Babcock & Brown Properties Pty Ltd v Valuer-General

In respect of all other sales at Bella Vista, it seems to me that the large percentage increases at Blacktown and Wetherill Park industrial estates should not be applied to the Norwest Business Park. But I am not satisfied that no adjustment for movement in the market should be made. In the absence of better evidence being available, I have adopted an increase of approximately 3.9% per annum, or on average 0.325% per month, for the movement in the market between the date of contract and 1 July 2003 for each of the sales at Bella Vista, this increase being slightly less than Mr O'Leary's original figure (par 38 above).

The estimated land value (excluding the value of any consent) of each comparable sale as at 1 July 2003 is given in column I and is then divided by the potential gross floor area relating to the sale (column E), to give a land value for the sale expressed as a price per square metre of GFA, which is set out in column J of Table 2.

In columns K to O of Table 1 the valuers adjusted the estimated land value of each comparable sale (column J) to take account of its various features compared with the subject land. I consider specifically below the more important factors of size (par 50 ff) and location (par 52 ff).

In respect of the factors in column M and N, I have had regard to the descriptive evidence of the valuers and to the figures that they gave in Table 1. In general, the subject land is generally slightly inferior, and I have reflected this by negative percentage adjustment in column M and N of Table 2. However, no overall adjustment is made in column M for sale R1, as the better topography of that property compared with the subject land is counterbalanced by the detriment of an easement crossing the property.

I have made slight changes to the valuers' figures for shape, having regard to what is shown on the deposited plans and in the photographs of exhibits N and O. Because the percentage adjustments are made to the price per sq m of GFA and not to the price per sq m of land, the adjustments should be small. For example, the irregular shape of the northern apex of the subject land probably has little or no bearing on value, as there is ample room for buildings and carparking areas for the agreed GFA elsewhere on the site.

In this case, it has not been necessary to have an additional column for "Other" adjustments. The overall total of the comparative adjustments for each sale is given in column O.

I have then applied the total percentage adjustment in column O to the $/sq m of GFA in column J to produce the $/sq m of GFA for the subject land shown in column P of Table 2. And the figures in column P are then multiplied by the potential GFA of the subject site of 51,740 sq m, to give an indicative land value (not including the value of development consent) for the subject land in column Q, to which must be added the agreed depreciated value of development consent of $1,500,000.

In column K the valuers considered what adjustment (if any) should be made to the comparable sale having regard to the difference in size of the developable gross floor area of the comparable property and of the subject site.

As a general rule, as the gross floor area becomes larger, the price payable per sq m decreases. Conversely, as the GFA becomes smaller, the price increases. Mr Juniper applied this rule when discussing in oral evidence sale A3 in comparison with another property at Bella Vista, but he made no size-adjustments in column K of Table 1 for any of the comparable sales. He agreed that it would not be unreasonable for another valuer to take the view that some adjustment should be made for size. Mr O'Leary considered that the general rule applied and made adjustments for most sales. I agree that adjustments should be made for the differing sizes of GFA, and it seems reasonable to make the adjustments set out in column K of Table 2.

As seen in column L of Table 2, the greatest differences of opinion between the valuers related to the adjustments to be made to the $/sq m GFA of each comparable sale for locational factors compared with the subject land.

The locations of all the comparable sales are shown on the map, exhibit M. A more detailed map showing the layout and size of the comparable sales at Bella Vista is exhibit L.

Both valuers agreed that there is a relativity between land value and rental value. But in the present proceedings there was a paucity of sufficiently reliable evidence to make a good comparison of the rental values between properties at Macquarie Park and the subject land and between properties at Bella Vista and the subject land.

Both valuers referred to a lengthy valuation report of LandMark White dated 7 November 2002, exhibit G, which was prepared for Westpac Banking Corporation. The authors of the report were not called to give evidence. The parties to the present appeal played no role in the preparation of exhibit G, and the Mr C Miller who appeared for the applicant was critical of certain findings in the report. Both valuers in the present appeal referred to the rental evidence that had been gathered by LandMark White at Epping and North Ryde as set out at pp 19-20 of the report, but they did not investigate that evidence themselves. No investigation is reported by LandMark White of rentals at Bella Vista.

At the subject site itself, the "passing rent", that is the rent payable by IBM under the 1993 lease, is not a good indicator of the market rental for the subject land, because the lease was entered into at the time of the sale in 1993, and the rents were then fixed at increasing amounts until June 2018.

Both valuers were of the opinion that the location of each of the two sales R1 and R2 at Macquarie Park are superior to the location of the subject land, but there was a difference between the valuers as to the magnitude of the negative adjustments in column L of Table 1. I agree that these two comparable sales have a superior location to the subject land. But (in the absence of the investigation of rental evidence), I am not inclined to make negative adjustments as large as submitted by the applicant. I also note that the size of the GFA of the properties in the LandMark White report was generally quite small in comparison with the GFA of the subject land, for which adjustments should probably be made. In addition, the location of sale R2 is probably superior to sale R1, because it seems that there would be the opportunity to display a sign on the upper part of the building on R2 identifying the nature of the business to the passing traffic on the adjoining motorway.

It is necessary to make a discretionary judgment to resolve the conflicting views of the valuers concerning locational factors. I have reflected what I consider to be the degree of superiority of the locations of the Macquarie Park sales compared with the subject land by inserting minus 12% for sale R1 and minus 16% in column L of Table 2.

When comparing the subject site with properties at Bella Vista, again the starting point is somewhat uncertain if one is taking the rental of $265/sq m (inclusive of carparking) from exhibit G. In relation to properties at Bella Vista, the valuers referred to the rental of a large property on Norwest Boulevard and Brookhollow Ave occupied by IBM-GSA, where the rental was said to be $238/sq m; an adjustment would have to be made for the older buildings on the subject land and for any other material differences.

Apart from rentals, I have also taken into account the other evidence of the valuers concerning the advantages and disadvantages of the location of the subject site compared with the Bella Vista properties in the Norwest Business Park.

It seems to me that it would be reasonable to allow in column L of Table 2 between about +8% to +15% on top of the price per sq m of GFA of those comparable properties at Bella Vista to take account of the likely overall superior location of the subject land as at 1 July 2003. This range of percentage adjustments also reflects some different locational features of the properties at Bella Vista. For instance, sale A4 has a longer frontage to Norwest Boulevard than sale A3 as well as having an extensive exposure to Old Windsor Road along its western boundary. In addition, sale A4 is at the western gateway entrance to Norwest Boulevard.

As can be seen from column Q of Table 2, the comparable sales indicate different land values for the subject land (exclusive of the value of any consent). Not all indicative values are equally reliable. A further step in the valuation process is therefore required.

I adopt what is said in the American textbook, The Appraisal of Real Estate (12th edition 2001) published by the Appraisal Institute, which is the top professional association in America for real-estate valuers: When many adjustments are applied by the appraiser and when their individual and collective amounts are substantial, the appraiser must ask, " Is the comparable property really comparable? " The greater the amount of collective adjustment, the more the appraiser may reduce the weight placed on a given comparable, or the appraiser may determine that it is not sufficiently comparable to be used at all. [pp 458-459] … Re-examining an appraisal helps ensure its accuracy, its consistency, and the logic of leading to the value indications. An appraiser relies more on professional experience and judgment in reconciliation than in any other part of the valuation process. The appraiser weighs the relative significance, applicability, and defensibility of each value indication and relies most heavily on the approach that is most appropriate to the nature of the appraisal problem. [p 600]

In exhibit W, the valuers gave a weight to each comparable sale according to its reliability. In general, Mr Juniper considered the applicant's sales (prefixed "A" in column A of Table 2) to be more reliable than the sales (prefixed "R") that Mr O'Leary used for the respondent; and Mr O'Leary considered his sales to be more reliable than Mr Juniper's. Mr Juniper attributed the greatest weight or reliability to sale A8, and Mr O'Leary gave the greatest weight to sale R1. It is therefore necessary to resolve the conflicting views on this part of the valuation process.

In column R of Table 2, I have inserted a weight factor for each comparable sale according to what I consider on the available evidence to be the reliability of the comparable sale to indicate an accurate land value for the subject land. I have taken a weight factor of 10 as the theoretical maximum. A factor of 10 would mean that the comparable sale in all respects is perfectly reliable and produces a highly accurate inferred land value. A sale of the subject land itself on the day on which the value was to be determined could achieve the maximum score. At the other end of the scale, a zero weighting means that the sale should not to be used in the ultimate determination of land value of the subject land, because the sale is not comparable or because for some other reason it would be unsafe to rely on the sale.

In general, all of the comparable sales when compared with the subject land have certain deficiencies, and I have not given any of the sales a higher weight than 6.5.

As indicated earlier, there were quite divergent views as to the adjustments to be made to the comparable sales for size and for the location. So even after I have determined what adjustments are to be made for size, those sales that have the least adjustment are likely to be more reliable indicators of value (subject to considering other adjustments). Sale R2 is most comparable to the subject land in size, and so I would have given it a higher weighting than 6.5 but for the fact that this sale suffers from significant adjustments that have to be made in column H (the contract price is altered by $2,380,000). Likewise, the fact that a particular sale occurred close to the base date would be one indicator of reliability, suggesting a higher weighting, but other aspects of the same comparable sale (for example, in respect of some uncertainty as to an appropriate adjustment for size) might indicate that a lower weight factor should be adopted.

I have given less weight to sale A11 than the other sales at Bella Vista. This is because the indicative land value from sale A11 (given in column Q) is somewhat out of line with the indicative values from the other sales at Bella Vista and significantly different from the values inferred from the Macquarie Park sales, as seen in Table 2. Sale A3 is also out of line using the unaltered figures of Mr O'Leary or Mr Juniper in Table 1.

Mr Juniper was of the opinion that sale R4 should not be used at all, as the land was bought to be developed wholly for residential use. In the present case, both valuers agreed that the highest and best use of the subject land was for commercial purposes, for a use similar to that of IBM, even though the zoning of the subject land would permit a residential use. On the present evidence, insufficient analysis has been made of the potential of the subject land to be developed from a practical point of view for residential use as distinct from the current use. Without necessarily precluding further analysis for future years, I am of the opinion that sale R4 should not be used in the present case for the ultimate determination of the land value of the subject land. Accordingly, I have attributed zero weight to that sale in column R of Table 2.

Taking into account the more reliable as well as the less reliable aspects of each particular sale, I have made a discretionary judgment as to an overall weighting factor for each sale, as set out in column R of Table 2.

Using the weights in column R, I have determined the weighed-average land value (not including the value of any consent) of the subject land as follows. In respect of each particular comparable sale, multiply the indicative value of the subject land (in column Q) by the weight attributed to that sale (in column R), to get a resultant value. Add up the resultant values from all the sales, to get the total of resultant values. Then divide the total of resultant values by the total of all the weights. This can be done laboriously with a hand-calculator. Alternatively, the weighted average can be calculated automatically in a Microsoft Excel spreadsheet. The underlying Excel formula in cell Q11 of Table 2 that automatically calculates the weighted-average figure (rounded to the nearest thousand) is: =ROUND(SUMPRODUCT(Q2:Q10, R2:R10)/SUM(R2:R10),-3)

It would not be appropriate, and indeed it would probably be an error of law, to merely adopt an arithmetic or simple average of the indicative land values in column Q, if an informed professional judgment can be made that some of the comparable sales are more reliable than others.

Likewise it would be an error of valuation principle to list the indicative values from all the comparable sales, commencing with the highest value and progressively decreasing to the lowest value, and then to adopt the middle or median value. A median value might be appropriate if (which is not the case here) all the sales were equally reliable or if a reasonable value-judgment could not be made as to the degree of reliability of any of the sales.

In this particular case, the simple average of all the inferred values in column Q of Table 2 happens to be $75,000 less than the weighted average. The difference between a simple average and a weighted average could be much higher or lower, depending on the range of possible inferred values and whether there is more variation in the weights to be given to the different sales.

Of course, the above procedure (of calculating a weighted average) is only one of the ways of weighing the evidence to come to an appropriate ultimate value, but it is a useful procedure where there are many sales involved.

Accordingly, the orders of the Court are: 1. The appeal is upheld. 2. The land value of lot 61 in DP 737386 at Coonara Avenue, West Pennant Hills as at 1 July 2003 is altered to $19,950,000. 3. The exhibits may be returned. __________ A J Nott, Commissioner of the Court

DISCLAIMER - Every effort has been made to comply with suppression orders or statutory provisions prohibiting publication that may apply to this judgment or decision. The onus remains on any person using material in the judgment or decision to ensure that the intended use of that material does not breach any such order or provision. Further enquiries may be directed to the Registry of the Court or Tribunal in which it was generated.