There was some information about the remissions, but these are not in issue in these proceedings.

1. HTI Watson appeals to this tribunal in relation to this decision.[18] The tribunal's role when reviewing a decision of an entity is provided for in section 68(2) of the ACT Civil and Administrative Tribunal Act 2008 (ACAT Act) which says: "The tribunal may exercise any function given by an Act to the entity for making the decision." Section 68(3) then provides that when reviewing an administrative decision, the tribunal must:
   • (a) confirm the decision;
   • (b) vary the decision; or
   • (c) set aside the decision; and

(ii) remit the matter for reconsideration by the decision-maker.

1. The tribunal can therefore make its own decision based on the evidence before it. On an appeal to the Supreme Court, Refshauge ACJ in Giusida Pty Ltd v Commissioner for ACT Revenue[19] adopted comments by Zelling J who said in Doherty v Commissioner of Highways:[20]

Judges do not have to accept the valuations of the valuers of either side and frequently arrive at a figure or figures, which constitute a modification or modifications of the figures submitted by one or more valuers. ... They are guided in coming to the conclusion by the evidence of the valuers together with the other evidence in the case.

1. The applicant filed an application for review dated 8 May 2019. The applicant provided a further report for these proceedings by Mr Richard Swinbourne attached to a statement dated 2 August 2019 (exhibit A1). Mr Swinbourne made valuations on a number of bases. For the before value, a comparable sales method was used, and this provided a before value (V1) of $7,100,000.[21]
2. Mr Swinbourne used a range of bases for assessing the after value (V2). One was what was called the hybrid method which gave an after value of $18,200,000.[22] Mr Swinbourne also had regard to three comparable sales to obtain an after value of $11,400,000.[23] He further had regard to a full hypothetical development method which gave an after value of $17,800,000, adjusted in the hearing to $12,700,000.[24] In conclusion Mr Swinbourne suggested an after value of $18,000,000.[25]
3. The applicant also provided a report from James Osenton dated 2 August 2018 (exhibit A10).[26] This formed an integral part of Mr Swinbourne's hybrid method. Both Mr Swinbourne and Mr Osenton gave oral evidence and were cross-examined.
4. The applicant provided a wide range of additional documentation. It also provided submissions on the correct and preferable decision (dated 8 August 2019), as well as applicant's closing submissions (dated 11 November 2019), and applicant's submissions in reply (dated 26 November 2019).
5. The respondent provided the T documents (exhibit T1). A witness statement was provided by Geoff McInerney dated 21 August 2019 (exhibit R2). Mr McInerney maintained the position of the respondent set out at above, namely:
   • (a) an after value (V1) of $27,500,000; and
   • (b) a before value (V2) of $3,530,000. [27]
6. Mr McInerney gave oral evidence and was cross-examined. The respondent provided a number of documents, submissions on the correct and preferable decision dated 30 August 2019, as well as closing submissions.

1. At the beginning of this hearing and throughout most of it the respondent maintained its position that the before value, V2, was $3,530,000. However in closing submissions, the respondent agreed with the applicant's before value of $7,100,000.[28]

1. Mr Swinbourne's primary comparable sale for the before value was a former research establishment at block 4 section 126 Symonston. This had sold in May 2018 for $3,810,000. Mr Swinbourne adjusted this to $3,924,640 to give a rate per square metre of $127. His second comparable sale was the Australian Defence College at block 1212 Weston Creek (or block 2 section 121 Weston). This had sold in October 2014 for $30,500,000. Mr Swinbourne adjusted this sale, to take out the value of the improvements, to $12,362,360 to get a rate per square metre of $137. A third before sale for the applicant was block 22 section 112 Symonston. This had sold in April 2015 for $6,500,000. Mr Swinbourne adjusted this sale to $5,338,800 to get a rate per square metre of $167. Other sales were also considered. Mr Swinbourne stated that he thought the direct comparison method, where the subject property is compared with other sales and adjustments made, was the most appropriate method here. Having regard to this process, he was of the view that the site value of $160 per square metre of land was appropriate, which gave a before value for the subject site of $7,100,000.[29] The Tribunal is of the view that this is a reasonable approach.

1. Mr McInerney, in his original assessment, looked primarily at the sale of the subject site. He took that sale price of $8,600,000, deducted the value of buildings and car parking, driveways, service lines and landscaping, which he assessed at $4,860,000, to give a before value of $3,740,000.[30] He used $80/sqm to get a rounded value of $3,530,000.[31] A similar approach was taken on the review, though with some regard to comparable sales.[32] Further, a similar approach was taken in Mr McInerney's statement for these proceedings (exhibit R2) [33] and his oral evidence, though with some regard to comparable sales. The Tribunal thinks there is some doubt about this approach, as there is significant evidence that the property was purchased for development, and was not therefore a useful basis for a before value.

1. The Tribunal therefore thinks it is appropriate to accept the agreement of the parties that the applicant's before value should be accepted.

1. The remaining dispute is therefore about the after value (V1).
2. Mr Swinbourne for the applicant and Mr McInerney for the respondent provided detailed reports, gave oral evidence and were cross-examined in relation to their assessment of the after value (V1). They each provided a list of sales which they considered to be evidence of the value of the subject land in the after sale (V1) scenario. They generally analysed the sales to derive unit rates, either on a per dwelling basis or a per square metre basis, to apply to the subject site. As to dwellings, it was accepted by both sides that the whole development would consist of 320 residential dwellings, made up of 170 apartments, 125 townhouses and 25 terraces and commercial and childcare uses.[34] Mr Swinbourne also used two other approaches, a full hypothetical development method and his preferred hybrid method for the after value scenario.
3. The relevant date of valuation is 31 October 2017. The land has an area of about 44,211- 44,217 square metres; the difference is of no consequence.
4. Mr Swinbourne tendered a table (exhibit A7) which indicated an increase in the moving annual median price for townhouses over the period March 2014 to October 2017 of 5.19%. Both valuers agreed that the market had been relatively stable over the period 2014 to 2018 and so no adjustment was necessary to most of the sales to account for time.

1. Mr Swinbourne's evidence was that the most appropriate method of valuation was by direct comparison sales, but that there were no sufficiently comparable sales to allow the direct comparison method to be meaningfully utilised, and therefore it was necessary to use another approach.[35]
2. Mr Swinbourne did identify 45 sales in his table 4 of his report, but he said that these generally represent the sale of development ready sites, and the subject site was not in this condition.[36] We make an important point here, which we return to below; in relation to the 45 sales, Mr Swinbourne provides very little information. Mr Swinbourne's comments about them are very generalised, broad brush and conclusionary. Mr Swinbourne does identify three sales of larger land areas in his table 5. He does seem to admit that these are more useful as comparable sales.[37]
3. It seems that in Mr Swinbourne's view the task of adjusting these development ready sites to be comparable with the subject site was a difficult one. Because of this Mr Swinbourne preferred what he described as a hybrid method, a blend of the full hypothetical development method and the comparable sales method. Mr Swinbourne stated:[38]

The above market evidence [the 45 sales in Table 4, then the 3 large site sales in Table 5] provides indicators of the value for a wide range of development opportunities. While sales of larger land areas are available, such sales are in fact less comparable due to being located along Northbourne Avenue, within the City area or within prime areas of Inner South Canberra.

... the property can be compared with a number of sales that have occurred across Canberra. The sales above of vacant land sites offered for sale by the ACT Government as part of the Land Release Program are of use. But the principle (sic) difference between the subject property and most sale properties is that this Watson property is much larger in land area.

1. His report then noted that there are 12 development sites within the subject site within which 320 dwellings of various types are proposed. It then stated:[39]

The sales evidence in Table 4 (Page 31) [the 45 sales] points to appropriate units of value to apply to housing types that can be achieved on development of the subject property. Knowledge of these sale properties assists in categorising the deduced values into the ... housing types.

1. Mr Swinbourne therefore proceeded to assess the after value by ascribing values to some development components across the whole subject site. This included consideration of the childcare, commercial and residential components. The residential split was 170 units/apartments, 125 townhouses and 25 terraces. In evidence, Mr Swinbourne stated that he had compared sites within the subject site development and the sales and considered differences such as sizes of blocks.[40]
2. Values were then assigned to the elements of the subject site development: childcare; commercial; apartments; townhouses and terraces. It was then said:[41]

While each of the 12 sites is developable in its own right (similar to the sale properties), there are costs relating to the servicing of the broader estate that must be brought to account for the sales evidence to be correctly applied.

1. The report of the quantity surveyors in relation to the civil works was referred to (exhibit A10), and then it was said:[42]

I have adopted the costs to get the individual sites to a serviced stage to be developable and comparable to the above sale properties.

1. That is, Mr Swinbourne assessed the after value by ascribing values to the development components across the whole site. This includes consideration of the 12 development lots created over the whole original block, including the childcare, commercial and residential components.
2. His calculations are as follows:[43]

(e) plus commercial site 752 sqm at $500 sqm = $376,000

__________________________________________________________

(f) to give a total gross realisation of: $32,726,000

1. This became the gross realisation for the development that fed into the hybrid development calculations. Mr Swinbourne then deducted an amount from this figure for profit and risk allowance, this figure was originally $3,292,609. It was later adjusted to $4,268,609.[44]
2. He then deducted the civil works costs to get the subject site to a development ready state. Initially Mr Osenton's evidence in these proceedings was that the applicant would incur costs of $10,236,934 in undertaking the civil works required to make the site ready for development. This was adjusted during the course of the proceedings to exclude the cost of demolishing the existing structural improvements.[45] The final figure was $9,378,559.[46]
3. In summary therefore Mr Swinbourne's calculation was as follows:[47]

(a) gross realisation (paragraph [48](f) above) $32,726,000

(b) less profit and risk (paragraph [49] above) $4,268,609

(c) less civil works (paragraph [50] above) $9,378,559

___________________________________________________________

Mr Swinbourne further adjusted this final figure in his oral evidence at the hearing to $19,350,000.[48]

1. Mr Swinbourne was asked in cross-examination if this was a valuation method which he used often and he said:[49]

No, it's not but for the benefit of the tribunal, I was trying to present a simpler - a simplified - more step by step, only three steps in it.... [It is simpler than] the method 3 [the full hypothetical development method], which is the customary method.

... it's a static - static approach, it doesn't involve a cash flow and - and if someone's not comfortable with cash flows then this would be a fair indicator and buy the - buy the land and service it and resell it and you take off - it's a calculation of more the hard costs when you're doing work for financiers or others, have regard to the hard costs and these are the hard costs of the development, except for the profit margin in the middle.

1. Then asked whether the comparable sales method would be more appropriate, he said:

If we had really good evidence but we don't have a really good evidence in this case.

... I don't regard the evidence generally as being comparable to the property, without significant adjustment. And this hybrid approach makes that adjustment.

1. The applicant's closing submissions described this approach as based on the hypothesis of a developer purchasing the subject site, planning an estate, servicing the land to a development ready stage, and then on-selling the vacant land to a third party to undertake the construction and sale of the planned dwellings. It was said that the approach was to use sales of development ready sites to ascribe values to each of the prescribed uses, then to calculate a gross realisation figure, and then deduct the amount which it would cost the developer to undertake the civil works required to make the subject site development ready.[50]

1. The appropriateness of this approach was a key remaining dispute in this case, so it is necessary to address a number of issues in relation to it. First, it is not a pure comparable sales approach. It is clear that such a comparable sales approach is the preferred method of valuation. (see paragraphs [81]-[83] below). Mr Swinbourne admitted this (see paragraph [40] above). Mr Swinbourne made some efforts to implement such an approach (see paragraphs [84]-[86] below). The respondent attempted to take this approach (see paragraphs [99]-[110] below). Of course if this approach is not possible, then other approaches must be investigated. But in our view some assessment needs to be made as to whether the comparable sales method is possible. The Tribunal makes further comments in relation to how this could be done below at paragraphs [118]-[155].

1. Second, the respondent argued that the hybrid approach was prohibited by the rule in Toohey's, Ltd v Valuer-General[51] (Toohey's case). It is necessary to spend some time discussing what that rule is, and whether it applies here.
2. The principle in Toohey's case appears to be that it is not appropriate when working out the unimproved value of a subject property to ascertain what the improved value of the subject property is, and then deduct the cost of improving it, to get the unimproved value. In Toohey's case the valuer took a figure for the subject land which included the bare land, the buildings on it and a licence to sell liquor. The valuer subtracted the value of the buildings but not that of the licence.
3. The Privy Council said that improvements are to be left entirely out of view when determining the unimproved value as required in that case; they are to be taken, not only as non-existent, but as if they never existed. It was then said:

It is, therefore, to approach the question from a completely wrong point of view to begin with a valuation which takes in the improvements and then proceed by means of subtraction of a sum arrived at by an independent valuation in order to find the required figure. What the Act requires is really quite simple. Here is a plot of land; assume that there is nothing on it in the way of improvement; what would it fetch in the market?[52]

It was said that the valuation should be of the land itself "without any buildings upon it, and without any consideration of the value of the subject as including de facto licensed premises."[53]

1. The respondent argued that this principle was breached by the hybrid approach taken by Mr Swinbourne. Further, it said that its position was also supported by Valuer-General v Fenton Nominees Pty Ltd (Fenton Nominees).[54] In that case, the High Court said that it was well settled that in establishing what the unimproved value was it is:

...necessary to inquire what the hypothetical purchaser would pay for the land in a notional condition shorn of its improvements and that it is not permissible to arrive at the figure by identifying the value of the site in its improved state and then subtracting the value of the improvements... [55]

But Toohey's case did not resolve the issue in Fenton Nominees. In that case there were two competing approaches to the valuation of comparable sales, not the subject land which was being valued. But the decision made clear that the principle in Toohey's case does not speak to comparable sales, because it focuses on the subject land; it does not deny that sales of improved property in the vicinity may be relevant for the purposes of valuing the subject land.[56]

1. The respondent also said that its position was supported by Maurici v Chief Commissioner of State Revenue (Maurici).[57] In that case it was necessary to arrive at the unimproved value of a developed site. The Commissioner fixed this value on the basis of only sales of scarce unimproved parcels of land in the same locality. The High Court said that this was wrong. Rather the valuer, as the appellant had argued:

... should have had regard to all the relevant facts including the scarcity of vacant land, the possibility of a particular and limited class of persons in the market for it, the scarcity or otherwise of improved land, the added value of improvements to comparable lands, and in particular, truly comparable sales, which ideally would include like land similarly improved to the subject land.[58]

1. This decision confirms the primacy of comparable sales. It confirms the need to look at a range of relevant facts, and use comparable sales of developed land to obtain a figure for the undeveloped subject site. The case supports using the improved value of comparable sites to determine the unimproved value of a subject site. It does not support calculating the unimproved value of a subject site by working out the improved value of that site, and deducting the value of the improvements.
2. This issue was further considered in detail by the NSW Court of Appeal in Commonwealth Custodial Services Pty Ltd v Valuer-General (Commonwealth Custodial Services).[59] This case concerned a particular statutory provision which was an amended version of the one considered in Toohey's case, in that it required the valuation on the basis that improvements, other than land improvements, had not been made, but specifically provided that it shall be assumed that land may be used for any purpose for which it is being used, and improvements may be continued or made on the land to enable that use.
3. Chief Justice Spigelman emphasised the importance of the terms of the relevant statutory scheme (see [4]-[6]), and noted he found Toohey's case unconvincing (at [7]). He stated at [11] that in Maurici, the High Court seemed to affirm that "long practice accepts that it is perfectly rational to value unimproved land by valuing a site in its improved state and deducting the value of the improvements". However, he stated, on the authority of Toohey's case, that:

... [T]his rational method for valuing comparable sales is impermissible for valuing the land in issue. ... there is a distinction between the two situations because in one case there have been actual sales.[60]

He thought the issue was determined by the emphatic manner in which the High Court expressed its approval of Toohey's case in Fenton Nominees.[61]

1. Santow JA seemed to agree that the subject land could not be valued on an unimproved basis by deducting the value of improvements from the improved value. However, he does seem to accept that it is possible to assume that the land is unimproved, and then theoretically build on it an imaginary building in order to arrive at a valuation.[62]
2. Tobias JA provided the most detailed judgment. The case involved the valuation on an unimproved basis of the heritage-listed Commonwealth Bank Building in Martin Place, Sydney. In simple terms, one valuer, Mr Jackson, determined the improved value of the property with its existing heritage listed improvements, and then deducted the added value of those improvements. The other valuer, Mr Hill, derived a land value of the property unaffected by heritage restrictions based on comparable sales of similar unaffected properties. He adjusted that figure to take account of the heritage restrictions and design and layout inefficiencies. Tobias JA accepted Mr Hill's approach but rejected Mr Jackson's. His Honour said at 213, [115]:

...in my view there is nothing anomalous in permitting the use of comparable sales of improved land to be used for the purpose of deducing therefrom an unimproved value of the same land which is then applied, with or without adjustments, to the land to be valued but outlawing the use of that method with respect to the land to be valued by valuing it in its improved state and then deducting, in order to obtain its land value, the added value of its existing improvements where the statute as a matter of construction mandates that result. [Emphasis added]

The reasons given for this are that the legislation required that improvements are to be out of view, and that there has been no actual sale of the subject site, as distinct from the comparable sites where there has.[63]

1. As the respondent pointed out, these principles have also been set out in Nap Nap Station Pty Ltd v Valuer-General (Nap Nap Station).[64] Further, the comparable sales approach makes the relevant adjustments to the comparable sales, not to the calculated improved value of the subject site.[65]
2. The applicant argued that the hybrid model was a rational approach, which should be allowed. In this regard it is true that Callinan J observed in Boland v Yates Property Corporation Pty Ltd[66] that there is no legal principle that purports to, or could close for all times, the categories of methods of valuation which might be acceptable to a particular case. But the cases noted above grapple with this very issue, and all find that the limitation in Toohey's case should be applied. It is true that some of the judgments discussed above raise issues with Toohey's case, but all in the end follow it.[67]
3. The decision in Estate of James v Valuer-General (James)[68] did take a contrary view. There Roper J stated that in certain circumstances there is no error involved in a valuer forming his opinion as to the unimproved value of the land by deducting the value of the improvements from the improved value. This is quoted in Nap Nap Station but not followed. There Justice Stein said that he accepted that it was permissible to deduct the estimated value of improvements from the actual sale of a comparable improved property to establish a deduced land value, but he did not consider, since he did not need to, whether this could be done in relation to the subject property. This judgment is another, which confirms that the comparable sale method is the preferable course_._[69]
4. But in Commonwealth Custodial Services Santow JA and Tobias JA specifically and expressly reject James on the basis of the High Court authorities_._ Spigelman CJ stated that he finds the reasoning in James to be entirely convincing as to the validity of the mechanism of valuation in issue, albeit Justice Roper's attempt to distinguish Toohey's case was difficult to sustain. Spigelman CJ indicated that he was bound by the High Court authorities to follow Toohey's case.[70] In our view this Tribunal is in the same position.
5. The respondent also referred to comments by Gobbo J in Waalt Homes Pty Ltd v Road Construction Authority.[71] As the respondent conceded in its submissions in reply at footnote 2 (to [16]), these seem to be comments about adjustments to a comparable sale. The respondent said that these comments should apply in the same way to the subject site. But the principle in Toohey's case as applied by the High Court and NSW Court of Appeal and noted above prevents this very approach to the subject site.

Does the rule in Toohey's case apply to the hybrid method?

1. The applicant argued that this case ultimately turns on its own facts.[72] That is true, but there is clearly a relevant principle here; it is not applied by simply saying the facts are different; it is necessary to ask does the principle apply to these facts.
2. Applying this principle in this case is a matter of some difficulty. As discussed, it seems that Mr Swinbourne did, albeit in a very broad-brush manner, look at comparable sales of what he regarded as partly improved, that is development ready, land to assess the value of the subject site. This seems to be consistent with the rule in Toohey's case, as discussed in the cases noted above. But Mr Swinbourne then deducted from this assessed value of the partly improved, that is development ready, subject site the cost of civil works to get it to this condition. This seems at least to some extent to be inconsistent with the rule in Toohey's case. That is it is taking a partly developed value for the subject site and then deducting from this the cost of that development on the subject site. Tobias JA suggested in Commonwealth Custodial Services, as set out in paragraph [65] above, that this is an inappropriate method, that is valuing the subject site in at least a partly improved state, and then deducting, in order to obtain its land value, the cost of the improvements. This seems particularly problematic because these deductions are not calculated on the basis of any actual land sales.
3. It is true as the applicant argued that it is not trying to include irrelevant improvements into the assessment.[73] It seems to be true that it is only once services are run to the edge of the relevant block that the Territory will issue Crown leases. It is also true that the applicant is by this method attempting to ensure a like for like comparison between the comparable sales and the subject site.[74] The problem is that the applicant used comparable sales to develop a part improved value for the subject site, and then used the cost of those improvements on the subject site to reach the unimproved value. It seems on the basis of Toohey's case, and the applicant's position in relation to the need for an assessment without the civil works, that what should have occurred was that the comparable sales themselves should have been adjusted. The respondent said that this approach was not reasonable or necessary, but we have gone to some lengths in this decision to show that this is probably not correct. We do not need to base our decision wholly on this conclusion however, since there are other issues in relation to the hybrid approach.

Other issues with the application of the hybrid method

1. There are other concerns we have about the hybrid approach as used by Mr Swinbourne. One is, as we have noted, that he used a very broad-brush approach to the comparable sales. He sets out 45 sales, but provides few details or analysis in relation to them. This provides a general and somewhat shaky basis for his hybrid approach. In particular Mr Swinbourne conceded that some of the properties were not development ready.[75] Further, as the respondent noted, while some of the sales were of small properties, some were of large properties, comparable to the subject site. Further, there was not much clarity in relation to what being development ready meant for the 45 comparable sale sites. There was certainly no assessment provided to the Tribunal that they had all the civil works, which were taken into account for the subject site.[76]
2. When he came to actually apply a comparable sales approach, Mr Swinbourne focused on the details of some specific sales, which in our view was the more appropriate and reliable approach.[77]
3. The second concern follows from this, and is that we are not convinced that the approach of looking at comparable sales and appropriately adjusting these was not available. The comparable sales approach undertaken by Mr Swinbourne and noted below shows this was possible. This approach has the benefit that it relies on arm's length comparable sales, that is what a purchaser actually paid for the comparable properties.
4. Third, as the respondent argued, if there is to be an estimate of costs, it should be at the date of the valuation, 31 October 2017. This seems correct, and was apparently accepted by the applicant.[78] Further, Mr Osenton's first costs summary was prepared in November 2016,[79] at which time he said that the relevant costs were about $1,551,000, similar to that assessed in the MMJ report.[80] But Mr Osenton then updated these costs in July 2019 to $10,237,000, which were then revised down to $9,379,000 to take out demolition costs.[81] This was not only done from an inappropriate date, but this was also a major increase from 2017 to 2019 of over 600%, for which little explanation was given. It was said to be based on recent actually incurred costs, which it was conceded might have involved overestimation,[82] and there was little other explanation for such a significant increase.
5. The fourth issue is that the applicants purchased one lease, block 2 section 64, Watson (original parcel), in April 2016. This block was subsequently divided into 16 blocks resulting in seven leases (subdivided parcels).[83] The assessment by Mr Swinbourne appears to be based on the proposition that the original parcel is developable in its own right (similar to what he says is the position with the comparable sale properties), but that there are costs relating to the servicing of the subdivided parcels that must be brought to account for the comparable sales evidence to be correctly applied to them. Certainly all the works, which are costed and then deducted, apparently took place within the original parcel, not outside it. However, all the relevant valuation calculations used by Mr Swinbourne, and everyone else, are of the original parcel. They are not of the subdivided parcels, which are the units which were said not to be development ready. If it is the cost of getting the subdivided parcels development ready that is relevant, then this should be applied to the value of each of the parcels, not the value of the whole parcel. Mr Swinbourne did discuss that the site area for the development is 26,014 square metres, 58.9% of the total original site area.[84] However, this does not flow into the relevant calculations.
6. Lastly, the applicant put forward no evidence of the general acceptance of this hybrid method. Mr Swinbourne himself admitted that this was not a valuation method which he used often.[85] Indeed there was no evidence he had ever used it before in a formal valuation process, or that anyone else had ever used it. This is not to say that this may not be a useful and rational approach. The question is should it be the principal approach adopted by the Tribunal in this case.
7. Therefore, we have concerns about the use of the hybrid method as the principal valuation method, based on the principle in Toohey's case and its adoption by Australian courts, and concerns about the application of the hybrid approach in this case. At very least, we do not think it is an appropriate principal valuation method here. In our view this should be the generally accepted comparable sales method.

1. The High Court said in Maurici:[86]

16 ... The traditional, and usually unexceptionable [valuation] method is to seek out relatively contemporaneous sales of comparable properties between parties at arm's length, unaffected by special circumstances, such as, for example, a strong desire by a purchaser to buy an adjoining property, and to use those sales as a yardstick for the valuation of the relevant land.

1. Justice Sheahan stated in Marroun v Roads and Maritime Services[87] that "it is well established that, if comparable sales are available, their direct comparison should provide the conventional method of valuation".
2. It is accepted that in applying this method some sales are highly comparable, others less so. There is a spectrum of possible comparisons.[88] Adjustments may need to be made to the sale prices to account for relevant differences. This approach was used to some extent by both parties, and accepted by them as appropriate. The dispute was as to the outcome of its use.
3. As we have discussed, Mr Swinbourne submitted in his report at table 4 a list of 45 vacant land sales.[89] In evidence and under cross-examination he suggested that the sale of the bigger blocks were most comparable, and he discussed these in his cross-examination.[90] We agree that the sales of larger blocks are the most comparable to the subject site, so we have focussed on these from Mr Swinbourne's table 4 list. We give these sales a number for the purposes of this decision, summarise the details here (and further at table 1 paragraph [88] below) and discuss some of them further below:
   • (a) A.1: 1/38 Wright sold on 6/12/17 for $10,250,000. 15,770 square metres site. 158 dwelling potential.

Reflects $64,873/dwelling; 100 square metres land area per dwelling; $650 per square metre.

(b) A.2: 4/28 Greenway sold on 28/02/18 for $18,000,000. 29,260 square metres site. 240 dwelling potential.

Reflects $75,000/dwelling; 122 square metres land area per dwelling; $615 per square metre.

(c) A.3: 4/64 Wright sold on 18/04/18 for $7,100,000. 10,750 square metres site. 107 dwelling potential.

Reflects $66,355/dwelling; 100 square metres land area per dwelling; $660 per square metre.

(d) A.4: 1/11 Coombs sold on 26/09/18 for $7,060,000. 9,831 square metres site. 98 dwelling potential.

Reflects $72,041/dwelling; 100 square metres land area per dwelling; $718 per square metre.

(e) A.5: 1/70 Lyons sold on 29/03/19 for $16,400,000. 23,289 square metres site. 492 dwelling potential.

Reflects $33,333/unit; 47 square metres land area per dwelling; $704 per square metre.

1. Mr Swinbourne also relied on more detailed information and discussion of the comparable sales at table 5 of his report.[91] These were as follows:
   • (a) A.6: 9/64 Watson sold on 28/3/14 for $7,300,000. 29,914 square metres site. The site had potential for development with 110 dwellings (21 separate title dwellings along Negus Crescent and 89 townhouses on three rear blocks).

Reflects a value of $66,364/dwelling; 272 square metres /dwelling; $244 per square metre.

This land required off-site civil works and an extension of Negus Crescent at an estimated cost of $3,200,000. When this was added to the purchase price a dwelling rate of $95,455 results; and a value of $350 per square metre is suggested.[92]

(b) A.7: 36/64 Watson sold in March 2017 for $13,750,000. 23,346 square metres site. This block is a consolidation of the three rear blocks and had the potential for development with 89 townhouses. This site was fully serviced and had water sewer and stormwater mains laid. There had also been extensive land filling and levelling of the site as well as the construction of retaining walls. These had been completed at the vendor's cost under the conditions of a holding lease. Apart from noting that the sale required adjusting down to take account of the works undertaken by the vendor, no further analysis was completed by Mr Swinbourne.[93] Such further analysis would have been very helpful.

Reflects a value of $154,494/dwelling; 262 square metres/dwelling; $589 per square metre.

(c) A.8: 1/82 Weston sold on 23/11/2017 for $32,210,000. 58,596 square metres site, proposed a development of 245 units. However Mr Swinbourne considered this sale to be 'out of line' in the face of other sales and so unreliable.[94]

Reflects a value of $131,469/dwelling; 239 square metres/dwelling; $550 per square metre.

1. There are no further calculations shown as to how Mr Swinbourne gets to his assessed value on a direct comparison basis of $11,400,000.[95] It is very difficult to see how he arrived at this figure. This represents $258 per square metre of site area and $35,625/dwelling. Nothing in the information provided by, or the discussion of, Mr Swinbourne, directly supports this figure. Figures which seem to emerge from the details of the sales provided by Mr Swinbourne are $350 per square metre, which gives a total figure of $15,475,000. Using $450 per square metre gives a total figure of $19,897,000.
2. In its closing submissions the applicant seemed to move away from this comparative sales valuation, and focussed on block 36 section 64 Watson (sale A7 above). As discussed, this sold for $13,750,000 in March 2017. The applicant said that the parties agreed that the sale of this site was the most important in terms of assessing the after value, which suggests that for them it was the most important comparable site.[96] Much evidence was taken in relation to this site, in particular to show that the vendor was liable to carry out civil works on the site. In its closing submissions the applicant stated that while it agreed that the site is important for the after value (V1), that an adjustment to take into account the fact that it was sold with servicing in place was necessary. It was said that: "When so analysed, the sale ... supports the V1 assessment of the Subject Site by Mr Swinbourne, particularly the results obtained from his hybrid method".[97] As noted Mr Swinbourne's hybrid model gave a value of $19,350,000. It is not at all clear how this sale (A7) supports this value for the subject site. We analyse this sale further below at paragraphs [143]-[145].
3. The details of these sales are summarised in the table below:

Original approval for 240 units changed to 150 townhouses.

plus offsite works said to be $3,200,000; total $10,500,000

Mr Swinbourne thought this needed to be adjusted down.

Applicant's method 3 - full hypothetical development method

1. Mr Swinbourne also undertook a full hypothetical development study for the site. This involved applying likely end sale prices to the completed units/townhouses to derive a gross realisation upon completion of $156,211,800. After allowances for developer's profit and risk (20%) and all other costs including holding charges he derived a residual land value of $17,800,000.[98] Following clarification of costs by Mr Osenton this value was reduced by a significant amount to $12,700,000.[99] The applicant said that this supported the after sale value (V1) of $18,000,000,[100] but there is a significant difference between the two figures.
2. Mr Swinbourne placed reliance on this approach as a check method only. Even so, we have significant concerns about this method. First, it has been questioned by a range of tribunals and courts. McLure J stated in Western Australian Planning Commission v Arcus Shopfitters Pty Ltd[101] that the full hypothetical development method is "a less reliable method because of the number and nature of the assumptions that have to be made".
3. This statement is demonstrated by the use of the method by Mr Swinbourne in this case. The starting point is a projected gross realisation upon completion from all sales of $156,211,800. In Mr Swinbourne's report, there is no information about how this sum is calculated. It is said in the report that it is "deduced from analysis of the proposed development and the sales potential of the dwellings"; no analysis was provided in the report, and no specific sales were provided.[102] Mr Swinbourne provided a spreadsheet of his calculations, which set out potential sale prices, but no basis for these prices; again no specific sales were provided. It was said that this was a culmination of a number of documents. It was said that these came from an analysis of a number of sales. But no information in relation to these sales was provided.[103] Mr Swinbourne conceded that this method was "alive with variables" where the smallest change on one could make a substantial change in the end result.[104]
4. Given this, the Tribunal does not think that it can accept the full hypothetical development valuation provided by Mr Swinbourne, at least as the principal valuation method.

1. There was some discussion in the hearing of an alternative method of valuation, namely taking the sale price for the block prior to the lease variation and adding to that amount the lease variation charge. Both Mr Swinbourne and Mr McInerney thought that this would be some indication of the after value (V1).[105]
2. But there is significant uncertainty in relation to what the lease variation charge is. This case demonstrates this starkly, with wide variations between the experts. Further, it does not seem appropriate, or possible, to use the figure of the lease variation charge to calculate the lease variation charge. The Tribunal therefore has no regard to this approach.

Summary of applicant's position on after value (V1)

1. In summary, Mr Swinbourne arrived at three figures, with a significant range:
   • (a) Hybrid method, originally $18,200,000, adjusted in the hearing to $19,350,000 (see paragraphs [49]-[51] above).
   • (b) Comparable sales method, $11,400,000.
   • (c) Full hypothetical development method, originally $17,800,000, adjusted in the hearing to $12,700,000 (see paragraph [89] above).
2. In his report Mr Swinbourne gave his after value (V1) as $18,000,000.[106] This was the amount put by the applicant for the after value (V1) in the hearing.[107]

1. The Tribunal also had before it a report from Knight Frank dated 7 November 2016 submitted by the applicant for the original assessment by the respondent. For the after sale value (V1) this report looked at a number of comparable sales, and assessed these on a dollar per square metre basis. It particularly looked at the adjacent block to the subject site, block 9 section 64 Watson (A6 table 1, paragraph [88] above). After allowing for $3,500,000 for offsite works, it analysed the sale rate as $361 per square metre. It adopted a rate of $350 per square metre for the subject site, to arrive at an after value of $15,475,000.[108]

1. For the review of the assessment by the respondent, the applicant provided a report from MMJ dated 18 October 2018, which the Tribunal also had before it. This report used land sales to ascribe land values for units (170 at $55,000 each), townhouses (111 at $100,000 each, and 11 at $120,000 each), terraces (25 at $120,000 each), the childcare centre ($600,000) and the commercial area ($225,600) to produce a value of $25,500,000. This was amended in light of proposed changes to the DA to $24,400,000. This was then adjusted by deducting an amount for required works of $1,253,897, to give a rounded after value (V1) of $23,150,000.[109] This was greater than the value put in these proceedings, in part because of, as noted above at paragraph [77] above, the very significant increase in the amount deducted for required civil works.

1. Mr McInerney's evidence was contained in his witness statement (exhibit R2). His first advice dated 19 December 2017 was in the Tribunal documents (exhibit T1) at pages T573-T585. The advice by Mr Pilat dated 4 April 2019 on the review is at pages T188-T196. As noted these provide an after sale value (V1) of exactly the same amount, $27,500,000.
2. The method of presentation of this advice, and in particular the statement for the Tribunal, made it difficult for the Tribunal to follow the rationale and application of the comparable sales to the subject property, and we principally had to rely more on Mr McInerney's oral evidence for this.
3. In his evidence Mr McInerney stated that he relied on the following properties.[110] As we did for the applicant's main comparable sales, we give these sales a number for the purposes of this decision. For each we calculate a value per potential dwelling, an area per potential dwelling, and a value per square metre, to obtain figures which can be used to calculate the value of the subject site:
   • (a) R1 (also A2 above): 4/28 Greenway sold on 28/02/18 for $18,000,000. 29,260 square metres site. 150 dwelling potential (under the revised proposal for 150 townhouses, see paragraph [146] below).

Reflects $120,000/dwelling; 195 square metres land area per dwelling; $615 per square metre.

Mr McInerney considered this to be an inferior location.

(b) R2: 5/13 Greenway sold in 12/16 for $15,100,000. 26,000 square metres site. 151 dwelling potential. This site required demolition of existing structures and payment of a lease variation charge, as well as offsite amplification charges. Mr McInerney's calculations suggest an adjusted price of $19,537,500.

Reflects $129,387/dwelling (after costs); 172 square metres land area per dwelling; $581 per square metre (before costs) or $751 (after costs). (Mr McInerney calculated this at $594 per square metre but this seems to be on the basis that the demolition costs are included but not the lease variation costs.[111])

(c) R3: 2/75 Bruce sold on 7/07/16 for $15,400,000. 19,590 square metres site.[112] 89 dwelling potential. Costs of $501,300 to remove easements.

Reflects $173,034/dwelling ($178,666 after costs); 220 square metres land area per dwelling; $786 per square metre ($812 after costs).

(d) R4: 1/82 Weston sold on 23/11/17 for $32,210,000. 58,590 square metres site. 245 dwellings potential.

Reflects $131,469/dwelling; 239 square metres land area per dwelling; $550 per square metre.

(e) R5 (also A6 above) 9/64 Watson sold in March 2014 for $7,300,000. 29,914 square metres site. 110 dwellings potential. Off site works said to be $3,500,000[113] (Mr Swinbourne thought this was $3,200,000) to be added, to give a total of $10,800,000.

Reflects $98,182/dwelling after off site costs are added; 272 square metres land per dwelling; $244 per square metre without the offsite works, $361 with offsite works.

(f) R6 (also A7 above) 36/64 Watson sold on 3 March 2017 for $13,750,000. 89 dwelling potential. 23,346 square metres site.

Reflects $154,494/dwelling; 262 square metres land area per dwelling; $589 per square metre.

(g) R7: 18/115 Watson sold on 8 November 2018 for $6,800,000. 8,975 square metres site. 33 townhouses dwelling potential.

Reflects $206,060/dwelling; 272 square metres land area per dwelling; $758 per square metre.

Original approval for 240 units changed to 150 townhouses.

plus demolition plus lease variation costs gives $19,537,500

Mr McInerney noted costs of $501,300 to remove easements

plus offsite works of $3,500,000 which gives $10,800,000

1. The applicant criticised reliance on block 5 section 13 Greenway (R2), in particular because of the fact that Mr McInerney did not produce any evidence to substantiate that works were required to be completed by the purchaser to enable development, and were said to be inside the block.[114] The respondent pointed out that the works are not referred to in the discussion of this property in Mr McInerney's report, and at any rate where the works are is not so relevant to a proper comparable sales analysis.[115]
2. The applicant also criticised the reliance on block 1 section 82 Weston (R4) on the basis that Mr McInerney claimed that the purchaser was required to carry out extensive bulk earthworks but did not produce evidence in relation to this.[116] The respondent stated that this ignores the detailed discussion in Mr McInerney's report and attachments G and H.[117] We think it is reasonable to take into account both those sales.
3. Mr McInerny's report also contained a number of other sales in Dickson (3), Downer, O'Connor (2), Throsby, Coombs (3), Moncrieff (2), Lawson (3) and Bruce, which were ostensibly in support of his after value (V1). There was no further reference to these sales in evidence and so their use to the Tribunal is limited. Many were analysed on a gross floor area basis which was carried through to the valuation table without explanation.
4. Mr McInerney then proceeded to assess the after value (V1) for the subject site by ascribing values to the individual development components across the whole site, using an earlier version of the subdivision plan and referencing gross floor areas for each of the components. This includes consideration of the 12 development lots to be created, including the childcare and commercial components. The residential split provided for is 175 units/apartments and 145 townhouses. Values for the townhouses vary depending on the area of the site from $75,000/dwelling for a 101 square metre dwelling, to $220,000 for a 273 square metre dwelling on a 200 plus square metre site. We note that there is very little information or calculations showing how Mr McInerney got from the comparable sales to these values. Like Mr Swinbourne he takes a general, broad-brush approach to this step.
5. The values are assessed individually in the reports, especially the first report dated 19 December 2017, and can be summarised as:
   • (a) Units/Apartments:
     • (i) 60 at $55,000 = $3,300,000
     • (ii) 115 at $60,000 = $6,900,000
   • (b) Townhouses:
     • (i) 11 at $75,000 = $825,000
     • (ii) 51 at $80,000 = $4,080,000
     • (iii) 22 at $85,000 = $1,870,000
     • (iv) 12 at $90,000[118] = $1,200,000
     • (v) 21 at $95,000 = $1,995,000
     • (vi) 11 at $200,000 = $2,200,000
     • (vii) 17 at $220,000 = $3,740,000

________________________________________________________

(d) Plus childcare site, 70 places at $10,000/place = $700,000

(e) Plus commercial, 1309 sqm GFA at $600/square metre = $785,000

________________________________________________________

(f) Total value ((c) plus (d) plus (e)) = $27,595,000.[119]

1. In essence this calculation uses comparable sales, broadly considered, to determine values for a realisation of the whole development. It is an approach similar to the applicant's hybrid approach. Mr McInerney agreed that his approach, to this point, is very similar to that of Mr Swinbourne.[120]
2. Mr McInerney stated that the values adopted by him had each been discounted by 20% to reflect the fact that they were part of a large development. We note that no such calculation is shown in the statement he provided for the Tribunal (exhibit R2). On this basis his position was that there was no need to further discount them for a "sale in one line". He disagreed that any further discount was necessary. He said in relation to the sales that he had chosen:[121]

Well, the majority of them aren't development ready. And that's why I picked out the ones which were bigger, which required something - they're more comparable to the subject, and compared them where they've got to incur like internal roads and extension of infrastructure and those sort of things. But the bulk of the ones on his table 4 [of Mr Swinbourne's report, exhibit A1] are just simply not comparable.

1. Mr Swinbourne had also made a deduction in his hybrid method calculation for profit and risk of about 15%, which Mr McInerney noted was similar to his 20% deduction. But Mr McInerney disagreed with the need for the further reduction for civil works which Mr Swinbourne had made (in an amount of $9,378,559 or about another 30% (see paragraph [50] above).[122]

1. A number of issues arise from the valuations by the parties. First, everyone recognises the appropriateness of the comparable sales approach. But Mr Swinbourne does not principally rely on a 'pure' comparable sales approach. He does include this as one of his three approaches, and discusses three particular sales in this regard, but there is little information about his application of, and calculations for, this approach. His principal approach is the hybrid method, which uses comparable sales to calculate a value for the site based on its proposed development, but then with deductions, including a significant deduction for the actual civil works to be undertaken on the subject site. Mr McInerney also uses a similar comparable sales method, with some deduction but not one for the civil works, to calculate the value of the site based on its proposed development. The Knight Frank report of 7 November 2016 seems to come closest to being a pure comparable sales analysis.[123] No one relied on the sale of the subject property, even though this was probably an after sale, that is a sale for development.
2. Second, in each case the mechanics of how the comparable sales were used by the valuer to get to the adopted values for the subject site was not revealed in much detail.
3. Third, neither valuer in their valuations addressed the subdivided sites. They did not take into account the granting of seven Crown leases comprising 16 lots, i.e. five residential leases being lots 7, 8, 9, 10, 11, a lease for communal uses being Lot 16 DP11671 comprising a total of 16,971 square metres, and a residue lease holding seven residential/commercial lots being lots 1-6, 12 and communal lots 13-15 DP11671.[124] Both valuers in effect valued the original unsubdivided lot. They also did not expressly take into account that this assumed the sale of the seven leases to a single buyer.
4. Fourth, the key difference between the approaches was the allowance by Mr Swinbourne for the costs of servicing the sites to make them 'development ready'. As discussed above at paragraphs [55]-[80], the Tribunal has a range of concerns about the use of the applicant's hybrid approach as the principal valuation method in this case.
5. There is an issue about what the Tribunal should do in this situation. As noted above, the Tribunal "may exercise any function given by an Act to the entity for making the decision."[125] Under section 68(3) of the ACAT Act it may:
   • (a) confirm the decision;
   • (b) vary the decision; or
   • (c) set aside the decision; and

(ii) remit the matter for reconsideration by the decision-maker.

1. Given the range of issues in relation to the valuations, consideration was given to remitting the matter to the original decision maker to resolve these. But this matter has been going on for some time; there were seven days of hearing in the Tribunal; to remit the matter would mean considerable further delay; and there is a wide range of evidence before the Tribunal. It seems appropriate for the Tribunal to use this material to reach a revised valuation using principally the comparable sales method.
2. We therefore consider a comparable sales approach based as Mr Swinbourne and Mr McInerney did on the commercial, childcare and residential elements of the subject site. The material does not allow us to look at each new individual lease (but we have noted this issue in paragraph [113]). The material does enable us to focus on what are the key comparable sales. The material does enable us to do what the valuers did and assess values for elements of the comparable sales which can be applied to the subject site.

1. The valuers each ascribed values to the child care centre and commercial components, separately from the residential components. In summary these were:[126]

1. The childcare site (837 square metres) is to be developed on lot 1 (2847 square metres) in conjunction with 55 units. In relation to the childcare site Mr Swinbourne provided five sales (table 6) that shows values ranging from $412/ square metre (1,943 square metres) in Rivett to $904/ square metre (1,991 square metres) in Red Hill. The other sales show values of $510 (3,503 square metres) in Macgregor, $569/ square metre (4,575 square metres) in Higgins and $596/ square metre (3,600 square metres) in Charnwood. There are no calculations to show how he gets to the final value of $600,000, which represents $717/ square metre of site area. Mr Swinbourne makes the point that the proposed site at 70 places is small by industry standards and this affects its value. [127]
2. On the same component Mr McInerney provided three sales which he analysed to a rate per place as follows:
   • (a) Higgins, 4575 square metres; 186 places; reflects $14,467/place;
   • (b) Charnwood, 3600 square metres; 176 places; reflects $12,520/place;
   • (c) Kambah, 4652 square metres; 120 places; reflects $8,558/place.[128]
3. He also pointed to the potential for the subject site to go up to 90 places, as per the DA lodged for the property. His calculation was however based on 70 places at $10,000/place, which equals $700,000.
4. Overall, we are inclined to adopt Mr McInerney's approach based on a rate per place. We note that the value applied represents $7,778/place when the 90 place maximum contained in the DA is applied. The difference between the two valuers on this issue was however minor.

1. The commercial site (752 square metres) is to be developed on lot 3 (2,574 square metres) in conjunction with 67 residential units.
2. Mr Swinbourne did not provide any sales evidence in relation to this component but applied a value taken apparently from his general knowledge of the Canberra commercial market. He determined the value of $376,000 by applying a unit rate of $500/ square metre of site area to the proposed 752 square metres site. We note that the development feasibility model provided (exhibit A11) confirms the building area for this component at 752 square metres.[129]
3. Mr McInerney provided four sales which he analysed on a gross floor area (GFA) basis to show values of between $577/square metre and $1,119/square metre. These sales also show land values of between $332/square metre and $495/square metre of site area, albeit generally for much larger blocks. The value has been determined by applying a unit rate of $600/square metre of GFA to the potential GFA of 1309 square metres. No supporting evidence was provided as to where this estimate of GFA came from but it is at variance with the provided material which as noted generally provides a higher rate.[130]
4. Mr McInerney's approach is preferred in this instance as it is based on the application of the listed sales. We need however to apply the actual potential GFA to the calculation, i.e. 752 square metres, which at $600/ square metre gives $450,000 (rounded).
5. The total value for these commercial components is therefore assessed to be $1,150,000.

1. The net area available for residential development, after the commercial and child care sites are removed, is 44,211 square metres, minus 1,589 square metres, which equals 42,622 square metres. Whilst the lease variation provides for the issue of seven new Crown leases, both valuers approached the valuation by aggregating the total development potential of the original site, and there is no information available for us to take a different approach.
2. The key sales common to both valuers appear to be:
   • (a) A7/R6: 36/64 Watson, 23,346 square metres, in March 2017, for $13,750,000.
   • (b) A2/R1: 4/28 Greenway, 29,260 square metres, in February 2018, for $18,000,000.
3. In addition, Mr McInerney cited the sale at 2/75 Bruce (R3), and Mr Swinbourne the older sale at 9/64 Watson (A6), which Mr McInerney also discussed (R5). We focus on this limited number of key sales as they are the most comparable to the subject site. They are larger blocks of land with larger developments, though not as large as the subject site, in locations similar to the subject site, indeed in two cases in the same suburb. The sales are summarised in the table below:

1. The sale price for A6/R5 which is 9/64 Watson has been shown as adjusted for off-site works in accordance with Mr Swinbourne's assessment.[131] Mr McInerney agreed that the cost of off-site works should be added to the price; he gave that cost as $3,500,000 while Mr Swinbourne gave it as $3,200,000,[132] We have used Mr Swinbourne's figure. All of the sales are for significantly smaller sites than the subject site. One sale A7/R6 at 36/64 Watson was also partly serviced internally and had significant earthworks completed by the vendor.[133]
2. The analysed values derived from the sales are in the table below:

Table 4: Key comparable sales with value per dwelling and square metre per dwelling

1. All of the sites were to be developed with townhouses. The Greenway site was originally approved for development with 240 units but the developer opted for 150 townhouses instead.
2. Applying this data unadjusted to the subject site gives the following figures:

Table 5: Key comparable sales applied unadjusted to subject site

1. We make some further comments about these key comparable sales.

1. The Bruce sale shows the highest values both as a rate per square metre of site and a rate per dwelling. This is to be expected as it is a higher value environment to the others. Mr McInerney noted that there would be added costs to this sale arising from demolition of structures and extinguishment of an easement.[134] He discussed this further in oral evidence and stated that:[135]

It's, well, it's comparable to the Watson one, the 36 of 64, which is correctly comparable to the subject. It's a site which is less than half the size of the subject. Its development density is 89 town houses, so it's of - it's an inferior development. ... This site, if you apply it to the subject site, if there was two of them, you'd show roughly $30 million on the subject site, which supports my after value. There's just another good supporting sale. It's debateable whether Bruce might be a slightly better location compared to Watson, they're probably similar distances to the city and whatever.

1. Mr Swinbourne said in oral evidence that he did not consider this property because he was not across the full circumstances of the sale. It was put to him that Mr McInerney calculated a value of $772 per square metre and considered the location comparable to Watson. He responded "yes, probably" but noted it was difficult to compare because of the density of the site.[136]
2. The value of $786 per square metre for this site applied to the subject site gives $33,500,000. The dwellings are all townhouses, so the value per unit of $173,034 gives a significant $55,371,000. The land area per dwelling is 220 square metre, 65% larger than the subject site. The full site is half the size of the subject site. This sale is considered a useful, but the site is not a particularly good comparable site to the subject site.

1. There was some discussion of this sale in the hearing. This was in particular in relation to the offsite civil works, given the particular relevance of this issue for the hybrid assessment by Mr Swinbourne.[137]
2. This original sale of 9/64 Watson is some 3.5 years prior to the relevant date and is complicated by the off-site civil works. The area per dwelling at 272 square metres is more than double the average area of the dwellings on the subject site. It is also a much smaller development than the subject site. The analysed dwelling rate of $95,455 reflects the fact that the whole site is to be developed with townhouses. It is considered to be a superior property on a dwelling comparison rate to the subject site.
3. Mr Swinbourne considered that the sale required some upward adjustment for the cost of the civil works to be undertaken. Mr McInerney agreed. This adjustment has been incorporated in the above figures (see table 3 and paragraph [130] above). Being so far removed from the valuation date it should also probably be adjusted for time. Adopting Mr Swinbourne's 5.9% growth in prices produces a value of $101,087/dwelling.
4. There seems to be agreement that this is a good comparable sale. An adjustment for the costs of significant off-site civil works has been included in the calculations, so it is a comparable sale even on the applicant's approach. The application of this sale on the adjusted basis to the number of dwellings for the subject site gives a value of 320 x $101,087 which equals $32,347,000.

1. The sale of 36/64 Watson was initially analysed by Mr McInerney at $154,494/dwelling and $589/square metre of site area. He was unaware that substantial site works including sewer, water and stormwater mains, and clearing and levelling of the site, which included some retaining walls, had been undertaken by the vendor as part of the contract. Mr Swinbourne was able to identify the works from aerial photographs tendered which tracked the development of the site from October 2016 to July 2019.[138] No costings or estimates of added value were provided for the works by either party. Mr Swinbourne observed that the price "needs to be adjusted down due to the benefit that the residual land achieved from the works already undertaken" but did not undertake the exercise.[139] It impossible for us to do so with precision.
2. We note that the site area per dwelling on 36/64 Watson is 262 square metres which is 97% larger than the subject site average of 133 square metre. There is also the matter of scale with this development of 89 townhouses being around 28% of the subject site at 320 dwellings. Even after adjustments for units versus townhouses, the subject site development is nearly three times the size. Adjusting $154,494 per dwelling down by 50% gives $77,250 per dwelling (rounded). Such an adjustment accommodates the applicant's concerns in relation to the civil works undertaken on the comparable site by the vendor, and the need to undertake these on the subject site.
3. The use of this basic value per unit for the subject site gives 320 x $77,250 which equals $24,720,000.

1. The land at 4/28 Greenway was originally sold as a multi-unit site with approval for 240 units. The developer subsequently lodged a development application for 150 townhouses. The sale was analysed by both valuers at $615/square metres of site area. The block was subject to affordable housing requirements. An aerial photo of the site was provided at exhibit A25.
2. Mr Swinbourne analysed this sale on the proposed development of 240 units at $75,000/unit with a land area of 122 square metres/unit.[140] He was cross-examined on this sale and he acknowledged it would need roads and other service lines, but thought that the roads would be fairly narrow, and therefore involve less cost than the subject site. He agreed however that it was a site which was somewhere in-between a development ready site and the subject site and would have had more on-site costs than some others. He implicitly agreed that such larger sites are likely to require more work and therefore be more comparable to the subject site.[141] He thought that some of the units in the development would have the benefit of being near the water and opposite parkland, and all would be close to the town centre.[142]
3. Mr McInerney analysed this sale on the basis of the actual development application at $120,000/townhouse site of 195 square metres. He suggested that there will need to be extensive internal road works and associated infrastructure service lines, implying that this makes it a good comparison for the subject site. He noted also that it was an inferior location and had affordable housing requirements. He commented that this sale demonstrated that the applicant's assessment of the subject site was too conservative.[143]
4. It was put to Mr McInerny in cross-examination that because this site was near the lake, it may have been a superior location. He did not agree, noting that you could only see the lake clearly from one part of the site, and that mostly you could not see the lake because of trees. He thought that there were no great attributes for the majority of the site. The proximity to a public park and Drakeford Drive he thought may in fact be detriments.[144]
5. This again is a much smaller site than the subject site yielding around half the number of dwellings and on areas which are 50% larger. But it was a larger site than many of those in Mr Swinbourne's table 4, and he conceded that some civil works would be required within this site, so it was a comparable sale even on the applicant's approach. The locality was considered to be comparable to Watson by Mr Swinbourne but inferior by Mr McInerney.In the Tribunal's view it is comparable to a significant extent to the subject site. The use of the site's value of $615 per square metre gives a value for the subject site of $26,212,000; the use of the site's value per unit of $120,000 gives a value for the subject site of $38,400,000; though we adjust this further below.

Subject site is a mixture of townhouses and units

1. The subject site is a mixed use site proposed for residential development of 150 townhouses and 170 units. Across the whole development the density is calculated at 133 square metres per dwelling.
2. For the purposes of making a comparison with some of the comparable sites it is necessary to view the subject site in terms of townhouse alternatives. We note two greenfield unit site sales which were modified post purchase to development for townhouses:
   • (a) 3/41 Coombs: 103 units. Development application for 60 townhouses indicates a 42% reduction in dwellings;[145] and
   • (b) A2/R1 at 4/28 Greenway: 240 units. Development application for 150 townhouses indicates a 37% reduction in dwellings.[146]
3. These variations support a value relativity factor for unit values as 60% of townhouse values. If this factor is applied to the 170 units proposed for the subject site you get 170 x 0.6 = 102 townhouse equivalents, or a total townhouse development potential for the site of 252 townhouses. Across the whole site the area per townhouse would be 169 square metres.
4. Mr McInerney adopted average values of $58,285/unit and $109,724/townhouse. Mr McInerney stated he applied a direct comparison with the sales to derive his values as well as applying a 20% discount for the sale 'in one line' of the blocks.[147] As we noted above few calculations were provided of these processes, and nothing in relation to the 20% deduction.
5. Mr Swinbourne adopted average values of $50,000/unit and $155,000/townhouse. Mr Swinbourne's townhouse values are predicated on the individual blocks within the site being fully serviced with roads and utilities to the boundaries. When looked at as a single large site with two road frontages they would necessarily be lower.

1. The Greenway sale (A2/R1) provides the best evidence of value at $120,000/townhouse. It is however a much smaller development (60%) and has a larger land area per townhouse (+26 square metres). We have allowed for a discount to the indicated value of 15% to take account of these differences, leaving an indicated value of $102,000/townhouse site, rounded to $100,000.
2. The sale of 9/64 Watson (A6) provides an adjusted value of $101,087/townhouse. It is also a much smaller development (44%) and has a larger area per townhouse (+103 square metres). We have allowed for a discount to the indicated value of 20% to take account of these differences, leaving an indicated value of $80,870/townhouse site. The multiple adjustments made to this sale in the course of its analysis makes it less reliable than the Greenway sale.
3. Adopting a value of $100,000/townhouse derives a value of $60,000/unit.
4. We therefore adopt the following values:
   • (a) 150 townhouses at $100,000 equals $15,000,000;
   • (b) plus 170 Units at $60,000 equals $10,200,000;

_____________________________________

(c) which gives a total residential value of $25,200,000.

1. We note that as a check this is very close to the value we got for the subject site using the sale of A7/R6 at 36/64 Watson of $24,720,000 (see paragraph [145] above). Also, using the value per square metre for sale R3 2/75 Bruce and applying this to the subject site gives $33,671,000 (see paragraph [138] above), a value appropriately well above $25,200,000. The value for A6/R5, 9/64 Watson of $32,347,000 (see paragraph [142] above) is in a similar position.

Summary of Tribunal's comparative sale analysis

1. The total value of the site on this analysis is therefore:
   • (a) commercial component of $1,150,000;
   • (b) plus residential component of $25,200,000;

_______________________________________

1. The range of analyses available to the Tribunal are therefore as follows:
   • (a) Applicant's hybrid method (realization less profit and risk and civil works) $19,350,000 (see paragraph [51]).
   • (b) Applicant's hybrid method but not less civil works $29,433,000 (see paragraph [51]).
   • (c) Applicant's comparable sales method $11,400,000 (see paragraph [86]).
   • (d) Applicant's full hypothetical development method $12,700,000 (see paragraph [89]).
   • (e) Knight Frank valuation $15,475,000 (see paragraph [97]).
   • (f) MMJ valuation $23,150,000 (see paragraph [98]).
   • (g) Respondent's comparable sales method $27,500,000 (see paragraph [107]).
   • (h) Tribunal's comparable sales method $26,350,000 (see paragraph [161]).
2. In our view the comparable sales method is the appropriate approach(see paragraphs [55] and [81]-[83] above). We have significant concerns about whether the applicant's hybrid method can be relied upon (see paragraphs [55]-[80] above). We also do not think that the full hypothetical development method can be relied upon (see paragraphs [89]-[92] above).
3. We have trouble following how Mr Swinbourne got to his comparable sale method conclusion in his report (see paragraph [86] above). Even the applicant in its closing submissions referred instead principally to the sale over what we have called A7/R6 at block 36 section 64 Watson (see paragraph [87] above). We have relied on this sale in particular for our comparative sales analysis (see paragraphs [143]-[145] above). Contrary to the applicant's submissions, we do not see how this sale supports its hybrid method valuation (paragraph [87] above).
4. The applicant did not press the Knight Frank and MMJ valuations.[148]
5. This leaves the most reliable valuations as the respondent's comparable sales valuation at $27,500,000 (at paragraphs [100]-[110]) and the Tribunal's comparable sales valuation at $26,350,000 (at paragraphs [118]-[161]), which are very close, especially given the range of the other valuations. The Tribunal has some concerns with the respondent's approach (see paragraphs [100], [106], and [111]-[113]). The Tribunal's valuation relies heavily on some sales which both parties accepted as comparable, and a closer analysis of these key comparative sales, and we think it is appropriate that it be substituted for the respondent's assessment.
6. The applicant's hybrid model valuation but without any deductions was $32,726,000. The Tribunal's figure is equivalent to a reduction of this figure by about 20%.This seems a reasonable outcome.
7. The Tribunal therefore finds that the before value (V2) is $7,100,000 and the after value (V1) is $26,350,000. The matter is remitted to the respondent for a determination of the lease variation charge based on these amounts.

[1] Exhibit T1 page T180; exhibit A2, tab 2

[2] Exhibit T1 pages T101 and following, especially pages T180-T183, and page T41 and following.

[4] Exhibit T1 page T101 and following. There was an amendment in relation to the purpose on one of the leases, see exhibit T1 page T39

[8] Respondent's closing submissions at [2]

[9] Respondent's closing submissions at [5]

[10] Applicant's closing submissions at [5] and [107]

[18] See sections 277(1), 277E, 407, 408A and Schedule 1, item 28 of the Planning Act

[19] [2016] ACTSC 275 at [154]- [157]

[20] (1974) 7 SASR 57 at 83

[21] Exhibit A1 at page 39 (30). Exhibit A1 has two numbering systems, the numbering system for the statement, and the numbering system for the report only which is attached to the statement. In these reasons we give the page number for the statement first, and then the page number for the report only second in brackets; applicant's closing submissions at [5] and [149]

[22] Exhibit A1 pages 47(38) and 50(41)

[24] Exhibit A1 page 50 (41); see also applicant's closing submissions at [120]

[25] Exhibit A1 page 51(42); applicant's closing submissions at [5] and [149]

[26] The date of this statement seems to be a mistake; it seems it should be 2 August 2019, since annexure C is dated July 2019.

[28] Respondent's closing submissions at [2]

[34] Exhibit A1 page 45(36); exhibit R2 page 32; exhibit T1 page T574

[35] Exhibit A1 page 43(34); transcript of proceedings 5 September 2019 page 152; applicant's closing submissions at [108]

[40] Exhibit A1 pages 45-46(36-37); transcript of proceedings 5 September 2019 pages 131-2

[43] Exhibit A1 page 45 (36) and 47(38)

[44] Exhibit A1 at page 47(38); exhibit A5; applicants closing submissions at [115]

[46] Exhibit A1 page 46(37); applicant's closing submissions at [116]

[47] Applicant's closing submissions at [116]

[48] Exhibit A5; transcript of proceedings on 5 September 2019 page 80; applicant's closing submissions at [107]; see also respondent's closing submissions at [4]

[49] Transcript of proceedings 5 September 2019 pages 128-9

[50] Applicant's closing submissions at [114]

[51] [1925] AC 439

[54] [1982] HCA 46; (1982) 150 CLR 160

[55] At [10], citing Toohey's case at 443

[57] [2003] HCA 8; (2003) 212 CLR 111

[59] [2007] NSWCA 365; (2007) 156 LGERA 186

[62] At 192, [21]-[22], quoting Estate of James v Valuer-General (1942) 15 LGR (NSW) 110, at 112

[63] Referring to the decision of Biscoe J in Krisgay Pty Ltd v Valuer-General [2007] NSWLEC 600 at [30]

[64] (1989) 72 LGRA 275, 280

[65] Respondent's closing submissions at [34]-[36]

[66] [1999] HCA 64; (1999) 74 ALJR 209 at [280]; quoted by Tobias JA in Commonwealth Custodial Services at 195, [37]

[67] See also Alan A Hyam, The Law Affecting Valuation of Land in Australia (4th ed, 2014) pages 92-94

[68] (1942) 15 LGR (NSW) 110

[69] ( (1989) 72 LGRA 275

[70] Spigelman CJ at 191, [13]-[18]; Santow J at 192, [21]-[22]; Tobias J at 201-204, [58]-[72]

[71] (1987) 64 LGRA 346 at 352-3

[72] Applicant's submissions in reply at [12]

[73] Applicant's closing submissions at [124]; applicant's submissions in reply at [23]

[74] Applicant's closing submissions at [124]-[126]

[75] Transcript of proceedings 5 September 2019 pages 137and 141

[76] Respondent's closing submissions at [30]

[77] Exhibit A1 pages 41-42 (32-33) and 47-48 (38-39)

[78] Applicant's submissions in reply at [35]

[80] Transcript of proceedings on 6 September 2019 pages 23 and 28; exhibit T1 page T302; respondent's closing submissions at [42] and [43]

[81] Exhibit A10; respondent's closing submissions at [6(e)], [40]-[51]

[82] Transcript of proceedings 6 September 2019 pages 24-25 and 28

[85] Transcript of proceedings 5 September 2019 pages 128-9

[86] [2003] HCA 8; (2003) 212 CLR 111 at 120, [16]

[87] [2012] NSWLEC 199 at [196]

[88] ISPT Pty Ltd and City of Melbourne [2007] VCAT 652 at [51]- [59]

[90] Transcript of proceedings 4 September 2019 page 54; transcript of proceedings 5 September 2019 pages 137-144, see especially page 141, line 10, which it seems should say "require more work", that is are closer to the subject site

[96] Applicant's closing submissions at [129] and [138]

[97] Applicant's closing submissions at [129]-[138] especially [129] and [138], and [142]

[99] Applicant's closing submissions at [119]-[120] and annexure A

[100] Applicant's closing submissions at [120]

[101] [2003] WASCA 295 at [65]

[103] Exhibit A6; transcript of proceedings 5 September 2019 pages 157-58

[104] Transcript of proceedings 6 September 2019 page 4

[105] Transcript of proceedings 4 September 2019 page 49; transcript of proceedings 16 September 2019 page 119; applicant's closing submissions at [143]-[146]

[107] Applicant's closing submissions at [5] and [107]

[108] Exhibit T1, pages T204-T221 plus attachments

[109] Exhibit T1 pages T258-T304 plus attachments

[110] Transcript of proceedings 6 September 2019 pages 59-81; exhibit R2 pages 18-26 and 32-35

[112] This is the area shown in the reports at exhibit T1 pages T581 and T244. This is changed to 19,950 square metres in Mr McInerney's statement, exhibit R2, at pages 25 and 34. We have used the area in the original reports.

[114] Applicant's closing submissions at [139]

[115] Respondent's closing submissions at [60]

[116] Applicant's closing submissions at [140]

[117] Respondent's closing submissions at [61]; exhibit R2 page 34, and attachments G and H

[118] It seems that this should be $100,000

[119] Exhibit T1 pages T582 and T237; exhibit R2 page 32; respondent's closing submissions at [13]-[16]

[120] Transcript of proceedings 6 September 2019 page 55

[121] Transcript of proceedings 6 September 2019 page 57.12

[122] Mr McInerney discussed these issues at the transcript of proceedings 6 September 2019 pages 55-59

[123] Exhibit T1, pages T204-T221, see esp. pages T219-T220

[124] Exhibit T1 pages T112-T178. The lots are shown at exhibit A2, document 28

[126] Exhibit A1 at pages 42 (33), 45 (36) and 47 (38); exhibit T1 at pages T583-T584 ; respondent's closing submissions at [22]

[129] Exhibit A1 page 45 (36); exhibit A11

[133] Exhibit A1 pages 41(32) and 47(36); transcript of proceedings 4 September 2019 page 62

[135] Transcript of proceedings 6 September 2019 page 80

[136] Transcript of proceedings 5 September 2019 pages 152-3

[137] Exhibit A1 page 47(38); exhibit R2 page 40

[138] Applicant's closing submissions at [129]-[137]; exhibit A2, tab 25; exhibit A20

[139] Exhibit A1 page 47(38). See also applicant's closing submissions at [129]- [138]

[141] Transcript of proceedings 5 September 2019 pages 140-141

[142] Transcript of proceedings 5 September page 142

[144] Transcript of proceedings 17 September 2019 pages 239-240

[146] Exhibit T1 page T192; exhibit A1 page 40(31); exhibit R2 pages 19, 25 and 33-34; transcript of proceedings 4 September 2019 page 54

[148] Transcript of proceedings 12 November 2019 pages 317-318