about the process of valuation; these observations dictated the way the Court insisted the rehearing be conducted.
Valuing the land directly v. valuing the non-land (and thus the land indirectly)
The rules in the duty legislation operated in a similar way to the principal asset test in the CGT provisions – to find out whether a share sale (or a seller) is liable to tax, the value of land inside the company must be ascertained. Tax or duty will be triggered if the value of the target’s land assets represent a sufficient proportion of the company’s total assets. (The issue in both sets of provisions is the same, although the Full Federal Court has suggested that the context somehow makes the CGT process different.)
The price paid for the shares represents a price for both the land and non-land assets of the company and so, conceptually, there are two ways to decide how much of the total was paid for each class of asset:
- value the land directly, so that the value of the other assets represents the remainder of the share price; or
- value the non-land assets and treat the remainder of the share price as representing the value of the land.
While either method should produce the same figure, the Court said only the first method should be used in this case; that it was not appropriate to subtract the value of all non-land assets from the total consideration and so treat the remainder of the amount as paid for the land. Rather the land should be valued directly because of the problems associated with the alternative:
- doing the reverse requires that ‘each and every item’ has been fully and carefully identified and that all those assets can be accurately valued; and
- more importantly, it will have the effect of attaching to ‘land’ any value which does not come from identified (or identifiable) assets.
The Court suggests this is a general rule and must be applied in cases where correctly identifying and valuing every asset ‘is not practicable or feasible.’ Certainly the Court gave the impression that it was sceptical about the ability of the revenue authority to correctly identify and value every asset of a company that operated over 4 continents, employed 13,000 people and made over A$300m profit each year:
with a business as substantial and complex as that conducted by [Placer Dome], it would be difficult, if not impossible, to exhaustively identify each and every component or attribute of [its] business which contributed to the value of [its] business as a whole.
This issue arose in three different ways in the case. First, there was a dispute about the relationship between the value of the mine (land) and the value of the taxpayer’s mining business (something which uses land and other assets). The Court accepted the distinction between the value of a mining business operating as a going concern and the value of discrete assets such as a mine, valued on asset-by-asset basis. Because the value of the business as a going concern had not been separately itemised and valued, this methodology effectively meant the difference (the amount by which the value of the going concern exceeded the value of the mine) leaked into the value of the land.
Secondly, there was a dispute about so-called ‘synergy benefits’ – the premium a buyer would be willing to pay because of the synergies or savings it would get from consolidating two businesses into one. There was evidence that Barrick estimated it could save A$200-$250m per year and this represented A$1.6-$2bn of the A$15bn share price. The revenue authority did not accept this – it argued synergy benefits were not property at all, and if they were property, they were not property of Placer Dome at the time of the sale (they only arose in the hands of a buyer once it had bought the business). The Court rejected the revenue authority’s view which meant the revenue authority’s indirect approach must produce an incorrect value: because the value of synergy benefits had been ignored, and thus not valued, their value would inappropriately enhance the value of land.
And the same issue turned up in another dispute in the appeal involving goodwill. The methodology adopted by the revenue authority would have the effect of drawing value away from land (or adding value into land) depending on whether goodwill was identified (or not) and how it was valued. Most of the judgment is concerned with describing the defects in the valuation of the goodwill of a gold mining business, but as the Court said, the goodwill issue only arose because the methodology depended crucially on being able to itemise and then value every non-land asset (in order to reach the value of the land).
Valuing an operating mine v. valuing the land on which the mine operates
A fundamental point of distinction between the two sides related to the difference between valuing an operating mine and valuing the land on which the mine operates. The Court agreed with the taxpayer’s argument that there was such a distinction but this then raised the question, how is a mine to be valued separately from the land on which it operates? The buyer of the company has paid for both without distinguishing between them.
The taxpayer’s expert differentiated the mine from the land by adopting a reconstruction notion – if the operating mine on the land was worth A$X, then the land would be worth A$X less what it would cost to construct an operating mine on it:
[the] benefits a hypothetical purchaser of an operating business obtains rather than acquiring land alone … are quantified by reference to certain costs, delays and risks that a hypothetical purchaser avoids by acquiring the operating business instead of the business's identifiable assets.
The Tribunal had rejected this approach but largely because it did not agree that there were two assets to be valued separately. The Court also rejected this approach but for a different reason: it challenged the reconstruction methodology.
The Court said the legal principles underlying valuation require the expert to value what was actually sold; the only hypothetical element is that the sale happens between two hypothetical parties; it is not proper to assume (and then value) a hypothetical asset as well:
no purchaser in a hypothetical transaction for the purchase of [Placer Dome’s] business as a going concern would contemplate offering, nor would a hypothetical vendor contemplate accepting, a price assessed by reference to something else – that is, a transaction which involved only the acquisition of [Placer Dome’s] land, after which the purchaser would be required to bring the mines on the land into operation.
This passage suggests reconstruction methodology is not going to be accepted as a valid approach in future valuation cases.
Predicting the future
Another area of dispute in the case arose from the use of discounted cash flow (DCF) methodology by the parties to value the mine.
A gold mine might possibly be valued by looking at historical data (sale prices from earlier sales of comparable property in similar circumstances) but comparable sales are likely to be rare and so both sides valued the mine using the expected cash flows from future extraction and sales of gold. DCF requires estimates of the size of the proven gold reserves, future gold prices, interest rates, extraction costs and so on. The taxpayer’s experts predicted a low price in the short term and the price would fall over time; the revenue authority’s expert predicted a higher gold price and assumed it would steadily grow over time, which accounted for how mines might be worth A$5.5bn or might be worth A$12bn!
The key difference between the two sets of experts was how to predict the future price of gold. The expert for the revenue authority had based his prediction on the prices from gold futures contracts and assumed a constant 2% p.a. price growth for years in which futures contracts were not written. The experts for the taxpayer argued that the best evidence of future prices was informed estimates by market analysts and industry experts.
The Court preferred the taxpayer’s view, noting that gold futures were not a predictor of future gold prices; they were essentially a financial product and were viewed that way by the industry. And the assumption that gold prices would continuously rise at an even 2% per annum was rebutted by evidence that, ‘an ever-increasing price for gold in real terms had never been observed in the previous 172 years …’
The Court remitted the matter to the Tribunal to determine the value but without allowing new evidence and excluding all the evidence led by the revenue authority. It is curious that, having found such fundamental defects in the revenue authority’s approach, and effectively ruling out introducing any new evidence, the Court was unwilling simply to accept the taxpayer’s valuations and give judgment accordingly. It seems the Court wanted to leave open the possibility that the Tribunal might not find the taxpayer’s evidence convincing and so allow the matter to be decided on the basis that the taxpayer had not discharged the onus of proof. That may not be so far-fetched – the Court expressed some doubt about the conclusion the taxpayer’s valuers had reached that the value of the land of a gold mining company might represent as little as 1/3rd of the value of its total assets; they were clearly inviting the Tribunal to scrutinise that figure very closely.
A key lesson from this saga is the reminder of just how fraught valuation disputes can be: they almost never end well for either side, with Courts often rejecting (or at least expressing scepticism about) the methods and conclusions of expert valuers. This case takes us a few steps closer to a more complete set of rules prescribing just how valuations are to be done for income tax and duty purposes.