EXECUTIVE SUMMARY WATERFRONT TENANCIES IN NSW. · separately valued wetland occupancies in Sydney...

W ATERFRONTACTIONGROUP A SUBGROUP OF THE BOAT OWNERS ASSOCIATION OF NSW INC

PO Box 639, Spit Junction NSW, 2088 - Phone: 9960 1859 Website: www.waterfrontactiongroup.com.au

""'. "', .... '"',#_-WHAT IS A FAIR MARKET RENT FOR ONE SQUARE METRE

OF THIS LAND? IS IT $1.17 OR $50.40 (BOTH AS CALCULATED BY THE PROPOSED IPART FORMULA IN IPART'S DRAFT

REPORT) OR NEITHER?

13 November 2011

EXECUTIVE SUMMARY

SUBMISSION TO THE IPART REVIEW OF METHOD FOR DETERMINING RENTS FOR DOMESTIC WATERFRONT TENANCIES IN NSW.

RESPONSES TO MATTERS RAISED WITHIN THE OCTOBER 2011 IPART DRAFT REPORT.

Pagelof8

GENERAL COMMENT ON APPROACH TAKEN IN SECTION 3 RELATING TO "VALUING OCCUPANCIES FOR THE PURPOSE OF CALCULATING RENTS"

The perfect matching of Rates of Return (RORs) and Statutory Land Values (SLVs) on a time, value and location basis is fundamental valuation theory. Rates of return are generally higher on low value properties and lower on high value properties (such as waterfront properties).

The approach taken by IPART of applying the median ROR of a postcode or group of postcodes to the median SL V of a group of waterfront SL V s is not sound valuation methodology and is not acceptable to us nor to our expert property valuation advisors.

IfIPARTwishes to use the median SLY ofa group of waterfront properties, then it must use the median ROR for that same group of waterfront properties. The problem is that such reliable data is not readily available. Therefore in our submission we suggested a completely new approach (not to repeat the 2004 fundamental errors) as follows:

SUMMARY OF WAG APPROACH TO SETTING A NEW RENT FORMULA STEP 1 - SET THE PRECINCTS STEP 2 - DETERMINE THE BASIS OF THE RATE OF RETURN (ROR) STEP 3 - MATCH SLVs TO RORs ON EXACTLY THE SAME BASIS STEP 4 - DETERMINE THE DISCOUNT FACTORS STEP 5 - APPLY STEPS 2 TO 4 TO DETERMINE A PRECINCT RENT PER SQUARE METRE STEP 6 - APPLY THE AREA OF THE OCCUPANCY TO THE RENT PER SQUARE METRE, TO ARRIVE AT THE TOTAL RENT PAYABLE.

There is no need to be focused on establishing a waterfront PSLV, as IPART and LPI have done, because the waterfront PSL V is irrelevant to setting a fair market rent. In order to set a fair market rent using a median ROR derived from a postcode or group of postcodes, the median SL V from that same postcode or group of postcodes must be used.

Therefore WAG believes that valuing the occupancies and setting a PSL Vas set out by IPART in Section 3 is a complete waste of time and effort, as it does not assist in determining a fair market rent. In fact, it does just the opposite and causes rents to be set that have no real relevance to the market.

Additionally, in 2006 and 2007 a senior LPI valuer, the NSW Valuer General and the then Chief Valuer ofNSW advised WAG that it is not fundamentally correct to divide their waterfront SLVs by the area of those SLVs, because the results are "meaningless figures" that should not be used to calculate rents.

WAG has avoided this second problem by moving away from the waterfront to the median postcode SL V, which is necessary in any case to match with the postcode median ROR.

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 20f8

We provide examples of proposed rental rate increases of up to 194% and which would not occur if the correct median SLY was applied to the selected ROR. Similarly in Table 9.3 (on Page 112) !PART advises that under its draft recommendations 1,620 rents (being 19% of the total) will increase by an average of 14% and that unacceptable outcome would not happen if the SLY and ROR were correctly matched.

By using the valuing of the occupancy as a starting point in 2004, IPART ended up with a doubly fundamentally flawed formula and it would seem that it is currently headed in the same direction, despite some worthwhile minor tinkering with the sourcing of data, establishment of precinct RORs, weighting by occupancy numbers, matching of ROR and SLY time periods etc. While WAG recognises that the appearance of the model has been greatly enhanced, WAG is disappointed that the fundamental valuation flaws within the formula have not been addressed, resulting in rents still being seriously miscalculated.

WAG is concerned that the LPI consultancy report does not warn !PART of these two fundamental problems with their approach, but notes that none of the 3 NSW Government valuers (including two at a very high level) who advised WAG on the flaws in the IPART fonnula have signed the LPI consultancy report.

SECTION 3.4 DETERMINING PRECINCT BOUNDARIES WAG generally does not have a problem with !PART's setting of precinct boundaries, except for Sydney Harbour. WAG suggested that the current 6 precincts be reduced to 3, not only for ease of administration, but also to overcome the unfair setting of abnonnally high rents for Precincts 2 and 5, because of their smaller than average (for Sydney Harbour) property sizes. WAG believes that these 2 precincts are still unfairly treated under the new proposal and would prefer to see Precincts 2, 3 and 4 combined into one Parramatta River precinct and Precincts 5 and 6 combined into one North Harbour precinct.

SECTION 3.5 APPROPRIATE DISCOUNT MULTIPLIERS

1. During the course of its investigations, WAG consulted numerous property valuation experts and several had stated that they had a high opinion of the valuation capabilities of the Valuer General, Mr Philip Western. We note that the LPI consultancy report has not been signed by Mr Western and we therefore do not have confidence in its fundamental correctness. LPI has confinned to WAG that Mr Western had no input into their report for !PART.

2. Following a long meeting with LPI, WAG and its property valuations experts believe that the LPI consultancy report has many short-comings and cannot be used by !PART as justification to retain a 50% discount multiplier.

3. With the assistance of LPI and NSW Maritime, WAG has located 145 samples of separately valued wetland occupancies in Sydney Harbour, Pittwater and Georges River and been able to compare them to freehold SL V s. Selecting the median (which ignores those valuations that are too high and too low) the discount multiplier results are as follows:

a) Applicable to the median freehold property within a precinct - 13.69%.

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 3 of8

b) Applicable to the median of the PSLV (which included occupancies) -16.59%

c) Applicable to the postcode median - 31.99%

4. WAG has now agreed to pay LPI $5,930 for a further 504 separately valued occupancies and their matching freeholds in the rest ofNSW and that will resolve the matter of the discount multiplier outside of Sydney.

5. At the meeting between WAG and LPI on 9 November, LPI advised that they had not considered the discount factor that relates to the ROR (because the ROR is for land and house packages and we need to bring it back to a ROR for land only). LPI acknowledged that the ROR might be different for land, but could not comment further, as they had not considered the matter. WAG and its property valuation experts have done extensive work in this area and found that the ROR for each precinct needs to have a discount multiplier applied of between 67.67% and 78.1 %, depending on the precinct.

SECTION 4 - THE RATE OF RETURN

We disagree with the IP ART opening statement "Once we have determined how to estimate the value of occupancies, the next element in the rental formula is the Rate of Return (RoR)". As in 2004, IPART has "put the cart before the horse". The RoR must be selected first and then an SLY selected to match with that RoR on a time, value and location basis. IP ART has headed up a "dead-end street" and has no-where to go, because there is no ROR available to match with its chosen PSLV. It must therefore go back and start again and decide on an approach where it can properly match the ROR and the SLY on a time, value and location basis (as we recommended in our June submission).

4.8 WAG and its property valuation experts disagree with IP ART's rejection of the arguments for a discount to be applied to the ROR, in order to reflect that renters are prepared to pay more rent for a house, than for land of equivalent value. Our arguments were supported by evidence, whereas IPART's arguments were not. The Crown Land Act states "the rent shall be the market rent for the land" (not land and house packages) and therefore one needs to determine the "market rent for the land" and IP ART has not done that.

4.9.2 WAG believes that IP ART has confused "capital expenses" (such as the cost of property improvements that are added to the cost base and thereby offset against capital gains) and depreciation, which is an allowable tax deductible expense against rental income. Having now investigated this matter further with several accountants, WAG is under no doubt that depreciation should be included in the differential figure that is applied to the gross ROR, in order to calculate the nett ROR.

WAG disagrees that "the agencies' approach of calculating land tax by using the median Sydney land value is reasonable". Land tax varies greatly based on the value of the property. Land tax forms a very large component of waterfront investment property outgoings and if the ROR is applied to waterfront SLVs, then it is important to include the land tax payable on waterfront SL V s.

IPART's recommended approach (using ATO postcode rental expense figures with a 2 year lag) - WAG is most concerned that the ATO figures include apartments, which have considerably lower expenses and therefore much higher net RORs. Unless the

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 4 0[8

ATO can separate the figures between houses and apartments, these unfortunately cannot be legitimately used.

SECTION 5 - RECOVERING ADMINISTRATION COSTS As per our earlier submission, WAG has a problem with a landlord charging an administration fee, in addition to the market rent. The Crown Lands Act, when it states "the rent shall be the market rent" does not add "plus an administration fee".

WAG notes that IPART proposes to abolish the minimum rent (currently about $430 plus GST) as an offsetting factor, but to all those DWF occupancy holders paying above the minimum rent, that is of no benefit and they will end up paying $220 more. WAG would prefer to see the minimum rent reduced to $275 (including GST) and no administration fee added. We believe that overall our proposal is fairer and in general more accurately reflects the "market rent" intention ofthe Crown Lands Act.

If IP ART insists on the addition of an administration fee, WAG is of the view that it should be based on $100 per occupancy (not $200) plus GST, but rather recovered as a small percentage surcharge on the rent, so that someone paying a fair market rent of $20,000 per annum will pay a 40 times greater administration fee than someone paying a fair market rent of$500.

CONCLUSION ON THE RENTAL FORMULA AND RENT CALCULATIONS

The proposed formula is still seriously flawed with the huge variance in current rents from $1.17 to $50.04 per square metre not representing anything close to market rents, as was required by the 2003 Terms of Reference, intended by IPART in 2004, as required by the Crown Lands Act and as we think was intended by the Government in the current Terms of Reference. Some of the reasons for the proposed formula not calculating fair market rents are as follows: I. In many cases, under what IP ART is proposing, the owners of smaller properties

will be paying higher rents and be subsidising the owners of larger properties, who will be paying lower rents (e.g. Parramatta River).

2. There is also a serious mis-match between SLVs and RORs on a value and place basis, which will cause some serious over-charging to occur.

3. The use of A TO data for apartments and houses, as the basis of the differential between gross and net ROR is flawed. That differential is not that far wrong at the current level of2.5%, as set by IPART in 2004. If the apartments figures are removed from the data supplied by the ATO, WAG believes that the houses figures remaining per precinct will vary between about 2.5% and 3.1 %.

4. The discount multiplier as applied to SLVs needs to be reduced from the 50% level to 32.% and be applied to the postcode based median SLY (to achieve proper matching).

5. A discount multiplier needs to be applied to the ROR, to allow for the fact that the ROR on land is lower than for land and house packages.

WAG recommends that the still seriously flawed formula is replaced with another, based on expert property valuation advice, to achieve fair market rents, and that the new formula is thoroughly tested before implementation. The proposed rent increases, averaging 14% for 1,620 DWF occupancy holders (see table 9.3 on page 112) and of WATERFRONT ACTION GROUP A Subgroup of the Boat Owners Association ofNSW inc Page 5 of8

up to194% in 9 current precincts, just because the ROR was not matched to the SL V, proves that while the old 2004 formula has a "flashy new coat of paint", underneath it is still built on a weak foundation, that will not withstand some serious prodding.

THE NEW FORMULA - AS RECOMMENDED BY WAG

STEP 1 - SET THE MEGA PRECINCTS The current proposal to reduce 61 precincts to 20 appears acceptable, except for Sydney Harbour, where we believe that the 6 precincts should be reduced to 3.

STEP 2 - DETERMINE THE BASIS OF THE ROR We agree with IPART's proposed setting of the ROR, except for the differential between gross and net.

STEP 3 - MATCH SLVS TO RORs ON EXACTLY THE SAME BASIS Fundamental valuation theory demands that RORs and SL V s are perfectly matched on a time, value and place basis. IP ART has correctly matched on a time basis, but not on a value nor on a place basis.

STEP 4 - DETERMINE THE DISCOUNT FACTORS In steps 1.2 & 3, we have determined the factors that would enable the calculation of a fair market rent on house and land packages within a precinct. But we are dealing with vacant land, on which the ROR is lower, so the ROR needs to be discounted accordingly.

And we are dealing with generally submerged land, held on uncertain tenure and which has severe restrictions placed on what can be done with it. This land is therefore less valuable than dry freehold tenure land, on which one can build a permanent dwelling. Therefore the median postcode SLY needs to be heavily discounted, according to evidence of how historically the Valuer General has discounted such wetland, compared to median postcode SL V s.

These 2 discount factors need to be included in the formula.

STEP 5 - APPLY THE AREA OF THE OCCUPANCY Steps I, 2, 3 & 4 will give us a precinct market rent per square metre. The fmal step is to multiply that by the occupancy area to arrive at the rent payable for a particular occupancy.

OTHER MATTERS

SECTION 6 - PENSIONER CONCESSIONS AND CASES OF HARDSHIP Pensioner Concessions - WAG would like to see the benefits offered to those on a full pension extended to those on a part pension. WAG is aware of people on a part


pension struggling to keep their family homes, because of the huge wetland rents being charged by Maritime and CLD. They also cannot afford to demolish their waterfront structures and surrender their lease or licence, so they are faced with having to seII the family home ..

Hardship Arrangements - In regards to Draft Recommendation 22, WAG would like to see the representative from "at least one independent organisation" on the review panel increased to two (to equal the two representatives drawn from the agencies) and for the "independent organisation" to be more clearly specified as an organisation with considerable financial skiIIs (such as the NSW Audit Office), so that as with the ROR, the agencies do not suit themselves, because a recommendation was insufficiently precise.

SECTION 7 - CONCESSIONS FOR OTHER CIRCUMSTANCES

7.3 Long Jetties - WAG disagrees with the recommendation not to grant a concession to the owners of long jetties, which are long because they have to be long to reach navigable water. It seems unfair that in the same precinct one jetty owner might have to pay say 4 times the rent for the land under a long jetty, that was expensive to build and is costly to maintain, compared to the rent that another owner pays, both being for the same purpose - to reach navigable water. The additional rent paid by the owner of the long jetty provides no additional benefit.

7.6 Other proposed discounts or concessions - WAG disagrees that some of its proposals for discounts "do not warrant special concessions or are already accounted for in the rental formula (via the PSLV or the discount multiplier)". For example-

• Someone with a jetty in shaIIow water and who can only use the jetty at high tide should not have to pay the same rent as someone else in the same precinct with a jetty in deep water and who can use their jetty 100% of the time.

• Someone who must provide access to utilities on a portion of the rented land is unable to build a stlUcture on that portion of the land, but must maintain it for the benefit of the utilities. The "market rent for the land" the subject of such impediment will be lower, but the rent calculated by the formula does not reco guise that.

8.1 NSW Maritime lease I right of appeal

a) NSW Maritime lease - WAG has some concerns with the rent clauses, including the proposed wording for clause 3.7.

b) Right of Appeal - WAG is most concerned that IP ART is recommending that NSW Maritime occupancy holders are left without any rights of appeal to an independent tribunal or court. That would aIIow the current situation to continue, whereby CLD occupancy holders can appeal to the Local Land Board orland the Land and Environment Court (and then to higher courts), but NSW Maritime occupancy holders have no such opportunity.

WAG has numerous other serious issues with the NSW Maritime lease, but the responsible Minister (the Hon. Duncan Gay MLC) has promised that a "high level

W ATERFRONTACTIONGROVP A Subgroup of the Boat Owners Association ofNSW inc Page 7 of8

working group" will consider the terms of the Lease, following receipt of IP ART's recommendations, and WAG will raise those other issues at that time.

9.1 Capping of rent increases - WAG agrees with the $200 annual rental increase cap, but would prefer to see that extended from I year to 3 years (i.e. $200 per year for 3 years).

9.3 Impact on the agencies - WAG understands that there are situations where structures have been built on wetland without the knowledge of the agencies and that rent is not being paid on that wetland. Better policing will result in the agencies (particularly CLD) increasing its revenue base.

9.4.4 Impact on W AO occupants - WAG considers that the proposed 39% average percentage rental reduction (based on IPART's assumptions) is far too small, considering that IP ART has recognised the validity of the argument that WAO (Water Access Only) occupants have a right to free primary access to their properties.

9.4.5 Impact on pensioners - WAG notes that under the draft recommendations 4 I pensioners will receive rental increases averaging $46 and believes that IPART should recommend that annual increases for pensioners should be limited to the CPL

W ATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 8 of8

WATERFRONTACTIONGROUP A SUBGROUP OF THE BOAT OWNERS ASSOCIATION OF NSW INC


WHAT IS A FAIR MARKET RENT FOR ONE SQUARE METRE OF THIS LAND? IS IT $1.17 OR $50.40 (BOTH AS CALCULATED

BY THE PROPOSED IPART FORMULA IN IPART'S DRAFT REPORT) OR NEITHER?

I3 November 2011

SUBMISSION TO THE IPART REVIEW OF METHOD FOR DETERMINING RENTS FOR DOMESTIC WATERFRONT TENANCIES IN NSW - WAG'S RESPONSE TO THE OCTOBER 2011 DRAFT REPORT

A. WHO ARE WE? The Waterfront Action Group (WAG) represents thousands of domestic lease and licence holders throughout NSW and who are wetland tenants ofNSW Maritime (Maritime) and the Land and Property Management Authority (LPMA). We are a sub-group of the Boat Owners Association ofNSW Inc. (BOA), the peak body in NSW representing the interests of boat owners. References to both WAG and BOA were included in the September 2009 Auditor-General's report to Parliament ("Administering Domestic Waterfront Tenancies") in terms of "represents stakeholders" and were the only such groups quoted in that report. WAG is also quoted hundreds oftimes in the 1P ART Draft Report.

Page 1 of 15

B. WHY ARE WE INVOLVED? There is a high level of concern and anger amongst our members, in regards to rentals which have escalated alarmingly since 2004, because of the failure of Maritime and LPMA to follow the recommendation of IP ART in its April 2004 report ("Review into Rentals for Waterfront Tenancies on Crown Land in NSW") being that "The rate of return will need to be regularly reviewed". The Terms of Reference of the IPART report include-

"to align rental returns to reflect and maintain market value", and refer to "The Department of Lands and the Waterways Authority wish to develop a market formula for rental returns to reflect market value at each rental review".

That has not happened since 2004, because of the failure to review the rate of return, so that rents have increased in line with waterfront property values, rather than in line with waterfront property rents as we believe that IPART intended.

Additionally, WAG has received advice in 2006 and 2007 from the NSW Valuer General and the Chief Valuer ofNSW and a property valuer representing them, that the April 2004 IP ART PSLV formula is "fundamentally flawed" and produces "meaningless figures". The proposed PSL V formula in the IP ART Draft Report still includes the same basic flaw, so that issue has not been addressed.

C. EXPERT ADVICE WAG has sought rate of return advice from numerous specialist academics in various universities in NSW and Queensland including:

a) Professor Alfred C. Eves, Professor in Property Economics at Queensland University of Technology.

b) Professor Bill Randolph, Faculty of the Built Environment, University of New South Wales, Director of the UNSW/uWS AHURI Research Centre.

c) Mr. Peter Wills, Senior Lecturer, School of Economics and Finance, University of Western Sydney.

WAG has sought advice from numerous valuation consultants in private practice including:

d) Mr Terry Dundas, Egan National Valuers (NSW)

e) Mr Paul Chaloner, Chaloner Valuations, but formerly with Egan National Valuers (NSW)

f) Mr. Frank Large, Egan National Valuers (NSW)

g) Mr Robert Rowlands, Landsbury's International Pty Ltd.

h) Mr Phillip Barlow, Landmark White (NSW) Pty Ltd.

The advice that WAG has received has been consistent from all sources and that is: 1. That both Rates of Return (RORs) and Statutory Land Values (SLVs) are value,

area and time specific, so that they apply to a particular value range, location (or area) at a particular time.

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 2 of 15

11. That as RORs and SLVs are generally in a direct inverse relationship, for one to be applied to the other, both must be treated in the same manner, on a time and place (or area) basis.

111. If one applies a ten year rolling average ROR for one area to a three year rolling average of SL V s for another area and of a different time period, the resulting answer is nonsense and cannot be used for the purposes of calculating market rents.

D. COMMENTS ON "THE MARKET" AND "MARKET RETURN" FOR DOMESTIC WATERFRONT OCCUPANCIES The Terms of Reference for the April 2004 IP ART report included two key statements in relation to market rents and market value as follows:

"The Department of Lands and the Waterways Authority wish to develop a market formula for rental returns to reflect market value at each rental review" and " ... .the Tribunal is to consider ... aligning rental returns to reflect and maintain their market value"

Additionally, the Crown Lands Act 1989 No.6 at Section 143(1)(a) states "the rent shall be the market rent for the land ... ".

WAG is of the view that the formula recommended by IPART in April 2004 and its poor application by Maritime and LPMA since then, has resulted in rents substantially above (and in a few cases below) market value and in that respect IPART and the two agencies have failed to meet the expressed key aim "for rental returns to reflect market value at each rental review".

The definition of "market" or "market value" is extremely important in this context and it is clearly understood by property valuers and lawyers. The Australian Property Institute defines "market value" as:

"Market Value is the estimated amountfor which an asset should exchange on the date of valuation between a willing buyer and a willing seller in an arms length transaction after proper marketing wherein the parties had each acted knowledgeably prudently and without compulsion."

The High Court of Australia in Spencer v The Commonwealth of Australia enunciated similar principles including "a willing but not anxious vendor and purchaser".

The current situation where NSW Maritime and LPMA demand huge rents from waterfront property owners, under duress and threats of termination of their lease or licence and demolition of their structures, or in the case of reclaimed land, to lease it to a neighbour or convert it to a kayak stop, does not reflect a true market situation. The comments within Paragraph 3.1 of the IP ART Issues Paper and which suggest that the current situation is reflective of a "market" are therefore not correct.

WAG and its advisors are of the view that the market rent for land below the mean high water mark is not the most that can be extracted from the owner of the adjoining freehold land (i.e. the current situation), but rather what rent others would be prepared

W ATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 3 of 15

to pay for that land, if it was able to be freely offered to the market in a normal manner.

E. RESPONSES TO MATTERS RAISED WITHIN THE IPART DRAFT REPORT

NOTE ON THE LIMITED SCOPE OF OUR RESPONSE - We are not attempting to respond in detail on every issue, but are limiting our response to a focus on key areas on the calculation of wetland rents, where we believe that substantial errors have been made and which if properly addressed will cause fair market rents to be charged for the benefit of DWF occupancy holders and the NSW Government (and as required by the Crown Lands Act). The current IP ART proposal does not fully eliminate present inequitable situations and it is our aim to assist IP ART to overcome these inequities. There are also some additional comments on some matters outside of rental calculation and where we feel greater equity could be achieved.

GENERAL COMMENT ON APPROACH TAKEN IN SECTION 3 RELATING TO "VALUING OCCUPANCIES FOR THE PURPOSE OF CALCULATING RENTS"

The perfect matching of Rates of Return (RORs) and Statutory Land Values (SL V s) on a time, value and location basis is fundamental valuation theory. This was made clear by us and our property valuation expert advisors in our and their submissions. The chart on page 143 of the IPART Draft Report is similar to charts in our submission and clearly demonstrates that rates of return are generally higher on low value properties and lower on high value properties (such as waterfront properties). Such a chart is included in Fact Sheet 6, which covers the important matter of matching SLVs and RORs by place and by value.

We refer to the letter of endorsement (dated 8 October 2009) of the Egan Pittwater Report from Professor Alfred Eves, which letter was included with our submission and which discusses the result of a 2005 study by Eves and Adair and which concluded that there was "a lower income return for waterfront property compared to lower value properties in the same postcode and Local Government Area location". That letter from Professor Eves is annexed to this submission, as it covers in some detail the matter of waterfront properties being in a different market to non-waterfront properties.

Surrounding an average precinct, as determined by IP ART, there will be many hundreds (if not thousands) of properties, the median rent and sales price of which will be selected to determine the ROR for that area. As waterfront properties, within a postcode or group of postcodes, are generally very few in number and have high sales prices and low RORs, the median rent and sales price will most certainly be of a nonwaterfront property with a relatively low sales price and high ROR. Therefore the approach taken by IP ART of applying the median ROR of a postcode or group of postcodes to the median SL V of a group of waterfront SL V s is not sound valuation methodology and is not acceptable to us nor to our expert property valuation advisors. More detail on this matter is covered in Fact Sheet B.


IfIPART wishes to use the median SLY of a group of waterfront properties, then it must use the median ROR for that same group of waterfront properties. The problem is that such reliable data is not readily available. Therefore in our submission we suggested a completely new approach (not to repeat the 2004 fundamental en-ors) as follows:

SUMMARY OF WAG APPROACH TO SETTING A NEW RENT FORMULA STEP 1- SET THE PRECINCTS STEP 2 - DETERMINE THE BASIS OF THE RATE OF RETURN (ROR) STEP 3 - MATCH SLVs TO RORs ON EXACTLY THE SAME BASIS STEP 4 - DETERMINE THE DISCOUNT FACTORS STEP 5 - APPLY STEPS 2 TO 4 TO DETERMINE A PRECINCT RENT PER SQUARE METRE STEP 6 - APPLY THE AREA OF THE OCCUPANCY TO THE RENT PER SQUARE METRE, TO ARRIVE AT THE TOTAL RENT PAYABLE.

There is no need to be focused on establishing a waterfront PSLV, as IP ART and LPI have done, because the waterfront PSL V is in-elevant to setting a fair market rent, unless the ROR is also set by reference to that same location and value range. The comments ofLPI on page 14 (4th paragraph) of their report are therefore not relevant to the setting of a fundamentally con-ect fair market rent and should be ignored. In order to set a fair market rent using a mediau ROR derived from a postcode or group of postcodes, the mediau SL V from that same postcode or group of postcodes must be used.

Therefore WAG believes that valuing the occupaucies and setting a PSLV as set out by IPART in Section 3 is a complete waste oftime and effort. as it does uot assist iu determiniug a fair market rent. Iu fact. it does just the opposite and causes rents to be set that have uo real relevance to the market.

Additionally, in 2006 and 2007 a senior LPI valuer, the NSW Valuer General aud the then Chief Valuer ofNSW advised WAG that it is not fundamentally con-ect to divide their waterfront SLVs by the area of those SLVs, because the results are "meaningless figures" that should not be used to calculate rents. That is because much of the value of waterfront properties is in the access to the water, fabulous views, potential lifestyle etc, rather than being related to size. It was explained that providing two waterfront properties in a particular location can only support one dwelling, whether one is twice or three times the size on another does not greatly affect its value. Therefore dividing the value by the area (as per the PSLV formula) is a nonsense. That original advice from senior government property valuers has been confirmed with numerous other property valuation experts in private practice and in various universities. More detail on this advice is covered in Fact Sheet B

WAG has avoided this second problem by moving away from the waterfront to the median postcode SLY, which is necessary in any case to match with the postcode median ROR.


In Fact Sheet A we provide 9 examples of proposed rental rate increases of up to 194% and which would not occur if the correct median SLY was applied to the selected ROR. Similarly in Table 9.3 (on Page 112) IP ART advises that under its draft recommendations 1,620 rents (being 19% of the total) will increase by an average of 14% and that unacceptable outcome would not happen if the SLY and ROR were correctly matched.

By usiug the valuing of the occupancy as a starting point in 2004. IP ART ended up with a doubly fundamentally flawed formula and it would seem that it is currently headed in the same direction, despite some worthwhile minor tinkering with the sourcing of data, establishment of precinct RORs, weighting by occupancy numbers, matching of ROR and SLY time periods etc. While WAG recognises that the appearance of the model has been greatly enhanced, WAG is disappointed that the fundamental valuation flaws within the formula have not been addressed, resulting in rents still being seriously miscalculated.

WAG is concerned that the LPI consultancy report does not warn IPART of these two fundamental problems with their approach, but notes that none of the 3 NSW Government valuers (including two at a very high level) who advised WAG on the flaws in the IPART formula have signed the LPI consultancy report.

SECTION 3.4 DETERMINING PRECINCT BOUNDARIES

WAG generally does not have a problem with IPART's setting of precinct boundaries, except for Sydney Harbour. WAG suggested that the current 6 precincts be reduced to 3, not only for ease of administration, but also to overcome the unfair setting of abnormally high rents for Precincts 2 and 5, because of their smaller than average (for Sydney Harbour) property sizes (which causes the PSLV to be artificially inflated by a smaller denominator in the flawed PSL V formula) . WAG believes that these 2 precincts are still unfairly treated under the new proposal and would prefer to see Precincts 2, 3 and 4 combined into one Parramatta River precinct and Precincts 5 and 6 combined into one North Harbour precinct. NSW Maritime's total revenue should not be affected by that change, but it will be more equitable for the DWF occupancy holders involved and remove a cause of frequent complaints.

There is no logical reason for occupancy holders in Bahnain to pay the proposed $45.64 per sqm, while occupancy holders on the northern side of the same river (e.g. Hunters Hill) pay the proposed $18.86 per sqm. Similarly, there is no logical reason why DWF occupancy holders in Drummoyne Avenue Drummoyne should pay either the proposed $45.64 per sqm or the proposed $20.81 per sqm, depending on which side of the Gladesville Bridge they reside. That is simply the product of a flawed formula, as was pointed out by the NSW Valuer General and Chief Valuer ofNSW in 2006 and 2007.

The reduction of precincts in Sydney Harbour from 6 to 3 does not improve the accuracy of the proposed rents. But it spreads the pain (of over-charging) more ·evenly, so that all DWF occupancy holders in the Parramatta River suffer equally, with the same result in the Kirribilli to Manly area. We have attached an "Analysis of NSW Maritime Precinct Data", which illustrates that Precincts 2 & 5 suffer higher


than normal rents, partly because the average property size is smaller than Precincts 1, 3,4&6.

SECTION 3.5 APPROPRIATE DISCOUNT MULTIPLIERS

1. During the course of its investigations, WAG consulted numerous property valuation experts and several had stated that they had a high opinion of the valuation capabilities of the Valuer General, Mr Philip Western. At a meeting on 1 June 2007 with the current WAG Chairman and the then BOA President, Mr Western indicated that he was prepared to provide advice to lPART, if requested. WAG was recently assured by lPART that the Valuer General was being consulted and the Roads and Ports Minister recently provided a similar assurance. We note that the LPI consultancy report has not been signed by Mr Western and we therefore do not have the same confidence in its fundamental correctness. On 9 November when WAG and LPI met to discuss their report, LPI confirmed that Mr Western had no input into their report for lPART.

2. WAG and its property valuations experts believe that the LPI consultancy report has many short-comings and cannot be used by lP ART as justification to retain a 50% discount multiplier. WAG has many reasons for being extremely disappointed with the LPI consultancy report and a list of some of those reasons is annexed under the title "Some Criticisms of the LPI Consultancy Report".

3. In its analysis of 144 examples of separately valued wetland occupancies (which was supplied by LPI and NSW Maritime), WAG found a large number of suspect valuations (some obviously too high and some obviously too low). For that reason we eliminated the top 25% and the bottom 25% and calculated the average from the middle 50% (being 72 of the original 144). LPI has criticised us for having too small a sample (72 compared to their sample of 52). NSW Maritime has recently assisted WAG to locate a few duplications (caused by different area sizes in their data base compared to the LPI data base). In addition, with the assistance of new data provided by NSW Maritime, WAG has located some additional samples, that were not previously used. At the time of writing we now have 145 samples in Sydney Harbour, Pittwater and Georges River and selecting the median (which ignores those valuations that are too high and too low) the discount multiplier results are as follows:

a) Applicable to the median freehold property within a precinct - 13.69%.

b) Applicable to the median of the PSLV (which included occupancies) - 16.59%

c) Applicable to the postcode median - 31.99%

NOTES: i) The reduction caused by the discount multiplier is 100 less the discount multiplier, so in the above 3 cases the reductions are a) 86.31 %, b) 83.41 %, c) 68.01 %.

ii) The above figures are derived solely from Sydney samples, so that the results do not necessarily apply to areas outside of Sydney. WAG has agreed to pay LPI $5,930 for additional data for the rest ofNSW and which will determine the discount factors applicable to areas outside of Sydney.

iii) The 13.69% discount mUltiplier in a) is derived from comparing the value of the occupancy to the combined value of the adjoining freehold and the occupancy


in each case, where possible. Where there was no adjoining freehold we compared to the median PSL V.

iv) The 16.59% discount multiplier in b) is derived from comparing the value of the occupancy to the median PSLV in every case:

v) WAG has a strong preference for the postcode median option, because under valuation perfect matching principles, the ROR, the SL V and the discount multiplier should all be matched. As the ROR is only available on a postcode median basis, that requires the other two factors to be matched to the postcode median.

vi) Fact Sheet C relates to this exercise.

4. We reject the criticisms of our analysis for the many reasons listed in the annexure titled "Some Criticisms of the LPI Consultancy Report".

5. We attach Fact Sheet 8 "Calculating the Discount Factor", which covers this matter in more detail.

6. At the meeting between WAG and LPI on 9 November, LPI advised that they had extracted another 504 separately valued wetland occupancies for areas outside of Sydney and that they would release those to WAG on payment of $5,930. WAG has agreed to purchase that additional data, believing that to consider ALL the separately valued occupancies in the state is the answer to this problem. These valuations are done by professional property valuers contracted by LPI and who therefore have LPI's confidence. While there may be examples of some valuations either too high or too low, selecting the median ignores those unusual samples.

7. At the meeting between WAG and LPI on 9 November, LPI advised that they had not considered the discount factor that relates to the ROR (because the ROR is for land and house packages and we need to bring it back to a ROR for land only). LPI acknowledged that the ROR might be different for land, but could not comment further, as they had not considered the matter. WAG and its property valuation experts have done extensive work in this area (see Fact Sheet 10) and found that the ROR for each precinct needs to have a discount multiplier applied of between 67.67% and 78.1 %, depending on the precinct. This discount multiplier is particularly important for regional areas, where the ROR for house and land packages is much higher than for land only. Fact Sheet D explains this in greater detail.

SECTION 4- THE RATE OF RETURN We disagree with the IP ART opening statement "Once we have determined how to estimate the value of occupancies, the next element in the rental formula is the Rate of Return (RoR)". As in 2004, IP ART has "put the cart before the horse". The RoR must be selected first and then an SL V selected to match with that RoR on a time, value and location basis. IP ART has headed up a "dead-end street" and has no-where to go, because there is no ROR available to match with its chosen PSL V. It must therefore go back and start again and decide on an approach where it can properly match the ROR and the SLY on a time, value and location basis (as we reconunended in our June submission).

4.3 We disagree with the term "a proper commercial return" in the Terms of Reference. We believe that in relation to domestic waterfront tenancies it would have

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 8ofl5

been more relevant and less confusing to use tbe term "a fair market rent", which has a clearly specified meaning to property valuers and to lawyers. In any case, we would suggest that in regards to property "a proper commercial return" is the same as "a fair market rent". The Crown Lands Act specifies (in Section 143 (1) (a) "the rent shall be the market rent for the land" and we believe that is the correct focus. IP ART's finding in 4.3.2 that" .... the NSW Government.. . .is entitled to a market return ... " is close to our thinking, but we would prefer to see that further refined to "a fair market rent", so tbat there is absolutely no confusion as to the meaning of "a market return". There are many market returns, but in the case of renting out land, the applicable "market return" is "a fair market rent" for that land.

4.4.2 We agree that "the residential rental ROR on houses" can be used "as tbe proxy for the ROR on DWF occupancies" providing it is adjusted to reflect that DWF occupancies are land and not houses and providing tbe ROR is properly matched to SL V s on a time, value and location basis.

4.5.2 We agree with IPART's fmding.

4.6.2 We agree with IPART's finding.

4.7.2 We agree with IP ART's findings except for the decision to apply a ROR based on postcode medians to a PSL V based on waterfront properties. There is a fundamental mismatch here on a value and location basis. The correct match to a postcode median based ROR is a postcode median based SL V. lP ART has recognised the inverse relationship between RORs and SLVs on a value basis and that RORs are location specific, but tben applies a correctly selected ROR to a wrongly selected SL V. The postcode ROR is calculated using a median postcode rent and a median postcode SLY. One cannot then apply that to an SLY of a much higher value and in a specific location that has a very different (i.e. lower) ROR.

For example, in Table 4.1 Brisbane Water is shown with a Gross ROR (GROR) of 4.28%, calculated from the postcode medians. But the postcode medians in tbe area range from $365,000 to $548,000, with an average not far from $430,000. The 2010 Postcode ROR - All Sales Value chart in Annexure 10 (or IPART's chart at page 143) shows that the ROR falls very steeply from that level, so that by the time it reaches waterfront property values of say around $1,000,000 the GROR has dropped to about 3%. That difference of about 1.3% in the GROR flows straight through to the net ROR (NROR) and if the differential between GROR and NROR is say 2.7%, then tbe NROR is about 0.28%, rather than 1.58% derived from the failure to properly match tbe ROR to the SL V, as IPART has done. This error alone has caused tbe Brisbane Water rent to be about 5 times what it should be and there will be many other similar examples throughout tbe state.

Table 4.1 The variation in the Gross RORs from 1.98% to 4.62% is very significant and illustrates that lP ART is correct in finding a single ROR cannot be applied across the state. It is interesting to note that tbe lowest ROR is in Sydney's eastern suburbs, where property values are high and that the higher RORs are where property values are low. That reinforces the fact that RORs tend to vary in inverse proportion to property values (so that a ROR for say a $600,000 median postcode property cannot be applied to say a $3,000,000 waterfront property).

4.8 WAG and its property valuation experts disagree with lPART's rejection of the arguments for a discount to be applied to the ROR, in order to reflect that renters are prepared to pay more rent for a house, tban for land of equivalent value. Our arguments were supported by evidence, whereas IPART's arguments were not. Land WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 9 of 15

is an appreciating asset, while houses depreciate in value over time. An investor needs a higher rental return on the house, that will eventually need to be replaced (or substantially upgraded), whereas a lower rental return on land is usually compensated by an eventual capital gain, when that land is sold. "The likely explanation" put forward by IPART is simply an untested and unsupported theory. Our work on the discount that should be applied to the ROR was checked and endorsed by Professor Alfred Eves, Professor in Property Economics at the Queensland University of Technology and an acknowledged expert on the ROR. IPART's finding that we are wrong therefore also flows through to the fact that they believe that Professor Eves is wrong.

The Crown Land Act states "the rent shall be the market rent for the land" (not land and house packages) and therefore one needs to determine the "market rent for the land" and IP ART has not done that. Annexed hereto is Fact Sheet 10 (from our June submission) and which fully explains why the ROR for house and land packages needs to be discounted to bring it back to "the market rent for the land".

4.9.2 WAG believes that IP ART has confused "capital expenses" (such as the cost of property improvements that are added to the cost base and thereby offset against capital gains) and depreciation, which is an allowable tax deductible expense against rental income. The deduction in tax returns is described as "Deduction for decline in value of depreciating assets". WAG has correctly apportioned depreciation against the rental return. WAG has supplied IP ART with the written opinion of its accountant, who states "Depreciation most definitely IS a tax deductible expense, claimable against rental income .... ". So WAG believes that IP ART needs to go further than recognising that "depreciation represents expenses borne by an owner of an investment property" and recognise that it is a legitimate tax deductible expense. Apart from the tax deductibility of depreciation, it is an expense recognised by the accounting profession and as well as recognising it as an expense against rental income, the value of depreciation is applied against the value of the building each year. That annual depreciation expense recognises that over time the building reaches a point when it either has to be replaced or substantially upgraded. Having now investigated this matter further with several accountants, WAG is under no doubt that depreciation should be included in the differential figure that is applied to the gross ROR, in order to calculate the nett ROR.

WAG disagrees that "the agencies' approach of calculating land tax by using the median Sydney land value is reasonable". Land tax varies greatly based on the value of the property. The tax free threshold of$387,000 (SLV) means that many low value properties pay either no land tax or very little. On the other hand, the rate increases from 1.6% to 2%, once SLVs exceed $2,366,000 (and many Sydney waterfront properties exceed that value). Once again proper matching is extremely important (this time on a value basis) and the median Sydney land value cannot be matched to anything else that IPART (or anyone else) is considering. Land tax forms a very large component of waterfront investment property outgoings and if the ROR is applied to waterfront SL V s, then it is important to include the land tax payable on waterfront SLVs.

IPART's recommended approach (using ATO postcode rental expense figures with a 2 year lag) - WAG is concerned about the 2 year lag, but this might be largely overcome by using the 3 year average ROR figure, rather than the latest year figure. What is achieved by dividing the total expenses within a postcode by the total rental income for that postcode is an average for that postcode. That does not quite match


with the ROR, which is median based, but is probably close enough. However, WAG is most concerned that the ATO figures include apartments, which have considerably lower expenses and therefore much higher net RORs. Unless the ATO can separate the figures between houses and apartments, these unfortunately cannot be legitimately used. This explains why the figures are so low in Sydney Harbour precincts, where there are large numbers of apartments and why they vary so much from the Egan National Valuers postcode based figures and which we still believe to be accurate (varying between 2.52% and 2.93% over 12 current precincts).

SECTION 5 - RECOVERING ADMINISTRATION COSTS As per our earlier submission, WAG has a problem with a landlord charging an administration fee, in addition to the market rent. The Crown Lands Act, when it states "the rent shall be the market rent" does not add "pins an administration fee". At the "ronndtable" discussion there were many complaints about the fact the agencies provide very little by way of service and that the only contact from the agencies is receipt of an annual rent invoice. WAG considers that an additional fee of $220 (including GST) for doing little more than producing an invoice is wildly excessive and will encourage continued inefficiency within the agencies. If, as a result of this current process, rents reduce back to a fair market rent, the agencies will find that DWF occupancy holders will pay their invoices promptly and the agencies' substantial collection costs (including legal costs) will reduce dramatically. WAG and its members will also cease legal actions disputing the rents charged and that will save significant legal expenses and executive time.

WAG noted that the lAB report was heavily qualified and in some cases expenses only allocated on the basis of staff interviews, rather than more reliable evidence. WAG therefore believes that not too much reliance should be placed on the lAB report. We note that the invoicing costs for 6,890 CLD waterfront tenancies is estimated by lAB to be between $8,000 and $12,000 per annum. Taking the mid point of $1 0,000, that calculates out at $1.45 per occupancy and WAG believes that particular figure to be a reasonable estimate. WAG has no issue with the $2,500 that IPART wishes to charge for annually updating the ROR, but that only adds another 36 cents per occupancy. WAG has great difficulty accepting that the costs of running an efficiently run division, solely devoted to administering 6890 DWF occupancies would cost more than $100 per annum per occupancy and considers that the job could be outsourced to private enterprise for about that figure.

WAG notes that IPART proposes to abolish the minimum rent (currently about $430 plus GST) as an offsetting factor, but to all those DWF occupancy holders paying above the minimum rent, that is of no benefit and they will end up paying $220 more. The addition of a $220 administration fee is a huge percentage addition to someone paying a low rent (it is 44% extra for someone paying $500 annual rent), but a small percentage addition to someone paying a high rent (it is l.l % extra for someone paying $20,000 annual rent). As stated in its earlier submission, WAG would prefer to see the minimum rent reduced to $275 (including GST) and no administration fee added. That will mean that in some cases, under our proposal, the agencies will receive $55 more than under the IPART proposal. But we believe that overall our proposal is fairer and in general more accurately reflects the "market rent" intention ofthe Crown Lands Act.

If IP ART insists on the addition of an administration fee, WAG is of the view that it should be based on $100 per occupancy (not $200) plus GST, but rather recovered as a small percentage surcharge on the rent (calculated by IPART and adjusted annually by the NSW Govennnent's percentage increase in public sector wages), so tl,at someone paying a fair market rent of $20,000 per annum will pay a 40 times greater administration fee than someone paying a fair market rent of $500.


CONCLUSION ON THE RENTAL FORMULA AND RENT CALCULATIONS

The proposed forn1Ula is still seriously flawed with the huge variance in current rents from $1.17 to $50.04 per square metre not representing anything close to market rents, as was required by the 2003 Terms of Reference, intended by IPART in 2004, as required by the Crown Lands Act and as we think was intended by the Government in the current Terms of Reference. Some of the reasons for the proposed formula not calculating fair market rents are as follows: I. In many cases, under what IP ART is proposing, the owners of smaller properties

will be paying higher rents and be subsidising the owners ofiarger properties, who will be paying lower rents (e.g. Parramatta River).

2. There is also a serious mis-match between SLVs and RORs on a value and place basis, which will cause some serious over-charging to occur.

3. The use of A TO data for apartments and houses, as the basis of the differential between gross and net ROR is flawed. The differential is not that far wrong at the current level of2.5%, as set by IPART in 2004. If the apartments figures are removed from the data supplied by the ATO, WAG believes that the houses figures remaining per precinct will vary between about 2.5% and 3.1 %.

4. The discount multiplier as applied to SLVs needs to be reduced from the 50% level to 32% and be applied to the postcode based median SLY (to achieve proper matching).

5. A discount multiplier needs to be applied to the ROR, to allow for the fact that the ROR on land is lower than for land and house packages.

WAG recommends that the still seriously flawed formula is replaced with another, based on expert property valuation advice, to achieve fair market rents, and that the new formula is thoroughly tested before implementation. The proposed rent increases, averaging 14% for 1,620 DWF occupancy holders (see table 9.3 on page 112) and of up to194% in 9 current precincts and the Brisbane Water example, where the proposed rent is about 5 times what it should be, just because the ROR was not matched to the SL V, proves that while the old 2004 formula has a "flashy new coat of paint", underneath it is still built on a weak foundation, that will not withstand some serious prodding.

THE NEW FORMULA - AS RECOMMENDED BY WAG

STEP 1- SET THE MEGA PRECINCTS The current proposal to reduce 61 precincts to 20 appears acceptable, except for Sydney Harbour, where we believe that the 6 precincts should be reduced to 3.

STEP 2 - DETERMINE THE BASIS OF THE ROR We agree with !PART's proposed setting of the ROR, except for the differential between gross and net.

STEP 3 - MATCH SLVS TO RORs ON EXACTLY THE SAME BASIS

WATERFRONTACTlONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 12 ofl5

Fundamental valuation theory demands that RORs and SL V s are perfectly matched on a time, value and place basis. IP ART has correctly matched on a time basis, but not on a value nor on a place basis.

STEP 4 - DETERMINE THE DISCOUNT FACTORS In steps 1.2 & 3, we have determined the factors that would enable the calculation of a fair market rent on house and land packages within a precinct. But we are dealing with vacant land, on which the ROR is lower, so the ROR needs to be discounted accordingly.

And we are dealing with generally submerged land, held on uncertain tenure and which has severe restrictions placed on what can be done with it. This land is therefore less valuable than dry freehold tenure land, on which one can build a permanent dwelling. Therefore the median postcode SLY needs to be heavily discounted, according to evidence of how historically the Valuer General has discounted such wetland, compared to median postcode SLVs.

These 2 discount factors need to be included in the formula.

STEP 5 - APPLY THE AREA OF THE OCCUPANCY Steps I, 2, 3 & 4 will give us a precinct market rent per square metre. The [mal step is to multiply that by the occupancy area to arrive at the rent payable for a particular occupancy.

OTHER MATTERS

SECTION 6 - PENSIONER CONCESSIONS AND CASES OF HARDSHIP Pensioner Concessions - WAG would like to see the benefits offered to those on a full pension extended to those on a part pension. WAG is aware of people on a part pension struggling to keep their family homes, because of the huge wetland rents being charged by Maritime and CLD. They also cannot afford to demolish their waterfront structures and surrender their lease or licence, so they are faced with having to sell the family home.

Hardship Arrangements - In regards to Draft Recommendation 22, WAG would like to see the representative from "at least one independent organisation" on the review panel increased to two (to equal the two representatives drawn from the agencies) and for the "independent organisation" to be more clearly specified as an organisation with considerable financial skills (such as the NSW Audit Office), so that as with the ROR, the agencies do not suit themselves, because a recommendation was insufficiently precise.

SECTION 7 - CONCESSIONS FOR OTHER CIRCUMSTANCES

7.3 Long Jetties - WAG disagrees with the recommendation not to grant a concession to the owners ofiongjetties, which are long because they have to be long to reach navigable water. It seems unfair that in the same precinct one jetty owner W ATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 13 ofl5

might have to pay say 4 times the rent for the land under a long jetty, that was expensive to build and is costly to maintain, compared to the rent that another owner pays, both being for the same purpose - to reach navigable water. The additional rent paid by the owner of the long jetty provides no additional benefit.

7.6 Other proposed disconnts or concessions - WAG disagrees that some of its proposals for discounts" do not warrant special concessions or are already accounted for in the rental formula (via the PSLV or the discount multiplier)". For example-

• Someone with a jetty in shallow water and who can only use the jetty at high tide should not have to pay the same rent as someone else in the same precinct with a jetty in deep water and who can use their jetty 100% of the time. The "market rent for the land" under those two jetties will not be the same, but the PSLV and discount multiplier are the same, as is the rent calculated by the formula that uses those two factors ..

• Someone who must provide access to utilities on a portion of the rented land is unable to build a structure on that portion of the land, but must maintain it for the benefit of the utilities. The "market rent for the land" the subject of such impediment will be lower, but the rent calculated by the formula does not reco guise that.

8.1 NSW Maritime lease I right of appeal

a) NSW Maritime lease - WAG has some concerns with the rent clauses, including the proposed wording for clause 3.7 as follows:

i) The question of how the initial rent is to be set has not been addressed.

ii) The use of the word "may" within the words [in 3.7(a)] "The Lessor may vary the Rent Formula to implement a recommendation of the Independent Pricing and Regulatory Tribunal .... " does not require the Lessor to do anything.

iii) The words [in 3.7(b)] "The Lessor must vary the Rate of Retum each year.. ... " may provide no real effect on the rent, because the Lease does not specify how the Rate of Return is applied or how any change in the Rate of Return is to be applied. In particular, if the Lessor chooses not to vary the Rent Formula under 3.7(a), then changing the Rate of Return under 3.7(b) will be of no effect. The problem would appear to lie in the definition of "Rent Formula" within Schedule 1 of the Lease.

b) Right of Appeal- WAG is most concerned that IP ART is recommending that NSW Maritime occupancy holders are left without any rights of appeal to an independent tribunal or court. That would allow the CUiTent situation to continue, whereby CLD occupancy holders can appeal to the Local Land Board orland the Land and Environment Court (and then to higher courts), but NSW Maritime occupancy holders have no such opportunity. WAG believes that as well as providing a normal legal review process, having a right of appeal will improve the quality ofNSW Maritime decisions and processes.

WAG has numerous other serious issues with the NSW Maritime lease, but the responsible Minister (the Hon. Duncan Gay MLC) has promised that a "high level working group" will consider the terms of the Lease, following receipt of IP ART's recommendations, and WAG will raise those other issues at that time. W ATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 14 of 15

8.4 Rights of appeal- This section covers basically the same ground as the similar section within 8.1 and comes to the same conclusion that NSW Maritime lessees should have no right of appeal to an independent tribunal or court, while CLD licensees should continue to have such a right. WAG disagrees with that recommendation.

9.1 Capping of rent increases - WAG agrees with the $200 annual rental increase cap, but would prefer to see that extended from I year to 3 years (i.e. $200 per year for 3 years).

9.3 Impact on the agencies - WAG understands that there are situations where structures have been built on wetland without the knowledge of the agencies and that rent is not being paid on that wetland. Better policing will result in the agencies(particularly CLD) increasing its revenue base.

9.4.4 Impact on W AO occupants - WAG considers that the proposed 39% average percentage rental reduction (based on IP ART's assumptions) is far too small, considering that IP ART has recognised the validity of the argument that WAO occupants have a right to free primary access to their properties. .

9.4.5 Impact on pensioners - WAG notes that under the draft recommendations 41 pensioners will receive rental increases averaging $46 and believes that IP ART should recommend that annual increases for pensioners should be limited to the cpr.

ATTACHMENTS

I. Fact Sheet A - Rent Increases as a result ofIPART Draft Report Recommendations.

2. Fact Sheet B - The Basic Flaws With the IP ART Formula 3. Fact Sheet C - WAG Analysis of Samples to Determine Discount Multiplier 4. Fact Sheet D - Overcharging of Lower Valued Properties by ROR 5. Fact Sheet 6 - Matching ROR and SLV by Place 6. Criticisms of the LPI Consultancy Report 7. Fact Sheet 8 - Calculating the Discount Factor 8. Fact Sheet 10 - Correcting the Flawed ROR Calculation 9. Letter from Professor Alfred Eves, Professor in Property Economics, Queensland

University of Technology, commenting on the Pittwater study by Egan National Valuers

10 .. Endorsement from Paul Chaloner of Chaloner Valuations, commenting on this submission and the attachments as they relate to valuation issues.

11. Endorsement from Professor Alfred Eves, Professor in Property Economics, Queensland University of Technology, commenting on this submission and the attachments as they relate to valuation issues.

12. Legal opinion of Michael Chapman, specialist lawyer and maritime consultant, on the matters relating to the NSW Maritime lease.

l3. Legal Opinion of Brian Hones, partner of property lawyers Hones La Hood.


FACT SHEET A

RENT INCREASES (PER SQM) AS A RESULT OF IP ART DRAFT REPORT RECOMMENDATIONS

CURRENT PRECINCT CURRENT RENT PROPOSED RENT % INCREASE

Georges River 5 $6.16 $6.82 10.7% Great Lakes $1.19 $1.91 60% Hawkesbury River 1 $1.43 $2.33 63% Hawkesbury River 5 $1.54 $2.33 51.3% Murwillumbah $3.38 $3.69 9.2% Pittwater W AO $7.62 $9.23 21.1% Richmond $2.11 $3.69 74.9% Wagonga-WaUaga $0.57 $1.47 157.9% Womboyn $0.50 $1.47 194%

PSLV* INCREASES (PER SQM) AS A RESULT OF IPART DRAFT REPORT RECOMMENDATIONS

CURRENT PRECINCT CURRENT PSL V PROPOSED PSLV %INCREASE

Georges River 5 404 582 44% Great Lakes 78 287 267.9% Hawkesbury River 1 94 257 173.4% Hawkesbury River 5 101 257 154.4% Murwillumbah 221 384 73.7% Pittwater W AO 500 1,388 177.6% Richmond 138 384 178.2% Wagonga-WaUaga 37 251 578.3% Womboyn 33 251 660.6%

*PSLV is the Precinct Statutory Value, which is derived from valuations provided by the NSW Valuer General.

NOTES: 1. The above rent and PSL V figures are quoted from pages 152 to 154 ofthe IP ART draft report.

2. Table 9.3 on page 112 illustrates that 1,620 wetland rents (being 19% of the total) will increase by an average 14%, as a result ofthe draft recommendations.

3. Of those receiving further rent increases, 41 are pensioners, who will receive rent increases averaging $46.

FACTSHEETB

TWO BASIC FLAWS WITH THE IP ART FORMULA

1. WATERFRONT SLVS CANNOT BE DIVIDED BY THEIR AREA

On 9 February 2006, Mr George Citer (current Chairman of WAG) wrote to the NSW Valuer General (VG), complaining about the VG's then approximate 60% increase in Seaforth waterfront SL V s and the effect on NSW Maritime rents. In that letter Mr Citer set out the 2004 IP ART formula. The VG responded by sending his valuer

to visit Mr Citer (on 10 March 2006) and explain that the 2004 IPART formula was fundamentally flawed and produced "meaningless figures". explained that Mr Citer's tiny 164 sqm property was worth a similar amount to much larger nearby waterfront properties, so that dividing the SL V by the area, as the IP ART PSL V formula did, was a nonsense, producing nonsensical rents.

On 30 August 2006 the VG confirmed in writing certain statements made by on 10 March 2006.

On 1 June 2007 Mr Citer and then President ofthe Boat Owners Association, Michael Chapman had a meeting with the VG and the then Chief Valuer ofNSW. During that meeting they confirmed the earlier comments of their valuer . When confronted with a NSW Maritime aerial photo of Middle Harbour with superimposed per sqm wetland rental rates varying between $55.02 (in Seaforth) and $6.38 (in Castlecrag) and asked if it was their valuations that caused such inequities, they responded that it was the fault of the flawed formula and the huge variance in property sizes that caused such a huge variance in wetland rental rates.

Every property valuation expert consulted by WAG has confirmed the advice of the VG and Chief Valuer, that with waterfront property one cannot divide the value by the area, as much of the value is related to access to the water, views and amenity, rather than size.

The current proposal ofIP ART in their Draft Report still repeats this fundamental flaw of dividing a waterfront SL V by its area. The proposed per sqm rents on the Parramatta River of $45.64 (Balmain area) , $20.81 (Hunters Hill area) and $18.86 (west of Gladesville Bridge) are examples of how the formula discriminates in favour of precincts comprised of generally large properties and against precincts comprised of generally small properties. So IP AR T still has not addressed that fundamental flaw in their formula.

2. MISMATCH OF RENTAL RATE OF RETURN AND STATUTORY LAND VALUE (AS SET BY THE VALUER GENERAL)

Every property valuation expert consulted by WAG has expressed that it is fundamental valuation methodology that the rental rate ofretum (ROR) must be matched to the statutory land value (SL V) on a location and value basis. They have advised that generally rental rates of return are higher for low value rental properties and lower for high value rental properties.

Postcode based RORs for postcode areas that include waterfront properties generally have a large number of non-waterfront lower value rental properties and a small number of waterfront high value rental properties. The median used to determine the ROR will therefore not be a waterfront high value rental property. It is fundamentally flawed to apply the median ROR for a postcode or group of postcodes to a waterfront property SL V (as IP ART has done), because that is a fundamental mismatch. What is fundamentally correct is to apply the median ROR for a postcode or group of post codes to the median SL V for that post code or group of postcodes. It is basic property valuation theory that one must "compare apples with apples", not "apples with apple seeds".

In its approach to setting a fair market rent, WAG has resolved both these issues. But while IP ART has tinkered with some fringe issues and by doing so made some worthwhile improvements, it has not addressed the fundamental problems with its 2004 formula. One property valuation expert commented "The appearance has been improved with a fresh coat of paint, but the structure is still unstable and will eventually collapse".

Fact Sheet C WAG Analysis of Samples to Determine Discount Multiplier

In putting together the original set of 144 separately valued occupancies WAG had data from both LPI and NSW Maritime.

In the case of LPI the data was extracted from the NSW Maritime precincts in Sydney Harbour, Middle Harbour and Parramatta River, as well as the Crown Lands Division's precincts in Pittwater and Georges River. The data was found by LPI by matching the address on separate valuation entries so that only properties that had 2 valuations at the same address were presented. They were then checked to ensure that one was the freehold and the other was an occupancy at that address, or a shared occupancy with the neighbour. The problem with this method was that it did not disclose separately valued occupancies on reserve land.

The data we obtained from NSW Maritime was only identified by postcode; there were no address details. We therefore had to rely on the order of presentation of the data in the spreadsheet. Most of the properties were a 'valuation' entry followed by a 'premise' entry and that reflects the situation with the majority of properties where the occupancy is valued with the freehold. The entries WAG could use were the valuation sets that contained a 'valuation/premises' entry either by itself or immediately after a 'valuation' entry. The single entry was almost certainly an occupancy on reserve land and as the LPI data search did not extract such occupancies most of this data was useable. There appeared to be a number of pairs of data (Freehold with separately valued occupancy) that had not been found in the LPI search.

The issue we found with the 'valuation' and 'valuation/premises' pair was that it was likely to be in the LPI data and a comparison had to be made to eliminate the duplicate. As we did not have address we could only make this comparison on the basis of freehold area and of occupancy area. However, there is a clear mismatch between the NSW Maritime and LPI databases and the stated areas can be quite different - they are rarely the same.

As a result of the limited supply of information WAG included a number of data sets from NSW Maritime that should not have been included, but also excluded data that should have been included. There were some instances where an occupancy valuation was included as a pair with a freehold, but instead it should have been included as a stand alone occupancy valuation.

NSW Maritime subsequently supplied address information (on a confidential basis) for the few properties where we had had difficulty in deciding if and how the entry should be included. As a result of this extra supply of data WAG has been able to isolate 145 separately valued occupancies. A summary is:

• 9 duplicates removed - they had been included as there were sizable differences between the LPI and Maritime areas in 7 cases and smaller differences in 2 other cases and we were unable to match

• 1 shared occupancy which was included twice was deleted as it was listed against both neighbours. LPI had it paired with 1 neighbour and NSW Maritime paired it with the other neighbour

• 5 occupancies were matched to the wrong freehold due to the lack of street detail, most were occupancies on reserve land

• 2 additional freehold/occupancy pairs were found • 2 occupancies that looked like they were on reserves turned out to be freehold/occupancies pairs

(included above) • 15 additional separately valued occupancies without matching freeholds were found however 2

were from previously mismatched freehold/occupancy pairs

Page 1

Fact Sheet C WAG Analysis of Samples to Determine Discount Multiplier

LPI criticised WAG for using both 2007-2009 3 year average data, as well as 2010 data in the one document. While the % was calculated by comparing like with like ie 2007-2009 to 2007-2009 and 2010 to 2010, to avoid any further unjustified criticism we were able to obtain 2010 data from NSW Maritime so that all values were based on 2010.

Given that !PART has calculated PSLVs based on 2010 median data, we have now included that in the analysis and a discount factor has been calculated on this basis.

The postcode SLV/m2 comparison is also included and this data is also 2010. SLY and area was sourced from LPI.

Initially we had calculated the discount factor as a percentage of the adjoining freehold, where that was available, and where it wasn't available we calculated it against an adjusted PSLV that did not include the occupancy. That was on the basis that we wanted to see what percentage the occupancy was of the freehold and not of the freehold and occupancy. This calculation has been changed to be the same as the PSLV ie occupancy and freehold. The reason is that the occupancies on reserve land had to be calculated against the median PSL V and as that contained both freehold and occupancy, and we did not have the information to recalculate the PSL V on a median basis, we had to change the occupancy to freehold calculation for consistency.

Based on this greatly improved sample and using the median (which ignores unusual valuations whether they are too high or too low), the discount multipliers are 13.69% on a direct match to freehold & occupancy and 31.99% on a postcode based median SLV/m2 .

. NOTE: WAG wishes to recognise the assistance ofNSW Maritime in providing data quickly and without charge. LPI also provided data in a speedy fashion, but WAG had to pay for that data.

Page 2

Fact Sheet D Overcharging of Lower Valued Properties More by ROR

Background WAG has produced' graphs, for many years now, that show that as the value of a property increases the Rate of Return (ROR) % decreases. These graphs show that this reduction in ROR is rapid at lower house/land values but then levels off and eventually asymptotes around a 2% ROR. See graph reproduced from Fact Sheet 10:

2010 Postcocle RCR - All Sales Values ($,000)

8.0:)'%

7.OCP/o

• 6.W/Q

5.CXJP/o

4.00'/0

3.(0'/0 • • • • •

2.(0%

1.0)'%

O.W lo

'Ill $500 $1,cx::xJ $1 ,500 $2,000 $3,= $4,000

!PART has extracted information for 9 years of data and graphed 3 year averages for 2001-2003 , 2004-2006 and 2007-2009. In all cases they support the WAG proposition that the ROR drops as the property value increases.

WAG has provided further evidence that much of this reduction in ROR comes about by the mix of the value of the house and the land in the overall value with the house having a higher ROR and the land having a lower ROR. Generally a houselland package with a very low total value will have a much higher percentage of the total being attributed to the house and a lower percentage to the land, whereas a house on a very expensive block of land will represent a lower percentage of the total.

!PART has dismissed this proposition, but has failed to give any facts , by simply stating that the house and land can't be separated and then goes on to use the ROR of the house and land as a surrogate for the ROR for the land.

Page 1

Fact SheetD Overcharging of Lower Valued Properties More by ROR

The Problem with the IP ART Draft Proposal While IP ART has recognised that there is a difference in the ROR at various price levels, it has proposed a system that ignores the very evidence that they have included in their own report (pages 143-145).

In all cases the application of a postcode based ROR to a Waterfront SLY will see a higher ROR than can be supported ie it is not market. However, it is the level of difference at various values that is of most concern.

What we have is the postcode median based ROR being used from a lower value on the graph, when the actual waterfront property value is higher value and in a lower ROR position on the graph. The big issue is that as both values move down e.g. regional NSW, you end up with the postcode based value being on the very high ROR section, but the waterfront value being in a still relatively low ROR section.

So IP ART's draft decision is to overcharge everyone, but overcharge the owners of the lowest valued properties by much more.

An example may assist with an understanding of this issue. Consider 2 actual properties on the Central Coast located about 4 kilometres apart, one a waterfront and the other a non waterfront: (This example was recently provided to WAG by a Central Coast real estate agent.)

• Non Waterfront o Annual rent $27,114 o Last sold December 2010 for $660,000 o I July 2009 SLY $288,000 o 4 bed, DLUG, 2 living areas o Implied house value $372,000 (Sales value less SL V) o Calculated ROR 4.11 %

• Waterfront o Annual rent $27,114 o Estimated value $900,000 (by local Real Estate Agent) o 1 July 2009 SL V $840,000 o 3 bed older style with jetty o Implied house value $60,000 (Sales value less SLY) o Calculated ROR 3.01 %

So one is a better house and the other has a better location (i.e. more expensive land). but the same rental. In essence an extra $552,000 in land value is an exact trade-offfor an extra $312,000 in house value. This suggests that return on house value is abnost twice the return on land value. See graph below from fact Sheet 10 on this issue:

Page 2

.. ~ 0 ~

Fact Sheet D Overcharging of Lower Valued Properties More by ROR

10.00"10

9.00%

8.00%

7.(10%

6.00"10

5.00%

4.00%

3.00"10

2.00"10

1.00%

0.00%

0.0% 10.0%

Rates of Return on Land Content with House and Land Packages

•

20.0%

• •

30.0%

(based on 2010 individual entries)

40.0%' SO.O"lo 60.0"10

Land ConlenW.

Trend line: ROR% = 6.3% - 3.92% x Land Conlenl% e9 50% land content i ROR% of:

70.0"10 80.0% 90.0% 100.0%

One can view this from another perspective. [PART is proposing to apply the median ROR of the postcode (essentially the non waterfront ROR of 4.11 % above) to the waterfront property value ($900,000). So !PART is ignoring reality and saying that the rent on the above waterfront property should be 4.11 % of $900,000 = $36,990, instead of (the actual rent of) $27,114. In this context it appears ridiculous, but that is exactly what !PART is proposing. And because of the way the ROR falls on the graph i.e.quickly at first and then more slowly, this has a much larger impact on the lower valued properties, which one find in regional areas.

The Resolution Trap On reading this !PART may see the error of its ways and assume it can be fixed by simply using the value per square metre of the postcode median SLY, to calculate the rent, as recommended by WAG. That is not sufficient.

The median postcode based property value is required to obtain a perfect ma tch between a median postcode based ROR and the value/location of the property to which it is applied, so that rents are market based.

But that does not allow for the fact that the ROR is based on a house and land package, rather than on land only. What is needed is the further adjustment of the ROR, as originally recommended by WAG to acknowledge the different ROR for house and for land.

In some areas (such as Sydney's Eastern Suburbs), where the values are high both on the waterfront and away from the waterfront, the median postcode ROR will be low and therefore relatively close to the ROR for land. In those cases , there will not be a large adjustment required to bring the ROR back to the ROR

Page 3

Fact SheetD Overcharging of Lower Valued Properties More by ROR

for land. However, in lower valued areas, including regional areas, the median postcode ROR is much higher and so a much larger adjustment is required to bring the ROR back to the ROR for land.

IPART dismissed the WAG advice and research that showed a difference between ROR for Land and ROR for the house but unless it recognises that issue it will be making the conscious decision to overcharge regional NSW by a substantial amount and others by a smaller amount.

Page 4

Fact Sheet 6 Matching ROR and SL V by Place

To achieve a market rent the rate of return (ROR) and SL V must match for time, place and value. Time has been dealt with in another factsheet but place and value have not been dealt with.

The remaining 2 issues are closely correlated eg Waterfronts are higher value than non waterfronts, Eastern Sydney is higher value than Western Sydney etc.

The easiest way to look at the issue is to graph the median sales value of each Sydney Postcode (as defined by the IP ART data supplier) against that postcode's ROR as follows:

S.W /o

4.OC)%

3.(0%

2.o:::P/o

1.00010

0.00'/.

ro

2008 Postcode ROR - All Sales Values ($,000) Postcodes: 2000 - 2249,2552 - 2570 &2740 - 2771

:.l.

~ •

• • ~

~ • • ! :~ • • .. ~. • ~ •• • • • •

•

$1 ,000 $1 ,:00 $2.000 $2,:00 $3,000

•

$3,:00 $4,000

The downward sloping trend line is a standard Excel Spreadsheet function and represents the line of best fit to the data points. The following table shows the ROR at various sales values, from the above graph, as well as the movement from the ROR at the Sydney median sales value.

Sales Value $485,000 $\,000,000 $1,500,000 $2,000,000 $2,500,000 Sydney

Median ** ROR 4.32% 3,56% 3,13% 2.83% 2.59% ROR reduction from 0.00% (0 ,76%) (1.19%) (l.49%) (1.72%) Sydney Median

** note that the trend line fitted to postcode medIans WIll not exactly equal the Sydney medIan ROR due to the different numbers of properties within each postcode

The reduction in ROR as property value increases is a known fact as is well documented in the opinion of Professor Eves dated 17 March 2009 in his peer review of the Egan Review of Maritime precincts, Copy of Professor Eves ' report attached.

Page 1

Fact Sheet 6 Matching ROR and SLY by Place

Egan National Valuers completed a review ofPittwater Macro Precincts, a LPMA area, in early September 2009. That review contains a number of relevant statements as follows:

• "Both SL Vs and rates of return are area and time specific, and both must be treated in an identical manner, to enable one to be applied to the other for any meaningful outcome."page 11 fourth paragraph section 6

• "I further believe that the rates of return should reflect residential house properties within the waterfront precincts in which the leased Crown land is contained, as occurs with the Statutory Land Values ..... I also believe waterfront properties being the more expensive would show an even lower return than those noted above." - page 11 third paragraph section 6

The NSW Auditor General (NSW AG) had the same opinion:

• "Advice from property valuation experts provided to the Audit Office was that both the net rate of return and the SL V are area and time specific. The relationship between SL V s and the net rates of return is indicated by their movement in opposite directionsgenerally as property values go up and the rate of return goes down. Therefore, it is preferable that both be reviewed at the same time and for the same area." - page 47 paragraph 1

• " .... The review should include assessment of the net rate of return from a location and time specific perspective, similar to the approach taken for assessing the statutory land value of precincts." - page 48 paragraph 3

Two experts and the NSWAG agree that the ROR and SL V must be for the same area and not a one size fits all. This is also obvious from the graph above.

The remaining issue is that it is unlikely that there will be sufficient data to complete the calculations for waterfronts only. The next best alternative is to use the ROR for houses for the postcode(s) in each precinct with all the SLVs in the same postcode(s).

The above would remove the final mismatch and would provide a system that is administratively efficient and easily updated and would provide the best surrogate for market rent available. Other alternatives would be to calculate ROR and SLY for groups of macro precincts or for waterways.

Page 2

W ATERFRONTACTIONGROUP A SUBGROUP OF TH;E BOAT OWNERS ASSOCIATION OF NSW INC


13 November 2011

SOME CRITICISMS OF THE LPI CONSULTANCY REPORT

[Including comments in square brackets relating to a full day meeting between WAG and LPI on 9 November)

NON-INVOLVEMENT OF NSW VALUER GENERAL

1. During the course of its investigations, WAG consulted numerous property valuation experts and several had stated that they had a high opinion of the valuation capabilities of the Valuer General, Mr Philip Western. At a meeting on 1 June 2007 with the current WAG Chairman and the then BOA President, Mr Western indicated that he was prepared to provide advice to IP ART, if requested. WAG was recently assured by IPART that the Valuer General was being consulted and the Roads and Ports Minister recently provided a similar assurance. We note that the LPI consultancy report has not been signed by Mr Western and we therefore do not have the same confidence in its fundamental correctness, as if he had signed it. [LPI have confirmed to WAG that Mr Western was not at all involved in their report for IP ART. LPI advised WAG that they agreed with LPI valuer the Valuer General and the previous Chief Valuer that it is fundamentally incorrect to divide the SLVs of waterfront properties by their area (as per IP ART's PSLV formula) , because the result is a "meaningless figure".)

COMMENT ON ANALYSIS BY LPI OF WAG SUBMISSION

2. In regards to the analysis by LPI of our work on the discount multipliers, we note the following;

a) There is no discussion about the fact that the rate of return relates to land and house packages and needs to be discounted back to an appropriate ROR for unimproved land. The Crown Lands Act at S. 143 (1) (a) states "the rent shall be the market rent for the land", not "land and house packages". [LPI conceded to WAG that the ROR on land might be different to the ROR on land and house packages, but stated that they were not asked by IP ART to consider that matter.]

b) Annexure 12 of the LPI report only covers Sydney Harbour, whereas our analysis also covered Pittwater and Georges River. [LPI agreed that our sample was for a broad sample of 3 waterways and could not answer why their map only dealt with Sydney Harbour samples.]

c) Annexure 8 of the LPI report notes that "22 water reserve properties listed in the WAG analysis were excluded, as they could not be identified from the information supplied". We provided all the source data, which came from LPI and NSW Maritime. We presume that the 22 properties which could not be identified were in the data supplied by NSW Maritime and a simple telephone call to NSW Maritime should have resolved the problem. We have now received from NSW Maritime (within 24 hours of our request) address details of those 22 properties. Why did not LPI make a phone call to NSW Maritime to resolve this problem? [LPI advised WAG that they decided to delete any properties that were not directly adjoining a freehold. That decision by LPI

Page 1 of 14

caused the discount multiplier to be higher than it should be, because the deleted samples had low discount multipliers.]

d) The complaint that our "valuation data .... contains significant inconsistencies, including duplication and inconsistent SL V base data" is curious when the data was sourced from LPI and NSW Maritime (which sources its data from LPI). We did check for duplication between the LPI and NSW Maritime sourced data and removed any where we could find identical or very close area sizes in specified postcodes. But the NSW Maritime data did not include street addresses. We have now received from NSW Maritime (within 24 hours of our request) that additional information (on a confidential basis) and which has enabled a more perfect cross-checking with the LPI data, which did include street addresses. We have now removed from our sample a number of duplications, but added in some others that we have found, with the help of NSW Maritime, so that we now have 145 samples in 3 waterways.

e) The comment "general over-representation of higher valued properties and use of a number of relatively low value occupancies in the WAG analysis resulted in higher discount factors" is strange when we used all the data supplied to us by LPI and NSW Maritime, and only removed the top and bottom quartiles, so as to ignore unusually high and unusually low observations that might distort the result. Taking the middle two quartiles provided a sound basis to our conclusions. Alternatively, WAG is quite prepared to leave in all the samples and use the median (which effectively ignores the extreme results at both ends), the result of which is now a discount of 86.31 % and a discount multiplier of 13.69%, as applied to waterfi'ont PSLVs, or a discount of 68.01 % and a discount multiplier of 31 ,99%, as applied to a postcode median, based on our latest exercise of 145 samples. Had we not removed any of the observations from both ends and used all 144 samples in our earlier exercise, and used a median instead, the result would then have been a discount of 87% and a discount multiplier of 13% related to the waterfront PSLV, or a discount of 66% and a discount multiplier of 34% related to the postcode median .. But it was obvious to us that a number of valuations were either too low or too high, so to be fair we removed an equal number at each end and worked with the remainder. [LPI agreed that WAG had, with its assistance, considered a broad range of samples in 3 waterways. LPI agreed that there were some values that were too high and some too low and agreed with WAG's approach to remove the top and bottom quartiles and work with the 2 middle quartiles. There was some discussion about an 80/20 rule, which would mean that only the top and bottom 10% are removed and which WAG is prepared to accept, as a compromise]

f) The complaint that we used a "geographically concentrated sample" seems unfair, when we had to pay for the data provided by LPI and when Sydney Harbour, Pittwater and Georges River, from where our 144 examples were drawn, provides most of the rental revenue base for the Government. We note that for its benchmark properties, LPI used only 3 Port Macquarie properties for the North Coast and only3 Batemans Bay properties for the South Coast, while for Brisbane Water they used 3 from Brisbane Water East, but none from Brisbane Water West and on the Georges River they included no samples from Precincts 2 and 4. We also note that Annexure 6 (relating to sale of small land parcels to adjoining owners) was restricted to the Woollahra Council area. LPI


with all its resources is using a very small sample of 52 (which is considerably smaller than our sample) to calculate the discount factor or mUltiplier. [LPI stated that is was limited to using data where there was good commercial marina rental evidence available and which they discounted by 20% to 30%] WAG strongly disagrees with this approach ..

g) We note that LPI is revaluing 37 of the 86 properties that they did not remove from our 144 samples, on the basis that they were wrongly valued. That represents a 43% error rate, which is very concerning. In regards to Sydney Harbour Precinct 1, they reduced our sample of 45 occupancies to 38 and then re-valued 27 ofthose (71 %), which is even more concerning. In June 2007 the Valuer General and Chief Valuer admitted that there was a level of inaccuracies in SL V s, but stated that it was in the region of 5% and that they were working to reduce that rate. WAG wonders whether in changing a high proportion of the valuations, the subject of the WAG analysis, LPI has "moved the goal posts mid-game". WAG is concerned that our objective data is being replaced by subjective data. [LPI stated that generally the objection rate had substantially improved to now be closer to 1 %. They agreed that in this instance the "error" rate was unacceptable.]

h) In their work on our 144 samples, LPI removed a significant number of occupancies where the discount was high, and then increased the valuations on a large proportion of the remainder. But there was a significant number of valuations in our sample which were obviously valued too highly. There were 3 where the wetland was valued on a square metre basis at more than 100% of the adjoining freehold (105.35%, 109.42%, 186.44%). LPI did not have those re-valued down, which one would have expected. And while we took out the "strange" valuations at each end, before taking an average of the middle two quartiles, it appears to us that LPI "adjusted" only the bottom two quartiles and left in all the "strange" figures in the top quartile, when taking an average. We have been asked "Was LPI trying to achieve a particular outcome, or were they simply sloppy?". [LPI stated that they will revalue the SLVs that were obviously too high, but their contractors had not yet got around to doing those]. Having only got around to revaluing the low ones obviously skews the results.

i) There is one further problem for LPI. WAG has received advice from the Valuer General and its own property valuation consultants that the recent revaluations are subject to the usual objection process. Property owners have 60 days to object and if they do, it may take many months until those objections are resolved. Some occupancy owners may even choose to challenge the revaluations in Court, in which case it could be upwards of 12 months until the re-valuations are final. But LPI has used the re-valuations in their exercise for IPART, as if they were final. WAG is of the view that LPI should have used the only final values currently available - the original values, as used by WAG. [LPI agreed that the new valuations were "fluid" and subject to change and should not be used in any valuation exercise.]

j) WAG notes that 26 occupancies were deleted by LPI from our sample, because they did not adjoin freehold properties and we compared them to the freehold component of the PSLV (which was the next best thing available to us). We note that LPI has compared its 52 benchmark samples to the PSLV (as we did with only 26 out of 144 samples). On the same basis that LPI deleted 26 of our

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 3 of!4

samples, they should delete all 52 oftheirs (i.e. wetland value comparison with non adjoining dry land value). [LPI and WAG failed to agree on this point.]

k) WAG and its property valuation experts are amazed at the statement by LPI "The comparison of occupancy land values to directly adjoining land values was considered inappropriate ...... LPI chose comparison to the existing PSL V s as more appropriate". There is nothing more appropriate fi'om a valuation perspective than the land that directly adjoins another parcel of land. LPI has rather chosen to compare the value of a parcel of wetland to the average value of (in most cases) hundreds of parcels of waterfront freeholds and adjoining wetland over a wide area. WAG and its property valuation expelis do not consider that LPI has adopted sound property valuation theory and methodology in this approach. [One of the issues raised by LPI is varying freehold sizes distorting the figures and WAG has some sympathy with that view.] But using a waterfront PSLV does not solve that problem (e.g. Sydney Harbour precincts 3 and 4 have hugely different average freehold sizes). This issue was raised by LPI valuer, , the Valuer General and the Chief Valuer in 2006 and 2007, when they stated that the PSLV formula is fundamentally flawed and produces "meaningless figures".

I) The statement on page 14"LPI considered that the use of median Land Values from postcodes did not provide the appropriate basis for determining the value of Waterfront leases" completely overlooks the fact that the aim of the exercise is not to value waterfi'ont leases, but to determine a market rent applicable to those leases. Valuing waterfront leases might be appropriate, if a waterfront rate of return (ROR) was calculated to apply to such valuations. But since the ROR is calculated using median postcode SLVs and median postcode rents, the only correct application of that ROR is to the same median postcode SLY. That is basic valuation theory and WAG is surprised that LPI has not recognised that and advised IP ART accordingly. [LPI stated that this was outside the instructions set by IP ART, but agreed that the ROR on waterfront properties within a postcode could be different to the ROR of the median property within the same postcode. LPI agreed that there was a mismatch involved in applying a postcode based ROR to a waterfront PSLV, but stated that IPART had not asked LPI to consider that matter.]

m) We agree with the LPI statement on page 14 "The properties within a postcode vary considerably and may have little in common with waterfront property". One of the things that varies considerably within a postcode is the ROR and the median postcode ROR will be very different to the waterfront ROR. Therefore, as property valuers, LPI should have pointed that out to IP ART. [LPI agreed that there was a need to "compare like with like" and suggested that WAG raises this matter with IP ART, as it was outside their instructions, as set by IPART.]

n) The statement on page 15 "If a postcode basis were to be used, LPI considers the average provides a better representation of the market movement, inclusive of waterfront properties, than the median" overlooks the fact that the ROR has been calculated using postcode medians and having one factor based on medians and the other based on averages is fundamentally incorrect. The median is generally used in the property market, as averages are more affected by a few very high or low observations. The median (being the one in the middle) ignores observations that are out of character with the bulk of

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 4 of!4

observations in a market. It appears to us that LPI have re-written property valuation theory just for this report. [LPI agreed that there was a difference between averages and medians and expressed a preference for averages. But they acknowledged that one could not apply a median based ROR to an average based SLY. When advised by WAG that IPART had decided to use a median based PSL V, LPI stated that their figures would need to be changed from averages to medians, to make them relevant. ]

0) On page 19 LPI states that we sourced data from them, NSW Maritime and Crown Lands. This is not accurate. Most ofthe data was supplied by LPI and a smaller number was sourced from NSW Maritime. None was sourced from Crown Lands.

p) On page 19 LPI states that it compared the valuations of the separately valued wetland to the PSL V (which, in the case of the median, might be many kilometres away), rather than the adjoining freehold (as WAG did). That is not sound from a property valuation perspective

q) On page 19 LPI states that there was some duplication of properties supplied by us. The data supplied by NSW Maritime was supplied on a postcode basis and did not include street addresses. We cross-checked with the data supplied by LPI and eliminated any within the same post code that had the same or similar area size. If there remained some duplication, then there is a discrepancy between the area sizes recorded by NSW Maritime and LPI. [LPI recognised that there were problems with the area sizes in their data base, compared to the NSW Maritime data base.]

r) On page 19 LPI criticises us for using 3 year averages in some cases (as opposed to 2010 values supplied by LPI). These would have been supplied by NSW Maritime in that form, and we had no means of unravelling a 3 year average. But it is the relationship of the SL V of the wetland to the SL V of the adjoining freehold that we are measuring and a 3 year average of both is actually better than a single year of both, because it is based on 3 times as many observations (i.e. enlarges the sample) and smooths any unusual valuations. NSW Maritime has now supplied us with 2010 values for those properties, so in our latest exercise we have been able to use all 2010 values.

s) On page 19 LPI claims that it compared the wetland land values to the average of the postcode SLVs, rather than the median. That is not sound from a property valuation perspective, as the average can be affected by a few very unusual observations at either end of the scale. [LPI agreed that there was a need for consistency and that it would need to rework its figures based on medians.]

t) On page 20 LPI complains that where a wetland occupancy did not adjoin a freehold, we compared it to an "average value", that they were unable to verifY. In fact we compared it to the freehold component of the PSL V (which is close to what they chose to do with all their 52 samples). We were only a telephone call or email away, for clarification purposes, but LPI rather chose to delete a large portion of our samples, which had lower than average values and which supported a higher a discount. [LPI and WAG disagreed on the need to include samples of occupancies adjoining public reserves, but LPI were quite interested in the concept that WAG had removed the occupancies from the PSL V before doing a comparison. They acknowledged that in not removing the occupancies

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 5 ofl4

from the PSLV before doing the comparison, they had partly compared the occupancy to the occupancy.]

u) On page 20 LPI seems to be suggesting that we selected a large number of properties in the "Woollahra and Hunters Hill Local Government Areas". We used all the data supplied to us by LPI and NSW Maritime, which covered Sydney Harbour, Pittwater and Georges River. We paid LPI about $2,500 for their time in searching their records for these samples and we felt that it would not be cost effective to spend much more time and money searching for samples in other waterways, when we were ofthe view (based on a LPI source) that we had captured about 75% of the total number in the state and that those were from the waterways where the majority of the rent revenue came. After we provided KPMG with some examples of separately valued occupancies in late 2009, they included a reference in their February 2010 report that this matter be further investigated. Nothing was done, until we did it this year and then with limited time and resources. The Government agencies should have looked at 100% of the available examples in 2010 and it is WAG's view that LPI should have done it now, as part of their research for their report for IPART.

v) On page 20 LPI states " .. .the Land value of a significant number of the occupancies used in the WAG analysis was too low, which produced a much higher discount factor than should have been the case" . We used all the samples provided to us and did not selectively use some and not others. We noticed some extreme examples at both ends and for that reason eliminated the top and bottom quartiles and calculated an average from the middle two quartiles. We wrote to the Valuer General and reported that we had noticed some extreme examples at both ends, with some obviously valued too low and others obviously valued too high (i.e. the wetland was valued at more than the adjoining fi:eehold). As stated earlier, we are prepared to work with the median, which leaves in all of the data, whether it is too high, to low, or about right. The result of that is that the Discount Multiplier moves from 16% to 13% if the median waterfront SL V is used. [LPI agreed that some valuations in our sample appeared to be too high, but stated that their contractors had not yet got around to revaluing those.]

w) [LPI stated that they had received one batch of reascertainments from one contractor prior to 20 September and that they used those figures for their report to IP ART. They stated that the reascertainment pro gram is still ongoing.] WAG is critical of the use of a small batch of "fluid" reascertainments including increases of up to 1,995%, as that seriously skews the results.

x) WAG has received details of 73 reascertainments and has subjected them to careful analysis. We note that there were only 4 reductions against 69 increases of up to 1,995%. The Valuer General has confirmed that these are subject to the normal objection process, which means that the new values are subject to challenge and that they should not have been used by LPI in the manner in which they have been used. Our analysis has thrown up the following interesting results on a dollar per square metre basis:

FAIRLIGHT - 2 neighbouring boatsheds - $200 & $500

MOSMAN -2 - $32 & 631

CASTLECRAG -3- $304 TO $1,292 WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 6ofl4

HUNTERS HILL - Vernon Street -2 $624 & $1,000

GLADESVILLE TO NORTH SYDNEY - Many - $148 to $1,666

DARLING POINT - Yarranabee Road - 2 - $1,480 & $11,000

DOUBLE BAY - I square metre - $10,000

POINT PIPER - Wunulla Road - 8 - $1,794 to $l3,333

[LPI agreed that they might need to look at some of the reascertainments. WAG agreed not to approach individual property owners about them, until they had been fmalised and the property owners notified.]

COMMENT ON LPI'S APPROACH

3. LPI has focused on determining a discount factor for wetland when compared to waterfront properties, which is fme if that can be used in a formula that includes a ROR relevant to waterfront properties. But since there is no such ROR available, one has to work with what is available (ROR for the median postcode property) and match all other factors accordingly. [LPI agreed on the need for a perfect match, but stated that this was outside their instructions from IPART.]

4. WAG has a philosophical problem with accepting LPI's argument that owners of reclaimed land, who paid to improve the wetland (or paid a previous owner for that improvement) should be penalised with a lower discount, which then would lead to a higher rent. Reclaimed wetland is valued higher than otherwise unimproved wetland and that is aireadyrefiected in SLVs and PSLVs. Applying a lower discount, as recommended by LPI would seem to be "double dipping". We believe that LPI are not fully taking into account the facts that reclaimed land is subject to uncertain tenure, strict restrictions on its use, in many cases is open to the public and the owner is responsible for the cost of returning it to its original condition, when the 'lease or licence is terminated. It also appears from LPI's comments that the records are not sufficiently reliable to be able to separate reclaimed wetland from other wetland [LPI stated that they had valued reclaimed land on the basis of 40% less than market for freehold land in Sydney Harbour and 50% elsewhere. They agreed that they had not taken into account the harsh and restrictive terms ofthe lease or licence. LPI advised WAG that they were unaware that NSW Maritime had stopped issuing 20 year leases and was now only issuing 3 year leases with a month to month hold-over provision, compared to the previous year to year hold-over provision (which represents a significant loss of tenure). They agreed that they had not taken that matter into account. They seemed to understand the argument about "double dipping", but did not express a view.]

5. We note the LPI comment (bottom of page 16) that "The rental charged for the swing mooring usually includes tender access to the stored vessel as well as access to any marina facilities with the operator responsible for maintenance". Similarly (on page 17) it is acknowledged that marinas offer additional facilities and services such as " .. . slipways, dry docks, repair and refuelling facilities. Power, water, sewage pump-out, car parking and security ... toilet and showers, Wi-Fi internet and are located adjacent to ancillary tourist facilities". At the bottom of page 17 it is acknowledged that a private mooring pen was advertised at 33% of the berthing rate at a nearby marina. That would seem to indicate that a very high value is placed on the additional facilities and services and a very low value placed on a mooring space. There are also different demand/supply

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 7ofl4

dynamics for commercial marina berths, because of the additional facilities and services offered. It is therefore clear that no useful comparison can be drawn from what commercial marinas charge and LPI recognises that in their conclusion on page 18, but not in their Finding No. 3 which states in part "The value of wetland .... is more related to marine rentals in a geographic location". [LPI stated that they valued wetland based on what commercial marinas charged for mooring a boat (irrespective of size), and they valued reclaimed land based on adjoining freehold values. They made an interesting comment that they regarded jetty size as being iITelevant and that occupancy-holders should be charged the same rent for a 20 sqm or a 100 sqm jetty. LPI agreed with WAG that owners of long jetties should not be penalised with higher rents, because they needed long jetties to reach navigable water.]

6. On pages 22 & 23 there is discussion of two legal cases relating to small parcels ofland and where the courts found that a 50% discount to market value would be appropriate. In Annexure 6 of the LPI report there are examples provided of discounts relating to sales of small parcels of dry land to adjoining property owners in the Woollaln'a Council area, at an average discount of 56% to the SLY. In these cases the discount relates only to the parcel ofland being very small and being of little value to anyone but an adjoining property owner. There is no discount relating to the lesser value attributed to the land being submerged or partly submerged, the inability to build a dwelling or permanent structure on it, the ability of the lease to be terminated at short notice, the restrictive nature and harsh conditions of the occupancy instrument, the lack of exclusivity of access etc. IP ART attributed a 50% discount to those factors (without providing any evidence for that figure), so that when one takes account of the 2 discounts suggested, it seems that a small parcel of wetland has a very low value of less than 25% of the adjoining freehold land .. [LPI recognised that they had not taken into account the disadvantages imposed by virtue of the lease or licence. There was disagreement between LPI and WAGs valuation expert on whether in the case of reclaimed land, one should value it on the basis of the adjoining freehold land less an appropriate discount (as they did), or on the basis of the additional value brought about by the addition of the reclaimed land and then less the appropriate discount.]

7. We agree with the comments in paragraph 3 on page 14 beginning with "LPI considered that the use of median Land Values ... ", but feel that LPI has "lost the plot". The aim of the exercise is to determine a market rent and for that purpose one has to match SLVs to RORs on a value, location and time basis. The ROR is median postcode based, so SL V s must be the same, or the result is nonsense. There are similar comments on page 15 that are similarly not soundly based. [LPI agreed with the principle ofperfect matching, but stated that IP ART had not asked LPI to consider or comment on that matter.]

8. The comment on page 15 about use of an average rather than a median goes against established practice. The average is influenced by unusual observations, whereas the median is not. [LPI and WAG did not agree on this issue.]

9. We note that in Section Three (Review of Rentals and other Market Data) LPI did not consider the following relevant data:

a) The rents paid by commercial marinas for undeveloped wetland and the SLVs of that wetland, compared to the SLVs of adjoining dry land.


b) Valuations by independent valuers employed by CLD and NSW Maritime to assist in determining the rents charged to commercial marinas.

c) Rents and SLVs applicable to undeveloped wetland beneath oyster farms and how those SLVs compare to SLVs of nearby dry land.

We note that LPI rather considered less relevant matters, such as:

i) Developed swing moorings in commercial marinas.

ii) Developed berths with jetty access in commercial marinas.

iii) Developed dry storage facilities (on dry land) for boats.

[LPI agreed that they could have looked at a, b and c above, but they didn't.] WAG is of the view that LPI took a more difficult and unreliable route to get to the value of the wetland. WAG believes that a very different result would have been achieved by following a more direct route of what is being paid for undeveloped wetland by commercial marinas, based on valuations prepared by independent valuers, than what those commercial marinas charge for developed wetland and other ancillary services.

10. We question the use of a 50% discount applied to small wetland occupancies in paragraph 3 on page 22. We feel that it should have been much higher than 50%, as not only can one not build a dwelling on the land, but the land is either fully or partially submerged, is subject to uncertain tenure and in many cases is open to puQlic. access. In any case, the results of this exercise support our contention that the discount should be considerably more than 50%. [LPI agreed that they had not taken all these matters into account.]

11. We note the comment on page 24 that "".it was clear that discount/actors varied significantly" between Sydney Harbour, outer Sydney metropolitan region and the rest of state. If after a proper analysis of all the separately valued wetland occupancies in the state and a proper comparison of those to appropriate median postcode SLVs, the result was still that there were significant differences throughout the state, then we would have some sympathy for LPI's view that there should be a number of different discount factors. Just as IP ART has decided to move away from a single ROR, there could be a similar argument in relation to the discount factor.

12. On page 24 "Methodolgy Used to Determine Discount Factors", WAG is critical of the limited selection of 52 benchmark properties. LPI used only 3 POli Macquarie properties for the North Coast and only 3 Batemans Bay properties for the South Coast, while for Brisbane Water they used 3 from Brisbane Water East, but none from Brisbane Water West, while on the Georges River they included no samples from Precincts 2 and 4 . This very limited selection of properties is hardly representative of about 8,300 wetland occupancies throughout NSW. Depending on how the 52 samples were selected, LPI could perhaps be accused of being selective, in order to achieve a particular outcome. [LPI stated that as it based its exercise on what commercial marinas charge, they were limited to using samples close to commercial marinas. They also stated that they had to work within a budget set by IP ART.]

LPI'S ANALYSIS OF 52 BENCHMARK PROPERTIES

13. Regarding LPI's spreadsheet exercise on the 52 selected benchmark properties "to determine Discount Factors" WAG is critical of the following:

WATERFRONTACTIONGROUP A Subgroup of the Boat Owners Association ofNSW inc Page 9of!4

a) No fonnulae have been included, to enable us to follow what LPI have done. [LPI advised that the figures were not calculated by the spreadsheet, but were fed in. They acknowledged that there were errors, as found by WAG and advised that they were now correcting those errors.]

b) Except for cases involving reclaimed land, the starting point is "net rental income". But there is no explanation of from where this figure comes. There is no gross rent figure and no expenses figure provided. We have telephoned many of the property owners involved and there is no rent being paid or received on any of the structures. It is simply a "manufactured figure" and bearing in mind that the whole IPART review is about how one should set wetland rents, to start with an artificial rent figure, involving the wetland, is circuitous and not acceptable. [LPI advised WAG that the ''Net Rental Income" figure was an estimate based on discounting by 20% to 35% what is charged by local commercial marinas (which is a gross rent figure) to moor a boat (and provide lots of ancillary services). They admitted that they did not then further adjust the gross rent figure to bring it back to net.] WAG is very critical of this method and regards the "net rent income" figures as highly unreliable. If the net rent figures were reduced by say 35% to 40%, to allow for the normal differential between gross and net rent and very substantial maintenance costs on waterfront facilities, the difference in the end result, being the discount mUltiplier would be huge.

c) WAG has further discovered that the ''Net Rental Income" figure is based on applying a small discount to the rate charged by a local commercial marina to moor a boat, but then not varied according to the size of the facility on the occupancy. For example, there are 2 Port Macquarie samples with the same ''Net Rental Income" figure 0 f $4,160, even though the area of one facility is 18.72 square metres and the other is 36 square metres. LPI's result of analysing those two is that the smaller wetland occupancy is worth 2,107% more than the adjoining freehold, while the larger one is worth 73.84% more than the adjoining freehold, on a per square metre basis. WAG's view is that both are crazy, but one is substantially more crazy than the other. There are numerous similar examples.

d) Then the artificial net rent figure is capitalised using 4.5%, which is the NSW Government current ShOli tenn bond rate and which is subject to regular variation. The 10 year NSW Waratah bond rate of 5.1% is likely to be more stable and reflective oflong tenn interest rates. [LPI stated that they felt the 4.5% was a reasonable rate to use and no agreement was reached on this issue.] WAG is of the view that the 10 year bond rate is more stable and therefore more appropriate for the setting of a discount multiplier that is going to be set by IP ART now and then not reviewed for many years. WAG is also of the view that because the term of CLD licences is 20 years and many NSW Maritime leases are also for a tenn of 20 years, that a long term bond rate is more appropriate than a short tenn bond rate. Using 5.1 % as the capitalisation rate makes a big difference to the discount multiplier end


result. However WAG has been advised by its property valuation experts that the use of a bond rate does not comply with valuation theory, which dictates that a capitalisation rate should be set according to an analysis of market evidence.

e) Then from an artificial "Estimated Market Value" of the improved wetland is deducted the "Cun-ent Replacement Cost" of the improvements. But that "Cun-ent Replacement Cost" has been generally significantly understated, according to our discussions with jetty builders. WAG found that the description and sizes on LPI's spreadsheet sometimes differed substantially from the description and size on the "Benchmark Valuations Report" and therefore had some difficulty in getting quotes from contractors. In some cases, because we could not rely on LPI's descriptions and sizes, WAG had to either physically inspect the facilities, or when that was not possible rely on Google Earth to verifY what facilities were actually involved. Therefore the "Wetland Deduced LV" and everything that flows from it is incon-ect, for mUltiple reasons. [LPI agreed that their method in an-iving in these figures was rough. They stated that they had not made site inspections and that the limited budget set by IP ART restricted them to a "desk top" study. They agreed that WAG's method of making actual site inspections and getting quotes from local contractors was more reliable.]

f) WAG has found some strange examples of "Wetland Deduced LV" (value of mud & water). For example the mud & water at· . ___ - lPOrt Macquarie is stated as being valued at $3,312 ~uare metre, while the PSL V (which includes both wetland and dry land) for that sample is shown as being $150 per square metre, so LPI has concluded that the mud & water is 22 times more valuable. (For comparison, they valued mud& water at Vaucluse at $1,666 per square metre, roughly half the value of the Port Macquarie mud & water.)

g) The whole exercise has so many "rubbery" variables, that it does not provide any solid basis for LPI's Findings and Recommendations. [LPI agreed that their study was based on some figures that were not reliably based and which were subject to possible considerable variation.]

h) The "Reclaimed Land Market Value" would appear to be a wild guess, not based on any evidence. [LPI did not agree that their "Reclaimed Land Market Value" figures were "a wild guess", but agreed that they had not taken everything into account.]

i) WAG has discovered that within the 52 samples there are 6 major and dozens of less serious mathematical en-ors, which tln'ows further serious doubt on the usefulness of the spreadsheet. [LPI agreed that there were multiple en-ors and stated that they were now rectifYing those.]

j) There are 3 examples within their 52 samples where the wetland is valued at more than the dry land by up to 2,107%. WAG is surprised that such huge premiums applying to mud & water did not set alarm bells ringing and cause LPI to review what they had done.


k) WAG has received advice from one of its propelty valuation experts that this type of valuation exercise is prohibited by the Valuation of Land Act, when sub-letting is prohibited by a lease or licence. [This was not discussed at the meeting with LPI.]

WAG is seriously disappointed at the quality of this whole exercise and WAG therefore believes that this exercise is of no use at all in determining a reasonable discount factor or mUltiplier.

14. WAG believes that a sounder method "to Determine Discount Factors" is to examine ALL the SLVs of separately valued wetland occupancies in NSW and compare them to the SLY's of adjoining dry land. WAG made that suggestion to LPI and IP ART on 19 October and offered to cover 50% of the cost of extracting the additional data. Since IP ART was not prepared to commit to that course, WAG has agreed on 11 November to cover the full cost of extracting and providing that additional data ($5,930, being for 504 matching pairs).

15. We note that IPART basically agrees with WAG, in regards to the close relationship between the wetland and the adjoining dry land in the statement (page 32, paragraph 3) "We also note that the method of only including the SLVs of waterfront properties with occupancies ensures that only those properties most closely related to occupancies are included in PSLVs ... "

COMMENTS ON FINDINGS & RECOMMENDATIONS

16. We disagree with Finding No.3 (page 26) that "Wetland does not have a direct ,r,elationship with the per square metre value of adjoining land ..... The value of

.- wetland as used for domestic waterfront occupancies, is mor~related to marine rentals in a geographic location .... " and which LPI has slated elsewhere are irrelevant, because those rental rates include many other factors. Based on LPI's statements that "wetland does not have a direct relationship with .... adjoining land" and marine rentals are based on "what the market will bear ... with variations based on the quality of service and facilities ... " (page 18) , WAG has great difficulty understanding how LPI arrived at its Recommendation 2 of a 52 % discount for wetland, compared to the average of SLVs of waterfront properties (including that same wetland) within a precinct. [LPI differentiated between reclaimed land, mooring pens and wetland used for jetties, boatsheds etc. They also differentiated between Sydney and regional areas and stated that it was very difficult to arrive at a common discount multiplier, taking all these variables into account. They stated that they had been asked by IP ART to do a very difficult jo b and within the constraints of the instructions and the limited budget set by IP ART felt that they had performed satisfactorily.]

17. We disagree with the suggestion in Finding No.9 that wetland rents could be set by reference to the local swing mooring rate. The swing mooring rates are set arbitrarily and are not scientifically based. [LPI clarified that this finding relates only to mooring pens. }

18. We disagree with the term set by IP ART, as stated in the preamble to "LPI Recommendations" (page 27) that LPI consider a "fair return to government for an asset". The Crown Lands Act at Section 143 (1) (a) states ''the rent shall be the market rent for the land" and that should be the focus ofIP ART and LPI.

19. We disagree with the setting ofa lower discount for reclaimed land as recommended in Recommendation 1. The Government did not pay for the


reclamation and is therefore not entitled to a higher rent, because the adjoining property owner spent money on improving the wetland. As stated earlier, reclaimed land is subject to uncertain tenure, strict restrictions on its use, in many cases is open to the public and the owner is responsible for the cost of returning it to its original condition, when the lease or licence is terminated. [LPI agreed that they had not taken into consideration the restrictions imposed by the lease or licence or the fact that NSW Maritime had recently switched from 20 year leases to 3 year leases.]

20. We disagree with the discounts recommended in Recommendation 2. They are far too low, considering that we are dealing with small parcels of submerged or partially submerged, generally poor quality land, of uncertain tenure, which is generally open to public access and on which one cannot build a dwelling and is severely limited to what can be built. The occupancy instrument is also very harsh, adding to the unattractive nature of what the Government is offering. Ifthe owner of the adjoining property decided against renting the wetland, its value in the market would be negligible and that very low value is the market value. [LPI agreed that they had not taken all these matters into account.]

21. We disagree with Recommendation 3. We disagree that reclaimed land should enjoy a lower discount for reasons already stated.

22. We also disagree with the recommendation that reclaimed land in Sydney Harbour should enjoy an even lower discount than reclaimed land elsewhere. The higher value of reclaimed land in Sydney Harbour is already reflected in SL V s used in the formula and which causes the rent to be higher. To then reduce the dis,count (to again reflect the same higher value) is double dipping. [LPI listened carefulIy to this argument and seemed to understand it, but did not express a view.]

23. While it may be financially attractive to our members, with the highest swing mooring rate being about $3 per square metre per year, we disagree with Recommendation 4 on page 9 in regards to the suggestion of "linlang of wetland area rentals with the rate for public moorings". This is because the rate for public moorings is set arbitrarily and there is no property valuation justification for the setting of a market rent for land by reference to an arbitrarily set mooring rate. [LPI expressed the view that mooring pens should be charged on a different basis to jetties and other facilities. They expressed the view that the charge for mooring pens should be based on the local swing mooring rate.] (In many NSW Maritime leases mooring pens were charged on the basis of the local swing mooring rate prior to 2004, so LPI is in favour of returning to that principle.)

SOME GENERAL COMMENTS

24. We note that IP ART has decided to use the precinct median, rather than the average, for purposes of setting the PSL V. But LPI has done its calculations based on averages (quoting as the reason that the PSLV is an average), meaning that there the figures calculated by LPI now do not match with the basis selected by IP ART. We would suggest that LPI needs to redo its calculations based on medians (and ideally postcode medians, to match with the ROR). [LPI stated that their figures being based on averages could not be applied to medians. They stated that IP ART would need to issue new instructions to LPI to rework the figures, based on medians.]

25. Philip Western (Valuer General) and Simon Gilkes (previous Chief Valuer) on I June 2007 confirmed the advice ofLPI valuer that with waterfront


properties dividing the value by the area produces "meaningless figures". On that basis they said that the IPART PSLV formula is fundamentally flawed. What LPI is recommending is a continuance of that fundamental flaw. In contrast, what WAG is recommending by moving to a postcode based media SL V overcomes that major problem. [LPI recognised the problem and since the Chief Valuer was no longer present in the room, they agreed to discuss with the Chief Valuer whether they might change their recommendation to using a postcode based median and postcode based discount multiplier, as recommended by WAG.]

26. [When questioned about the discount multiplier (or factor) that WAG had recommended in regards to the fact that the ROR relates to land and house packages, rather than land and that WAG had established that the ROR on land was lower, LPI stated that IP ART had not requested LPI to look at that, so they didn't. However, WAG observes that in their instructions, IP ART did request LPI to "review and provide comment on ... ... WAG's proposed discount factors ... " (i.e. more than one) and WAG proposed two discount fuctors, only one of which has been subject to review and comment by LPI.

27. In its instructions to LPI, IPART stated " ..... LPI must also clearly explain the rationale and methodology behind its analysis and conclusions, cfocument key assumptions and provide relevant supporting evidence." WAG does not believe that LPI has fully complied with that instruction.

28. WAG has developed a sensitivity analysis that shows that even minor adjustments to the ''Net Rental Income", Capitalisation Rate and "Current Replacement Cost" makes huge differences to the discount factor derived from this exercise. That would partly explain the huge variances in LPI's discount factor, ranging from a 90.36% discount to a 2,107% premium. The truth is somewhere between those two figures, but one cannot determine what it is from LPI's analysis.

CONCLUSION - WAG is of the strong view that the LPI report cannot be relied on by IPART, because of the many serious problems with the LPI analysis, which have led to unreliable Findings and Recommendations. WAG is ofthe view that a far better alternative is to examine ALL the separately valued wetland occupancies within NSW and observe the discount factor that has been applied to those by professional property valuers, which have been engaged by LPI and therefore have LPI's confidence. That information has already been extracted by LPI and will be provided to WAG on payment of a $5,930 fee. WAG will combine that new data (504 matching pairs) with the 145 samples already in its possession, analyse the combined data and provide the analysis for the whole state and the supporting data to IPART.


Available Data

Fact Sheet 8 Calculating the Discount Factor

(as applicable to SLVs)

The Valuer General (VG) is required to provide an SL V for an occupancy in 2 circumstances.

The first is where an occupancy is adjacent to freehold(s) but where it has different ownership to the adjacent freehold(s). This usually occurs where an occupancy is shared between neighbours.

The second circumstance is where an occupancy is adjacent to a waterfront reserve and not the freehold of the owner.

WAG has been advised that there are approximately 200 such occupancies. These valuations can be compared to either the adjoining freeholds, where they exist, or to the Precinct SL V (PSL V) to arrive at a scientifically calculated Discount Factor (DF) in place of the current 50% discount factor (which has no basis in fact).

PSLV The original calculation of the PSLV per square metre had inconsistencies in the formula versus the surrounding text. However, it appears that IP ART intended to calculate the PSLV using the combined value of the freeholds and the occupancies divided by the combined area of the same freeholds and occupancies. The key point is that the discount factor applies to the average PSL V per square metre of both the freehold and the occupancy whereas it would have been more sensible to apply it to freehold value only.

Using the average value of freehold and the occupancy as a base introduces variability into that base figure. The rate per square metre can change just because, within a particular precinct, the occupancy represents a different percentage of the combined area of the freehold and the occupancy. To resolve this issue you either need a different discount factor for each precinct or to use an SLY basis that excludes the occupancies as the basis of the PSL V. This could be the PSL V of freeholds only or the median SL V for the postcode in which the occupancies are situated. The latter is preferred by WAG.

WAG proposes that, at the very least, the PSLV is redefined for this purpose to be the freehold SLY per square metre. This is the basis of the following calculations.

Calculating DF using Shared Occupancies Valuations There are 2 possible circumstances which need to be taken into account. The first is where the freehold SL V does not contain any occupancy area. The second is where one or both of the adjoining freehold SLVs includes some private occupancy areas eg a reclamation. In both cases there will be an SL V for the shared occupancy area. For simplicity the following abbreviations are used in the calculations below:

TotFHPOSLV

TotFHArea

TotPOArea

SOSLV

refers to the total SLY of the 2 adjoining FreeHolds and includes the value of Private Occupancies (if any) refers to the total area of the 2 adjoining freeholds without the area of any Private Occupancies refers to the total area ofthe Private Occupancies included in the freehold valuation refers to the SLY of the Shared Occupancy with a separate valuation

Page 1

SOArea SOVPSM DF FVPSM



refers to the area of the Shared Occupancy with a separate valuation refers to the Shared Occupancy Value Per Square Metre refers to the Discount Factor implicit in this valuation set refers to the calculated Freehold only value Per Square Metre

SOSLV SOVPSM =

SOArea

FVPSM (TotFHPOSLV -(TotPOArea x SOVPSM»)

TotFHArea

DF (as a %) = SOVPSM x 100 FVPSM

The same procedure can be applied to each such property set and all the DFs averaged to find a DF to be used in future calculations. This calculation uses the closest possible fit between occupancy and Freehold but cannot be performed where the separate occupancy SL V is on a waterfront reserve.

If the postcode median SLY and area are used the calculation is substantially simplified and it is just each occupancy value per square metre divided by the postcode median value per square metre.

Calculating DF using all Occupancy Valuations In this case we don't have freehold properties adjoining the occupancy in all cases. The best option is to compare the occupancy value per square metre to the available freehold only valuations per square metre where the comparison to freehold method is used. If IP ART choose to use the postcode median method then the calculation is exactly the same as noted above in the 'shared occupancy' calculation.

Obtaining the Data WAG has been able to locate 144 separate occupancy valuations by various methods. We set out below how we were able to obtain this information.

Staff of the VG's office were able to look at their database for various keywords. In the case of Maritime they searched for 'Maritime' and (either 'lease' or 'reclaimed'). They then searched for a matching pair to such record on the basis of matching address but excluding any suffix on the number. For example, 2a would match 2b and 5 would match 5-9. We then reviewed the results and eliminated any that were not perfect matches.

The same procedure was used for LPMA data but the keywords were 'licence' and 'permissive occupancy'

Ideally we would have searched for numerous other combination of keywords eg 'piles', 'jetty' etc but time and expense did not permit therefore we had to be content with about 80 sets of data within the Sydney region.

Page 2



We were the able to go back to the Excel spreadsheet provided by NSW Maritime and extract further iofonnation. This data, while it did not have address iofonnation, was provided io sets of generally 1, 2 or 3 lines. The 'valuation' line gave the area of the freehold and either the valuation of the freehold only or the combioed valuation of the freehold and the occupancy. The 'premises' line gave the area of the occupancy so if present its value was in the 'valuation lioe'. The 'valuation/premises' line represented a valuation of an occupancy. Any set that had 'valuation/premises' was an opportunity to calculate a discount factor.

Where 'valuation/premises' appears alone or in groups this seems to iodicate an occupancy on a waterfront reserve ie no attached freehold.

Where a 'valuation' and a 'valuation/premises' appear together this is the matched pair type infonnation provided by the VG. We then had to look at each set and match what we had from Maritime to the VG data and elimioate duplicates.

Where the was a 'valuation', 'premises' and 'valuation/premises' set together this was indicative ofa shared jetty (thus the valuation of the occupancy) with a private occupancy such as a reclamation adjoioing the freehold.

In all we found 144, however we are aware that there should be separately valued occupancies at Bar Poiot, Woy Woy and Tea Gardens at the very least, as well as a range of other which should be able to be located via 'the use of the spreadsheet in use by LPMA for calculation of rents.

The Benefits of using the Median SL V per Postcode While calculatiog the discount factor agaiost freehold only, rather than freehold and occupancy, will elimioate some additional variation, WAG's strong view is that the DF factor should be calculated against the postcode median SL V per square metre.

The benefits are: • The reference poiot (postcode wide) contaios thousands or tens of thousands of properties

giving a very stable median value and area. Our research shows that some of the current precincts only contaio a handful of properties

• The discount factor calculation is very simple • The differences between the areas on which precioct values are based, and which gives

rise to strange variances between adjoining or nearby preciocts, are substantially elimioated

• The calculation is fundamentally correct for application in a simple fonnula based on median postcode RORs applied to median postcode SL Vs, which are readily available from NSW Government sources.

Page 3

Fact Sheet 10 Correcting the Flawed ROR Calculation

The Issue The current IP AR T method of calculating the net rate of return (NROR) has a major flaw. It purports to calculate the NROR applicable to the land by deducting an outgoings % from the gross rate of return (GROR) for house and land packages for wider Sydney. At the core of this is an assumption that the GROR% is the same for the land $ value component and for the house $ value component.

Calculating the Implied GROR A simple method of ascertaining the implied GROR% applicable to house value and to land value is to perform a simple linear regression analysis of the median GROR % by postcode for NSW versus the % the land value represents of the median sales price for the same postcode.

The median sales price and median rent for 2010 by postcode was purchased from Housing NSW and the GROR% calculated as per normal practice. The median Land Value (LV) for 2010 by postcode was purchased from the Valuer General. The LV % was calculated as the median postcode LV (times 100) divided by the median postcode Sales Price.

When this is graphed there is a clear negative relationship between GROR % and L V% ie when the LV% increases the GROR% decreases.

The data set we used was state-wide and contained 307 postcodes. The chart is as follows:

~ ~ 0 ~

10.00%

9.00%

8.00%

7.00%

6.00% r-

5.00%

4.00%

3.00%

2.00%

1.00%

0.00% 0.0% 10.0%

Rates of Return on Land Content with House and Land Packages (based on 2010 individual entries)

Trend line: ROR% = 6 .3% - 3.92% x Land Content"!.

eg SO";. l an6d;~n_t~~t gives; ROR% of: 3.92% , "'%. 4.34'.

•

. ..!.'.1_.:~~ #,! :... ~~w 20.0% 30.0%

• - ':. #-. :!'

40.0%

• • '-.' r,.. ~

SO.O%

land Content'l.

60.0% 70.0%

•

-80.0% 90.0% 100.0%

This linear regression shows house value with a return of6.3 % and land value with a return of2.38%.

This is a mathematical way of exposing the flaw in the basic, and vital, assumption that the GROR for land and house are the same. Essentially this is saying that people need accommodation first and are willing to pay more for that essential item and secondly they want a desirable location, but are willing to pay a lesser % rate to fill that requirement.

Page 1


Analysing the variation If we look at the average GROR for the 10 lowest GRORs by postcode in 2010 we find:

• the average GROR is 2.61 % • the land represents 59.1 % of the total value • the average sales price is $1.624 million

The 10 highest GRORs by postcode in 2010 had: • an average GROR of 6.39% • the land representing 34.9% of the total value • with an average sales price of $236,100

In effect what this calculation is saying is that basic accommodation attracts a very high GROR in low valued and low % land content areas, but prime accommoqation attracts a very low GROR in high valued and high % land content areas . Note in particular the similarity of the GROR to the linear regression calculation above.

Comparing GROR across Types and Values of Accommodation LPMA, Maritime and WAG propose that the GROR be based on Houses rather than on Flats and Units as the latter has a higher GROR. Why do Flats and Units have higher GROR?

Flats and Units have more accommodation per unit of land and that is consistent with the mathematical calculation above as well as the analysis of the top and bottom end of the market. The higher the land % is of the total value the lower the GROR.

In the KPMG report, commissioned by LPMA, they include a chart ofGROR% versus Sales Value. That chart was first produced by WAG (fact sheet 6). It shows that as property value goes up the GROR reduces . The chart below is the equivalent of that KPMGfW AG chart but using 2010 data.

2010 Postcode RCR - All Sales Values ($,000)

8.W/o

7.0C1'/o

• 6 .Wlo

S.fD%

4.CXJ'1o

• 3'(Xl"/o • • • •

2.0)%

1.CXJ'1o

D.Wlo

:00 $500 $1 ,(0) $1 ,500 $2,(0) $2,500 $3,(0) $4,(0)

Page 2


Waterfronts are higher valued properties and generally much of this is in the LV. And it is clear that those higher values have lower GRORs.

We now have 3 indicators pointing to the fact that the more LV there is the lower the GROR and the less LV there is the higher the GROR.

Investment Theory Investment Markets tend to equalise return where the same risk is involved.

In looking at a house and land package we have 2 different types of assets. The house is a nonappreciating asset whereas the land is an appreciating asset. If we consider that the land might appreciate at say 8-10% per annum (i.e. the sort of number published by the popular press in their annual property surveys), and the house is likely to depreciate over time. If the same GROR were to apply to both, and for the sake of this exercise let's use a GROR of3%, we would have the land earning 11-\3% (after adding capital appreciation) and the house earning 3%. On that basis a rational investor would buy land and not waste his money on the house value. However, if we believe that there is a different GROR to the 2 assets, and let's take the numbers from the mathematical approach above, then the house GROR would return 0% capital and 6.3% income and the land 8-10% capital and 2.38% income.

Sample Data WAG has obtained, from a small number of Sydney real estate agents, recent rent and sales information for both waterfront and non waterfront properties allowing it to calculate actual GROR for properties in high valued areas. WAG is aware of 4 properties from within that limited sample (3 waterfront and 1 non waterfront) that have GROR below 2% (1.92% on $4.75m, 1.76% on $2.7m, 1.59% on $3.6m and 1.57% on $2.6m).

In addition we note this advertisement:

)lr , Nab

This is a GROR of 1.1 % at the advertised price and 1.2% if the actual sale price was 10% lower than advertised.

Page 3


Bringing it Together It should be clear now that there is a different GROR applying to land than there is to the house and that the GROR on land value is considerably lower than that of the house value. So we now have some limits to work within. Mathematically we can see that 2.38% is the maximum that can be expected from land. These are state-wide GRORs.

We are now nan-owing in on a GROR for land which only needs a small number of deductions. They are land tax, council rates on unimproved land, the charge for having water running past the property, vacancy allowance, land portion of lease up fee and managing agent fees. The matter of depreciation on house and contents is avoided as is house and contents insurance and repairs and maintenance on any improvements.

At this point it should be clear that it is not appropriate to assume that house and land have the same NROR and it is necessary to use a better sun-ogate for the NROR for land.

The Solution The GROR needs to be tied into the 'local' market given that there can be large differences between areas e.g. Western Sydney suburbs and Eastern Sydney suburbs, as has occun-ed recently. Markets can, and do, move in different directions so the application of one GROR to a different market will result in rents that do not represent 'true' market and can give widely fluctuating results.

WAG proposes a GROR discount factor. The discount factor would essentially standardise the GROR for variation in land content.

WAG has reviewed the SLVs and Sales Prices by postcode for 2007 to 2010 and it appears, on average, that the GROR on land is one half of the GROR on the house, with the weighted average being the GROR on the house and land package. A simple mathematical equation can be used to deduce the underlying numbers.

For the 3 years ending 2010 the GROR for Sydney Harbour was 3.17% and the land value represented 64.20% of the total value, whereas for the other Sydney waterways the GROR was 3.48% (pittwater), 3.83% (Georges River) and 3.75% (port Hacking) the land represented 65.06% (pittwater), 69.93% (Georges River) and 71.97% (port Hacking) of the total value. For regional areas the GROR was 4.33% and the land value represented 52.22% of the total value.

The calculated Land Value GRORs are: Sydney Harbour 2.33%, Pittwater 2.58%, Georges River 2.94% and Port Hacking 2.93% and Regional is 2.93%. Importantly these numbers will tie into the market fluctuations of the region in which they are situated.

The GROR Discount Factor (GRORDF) is simply a corollary of the discount required to bring back (the house and land) GROR from 3.17% for Sydney Harbour to 2.33% (being the Sydney GROR for land). For the other 3 Sydney waterways the GRORDF is applied to bring the (house and land) GROR of 3.48% for Pittwater back to 2.58% (being their GROR for land), 3.83% for Georges River to 2.94% and 3.75% for Port Hacking to 2.93%. In the Regional Areas the GRORDF is applied to bring back (the house and land) GROR from 4.33% to 2.93% (being its GROR for land). Those Discount Factors are-Sydney Harbour 73.64% Pittwater 74.11% Georges River 76.88% Port Hacking 78.10% Regional areas 67.67%

Page 4


NOTE: These figures are calculated from the most recent 3 years of data and for 100% accuracy they should be updated annually (i.e. a 3 year rolling average)

Possible Further Simplification - Rather than calculating a GROR on house and land packages, which is not appropriate to DWF occupancies, and then discounting it back to a GROR on land only, which is rather circuitous, it might be more appropriate to simply use the GROR on land for Sydney and regional areas, calculated by the method that we have used. In that case the GROR, outgoings and net ROR (NROR) 3 year rolling average figures are as follows:

LOCATION GROR OUTGOINGS NROR SYDNEY HARBOUR 2.33% 1.30% 1.03% PITTWATER 2.58% 1.17% 1.40% GEORGES RNER 2.94% 0.91% 2.03% PORT HACKING 2.93% 1.01% 1.92% REGIONAL AREAS 2.93% 1.15% 1.78%

Page 5

School of Urban Development Ph: 07313S9Il2 Sib October 2009

Waterfront Action Group (WAG) 152 Cammeray Rd., Cammeray, NSW 2062 Australia

Queensland University of Technology Faculty of Suitt Environment and Engineering

2 George Street GPO Box 2434 Brisbane Qld 4001 Auslraf", Phone +61 731381433 Fax +61 731381529 Email [email protected] www.bee.qut.edu.au

Re: Report on "A review into rentals for waterfront tenancies on Crown land in New Soutb Wales" as compiled by Egan National Valuers (NSW)

Dear SirlMadam

Thank you for the opportunity to review the information, data and conclusions compiled by Egan Natioual Valuers (NSW) based on data provided by Mr G Forster in relation to the comparison of rates of return for higher value property in comparison to lower value property based on NSW postcodes, with particular reference to Pittwater postcodes.

This report follows a similar report carried out by the same authors for a study of waterfront rates of return for Sydney based postcodes and my subsequent review of this report. I once again purport that the uoderlying research and analysis of both these reports are consistent and provide a logical and sound methodology to assess the variation in the gross and net rates of return for property that is in postcodes located in waterfront locations ofPittwater compared to adjoining and nearby postcodes that do not have property with waterfront access.

Again, as I stressed in the previous work by these authors, the work carried out for the Pittwater study by Egan National Valuers (NSW) is based on souod research and valuation analysis approaches and the findings are supported by me and considered in line with current property valuation and investment practice and theory.

As background to this report, it is important to note that waterfront property in the major residential cities and towns of Australia, is a separate market to any other residential market in the same location that does not have direct waterfront access, even for property in the same post code areas and any comparative analysis from a valuation or investment perspective has to address these differences.

I refer to a study by Eves and Adair (2005) that found that there was considerable difference between the capital growth (price) in prices for waterfront property (waterfront was differentiated from property simply with water views), in the inner harbour locations of

CRICOS />10.00213.1 ABN 83791 724 G22

Sydney compared to similar residential property in the same Local Government Areas and the same postcode locations. On an average annual basis this difference was up to 10% for the same postcode and up to 60% higher than the average capital return for the median property price for Local Government Area. Although this study was based on waterfront property in the Sydney Harbour locations, it is applicable to other areas, such as the Pittwater location in Northern Sydney.

This 2005 study /llso showed that the higher prices for waterfront property did not result in an equivalent increase in weekly rental values and the end result was a lower income return for waterfront property compared to lower value properties in the same postcode and Local Government AJea location.

This study also showed that this inverse relationship between higher value property and low rates of income return (rate of return) is even greater as the price differential of the property increases. Again, this was evidenced in the study by the decrease in the average annual capital return from the high value sub-sector to the immediate postcode area to the Local Government Area median house prices, over the study period of 11 years.

Since this study a review of the Brisbane residential rental and sales markets by the Property Economics department at Queensland University of Technology (2009) confinns that the gross rate of return for residential property in low and medium socio economic area~ of Greater Brisbane have significantly higher gross rates of return compared to suburbs in high socia-economic locations, and in particular those suburbs which have direct access to the Brisbane River.

Based on average rentals and median prices for 3 and 4 bedroom houses in 24 suburbs pf Brisbane, there was a significant variation in gross ,rates of return based on suburb location and economic status. A stimmary of the findings based on data as at June 2009 is below .

Table 1: .Brisbane Socio-economic snburbs: Rate ofRetnrn: June Quarter 2009

LowSocio- Middle SociI)- High Socio-economic Suburbs economic Suburbs economic Suburbs

Gross Rate of 3.92 to 4.97 3.02 to 4.13 2.05 to 3.13 Return: Ran2e (%) Average Gross Rate 4.44 3.55 2.59 of Return

In a comparative study by Eves and Wills (2003) of residential property investment performance across the capital cities of Australia, it was found that although there was a narrower range in total return performance for residential property across all state capitals of

Australia, there was considerable difference between the capital and inc.ome returns of residential property across these cities. The results s~.owed that the cities with the l.ower residential hou~ing median price had the higher inc.ome returns c.ompared to the cities with higher median h.ouse prices. The t.otal returns in the higher value markets being m.ore a functi.on.of capital growth rather than inc.ome gr.owth. In these cases the rate .of return based .on rental inc.ome f.or cities such as Sydney, Brisbane and Melb.ourne were far l.ower than the rate .of return fr.om rental income in Adelaide, H.obart and Cimberra

These previous studies supp.ort the research and [mdings in the rep.ort by Egan N ati.onal Valuers (NSW) that the gr.oss rate .of return (inc.ome return) is l.ower in higher valued property markets c.ompared to l.ower valued property markets and that an average .of a large market cann.ot be c.onsidered t.o be representative .of a verY small subset of a particular property market, especially if that subset market has attributes that are n.ot available t.o the t.otal market, which is the case f.or the situation in relati.on t.o the Pittwater waterfr.ont property market.

In additi.on to the analysis and interpretati.on .of the gr.oss rate .of return data provided in the Egan National Valuers (NSW) report, I als.o reviewed the meth.od.ol.ogy and conclusions quoted in relati.on to the determination of a net rate of return for the subject postc.ode areas. The method used by the writer is based ou standard property industry practice to determine a net rental figure. All statutory charges, depreciation allowances, insurance, management and maintenance costs associated with investment property ownership have been deducted in acc.ordance with standard industry practice in respect t.o the calculati.on .of a net rate .of return for income producing rental property. Adequate c.overage has also been given to the pr.oblems and issues ass.ociated with establishing an average figure t.o be ad.opted acr.oss a range .of property types within a specific p.ostcode area.

Please d.o n.ot hesitate t.o c.ontact me t.o pr.ovide further clarificati.on .of this .opinion.

n.

Y.ours faithfully

Pr.ofess.or Alfred C Eves B.Bus (Banking and Finance), Grad Dip Ed (Technical), Master.of C.ommerce(H.ons), PhD, Dip App Sc (Ag), Ass Dip Val, AAPI, FRICS

Pr.ofess.or in Pr.operty Ec.on.omics

14 November 2011

The Chairman Independent Pricing and Regulatory Tribunal PO Box 0290 OVB Post Office NSW 1230

Dear Sir,

CHALONER ATIO S

Suite 618, 368 Sussex Street,

Sydney NSW 2000

PO Box 20300

World Square NSW 2002

t. +61 2 8064 3434

1. +61 2 8064 3432

e. [email protected]

Re: Waterfront Action Group Submission on the Review of the Current Method of Rent Determination for Domestic Waterfront Tenancies in NSW

I refer to the Waterfront Action Group (WAG) submission dated 13 November 2011, in response to both

your draft report dated October 2011 and an attachment to this submission titled "some criticisms of the

LPI consultancy report".

I do not intend to revisit in depth the information and opinions expressed in this very extensive

submission, however I will highlight some critical points made by WAG from a property valuation

perspective and provide a critique of the Land and Property Information (LPI) submission.

One of the most critical issues to attempt to arrive at a market based outcome is the application of rates

of return (ROR) must be on a time, value and place specific basis, this is basic valuation theory and

cannot be ignored if a legitimate attempt to interpret market is to undertaken. IPART's draft report has

proposed applying a median ROR of a postcode or group of postcodes to the median Statutory Land

Value (SLV) of a group of homogenous waterfront properties. Such an approach will I believe not

determine in any form a result that is based on the market as dictated by Section 143 of the Crown

Lands Act.

If IPART wish to use a median SLV for a group of waterfront properties, then clearly the ROR for a

similar group of waterfront properties must also be applied. Unfortunately, such data ROR is not readily

available. As a consequence to correctly interpret market the median SLV for a group of properties

within the same postcode area should be used.

WAG has noted within their subrnission numerous anornalies resulting from the application of this part

of the proposed formula. This confirms in my opinion the draft formula is seriously flawed.

With respect to the LPI submission, I have numerous criticisms of this report and I believe the

recommendations this organisation have provided to IPART are clearly incorrect.

Property Valuation 8< Consultancy

Liability Limited by a Scheme approved under Professional Standards legislation

Chaloner Valuations Ply Ltd

ABN 27 145 343 862

A summary of my criticisms are as follows:

CHAlONER VALUATIO S

• LPI have ignored WAG's preferred approach of comparing SLV of the occupancy to the SLV of

adjoining waterfront freehold. Reasons given include there is disconnection between the value of

the occupancy to the value of the freehold. Yet LPI prefer to compare the value of the occupancy

with the precinct average SLV (PSLV), which in my opinion is even a more substantial disconnect.

• LPI stated in a meeting on 9 November 2011 that they believed there were inaccuracies in their

SLV's of the occupancies and that basically this information cannot be relied upon, preferring

instead their "benchmark" valuation approach. This simply just does not wash - surely LPI would

stand behind all the valuations they issue, such valuations having been carried out by registered

contracted valuers.

• Instead LPI has preferred to use a benchmark valuation approach, whereby 52 occupancies have

been valued by a convoluted method to arrive at a rate per square metre ($lm 2) and then compared

to the precinct SLV also on a $lm 2 basis.

My criticisms of this approach are as follows:

1. The benchmark valuations are on a market value basis whilst these are compared to a statutory

land value. I would suggest like with like is not being compared.

2. In assessing an improved rental for each benchmark, LPI assumes the adjoining freehold owner

would lease it for full market rates, when he is in fact the only potential lessee for such an area.

Such an assumption is flawed and does not reflect market rental value in any way. As LPI points

out in their report (but then ignores) the evidence supports a minimum 50% discount just for this

factor alone.

3. No methodology as to how such improved rentals have been assessed in comparison to

commercial marina rentals has been provided. It would appear a discount in the range of 20% to

30% has been applied. In my opinion, such a small discount is manifestly inadequate.

4. Differentiation between net and gross rentals has not been explained. When questioned about this

matter, LPI seemed to have overlooked it, but suggested that such differentiation might be covered

in the 20% to 30% discount noted above. The expenses on waterfront facilities are substantial and

are themselves likely to exceed 30% of any rent received.

5. Application of a capitalisation rate by reference to a bond rate is just ridiculous. Basic valuation

theory dictates the only way to assess a yield (or capitalisation rate) is by reference and analysis of

market evidence. I acknowledge that it is likely no such market evidence exists, however

application of an artificial rate is plainly wrong. It is further evidence that LPl's approach is flawed.

Our Ref: Letter to IPART 14 Nov 11NJAG

CHAlONER VALU 10 S

6. In arriving at a land value for the mud and water component of their benchmark occupancy, LPI

have deducted the replacement cost only of a new jetty and other improvements. I believe this is an

incorrect application. As well as deducting replacement cost, other deductions should also be made

including holding costs, interest, profit and risk allowance, stamp duty, legal fees etc. This is what is

known as a hypothetical development exercise. Further, LPI's approach of assuming the

improvements exist then deducting their cost is against long established court precedent. I would

note the 1924 case of Toohey's Ltd VS The Valuer General an authoritive case in such mailers

states in assessing an unimproved value (land value) improvements are assumed to be non

existent as if they never existed. Therefore, I believe LPI's approach would not stand scrutiny in a

court.

7. I believe that in assessing the value of reclaimed land by deducting 40% or 50% from the freehold

of the adjoining land value as LPI have done and quoting the "Morts Dock" case as precedence and

guidance for such approach is the wrong application of the principles of this case. Basically "Morts

Dock" principal applies when valuing a remnant parcel, when there is only one potential purchaser.

The principle is to split the difference of the added value to the main parcel when the remnant

parcel is added. LPI have not done this, they have assumed the remnant parcel (in this case the

occupancy) has the same value of the adjoining freehold (not the added value) before applying

"Morts Dock" principal. This is a very serious flaw in their approach and if the right application of the

principle was made the discount would increase substantially, for the reclaimed land component in

their exercise

8. I do not believe it is appropriate to assume an occupancy can be leased at market on commercial

terms, when the Crown Lease (on which the point of this exercise is to estimate market rent)

specifically prohibits such leasing.

In conclusion with respect of the LPI Consultancy Report, I believe there are far too many estimates,

guesses, undisclosed rational, incorrect application and ignoring of basic valuation theory to arrive at

any meaningful conclusions that may assist IPART in their deliberations, in determining what an

appropriate discount factor may be.

The most sound method using currently available information is the application of analysis of available

valuation evidence by way of comparison between SLV's of the occupancy area and the adjoining

freehold, being the approach recommended by WAG.

Our Ref: letter to JPART 14 Nov 11NoiAG

CHALONER VA UATIONS

I would further note that having carefully reviewed the WAG submission dated 13 November 2011, I

fully agree with and endorse all the rationale noted in such submission.

Yours faithfully, Chaloner Valuations

Paul Chaloner, FAPI Registration No. 2577 Certified Practising Valuer

Our Ref: Letter to IPART 14 Nov 11fWAG

School of Urban Development Gardens Pt Campus Brisbane

13th November 2011

Waterfront Action Group (WAG) PO Box 639

SPIT JUNCTION, NSW 2088 Australia

Queensland University of Technology

Re: Report on "The WAG submission and response to the draft October 2011 report in relation to the IPART review of method for determining rents for domestic waterfront tenancies in NSW.

Dear Sir/Madam

The following endorsement is in relation to the overview of the report the Waterfront Action Group (WAG) has supplied in response to the draft IPART report that has been recently issued. Following the release of this draft report WAG has carried out a significant review and critique of the document and other material supplied to them. My endorsement covers the WAG report and the various documents that they constructed or accessed to develop this rejoinder report.

Over the past two years I have had the opportunity to review, critique and comment on the issues of calculation of rates of return and valuation theory, as raised by WAG in a number of submissions that they have prepared on the issue of rent determination for wetland leases.

As an academic, with substantial experience in the market analysis of a wide range of property sectors, including waterfront residential property, and a leading educator in property valuation theory and practice, I have been able to review the data, research methods, market analysis and formula calculation approaches adopted by WAG in their various submissions to the IPART review and legal proceedings in relation to the calculation of rents for land below water in which they have been involved.

In all cases I have endorsed the approach, market analysis and conclusions that WAG has presented in these submissions, as their research methodology has always followed current industry practice and has been supported by grounded property, investment and valuation theory and practice.

Having completed the review of the current WAG submission I fully endorse the submission that WAG has prepared in response to the IPART October 2011 Draft Report and offer the following points.

There are several issues that have been raised in the WAG report that are crucial to the determination of a fair market rent for land under water (wetlands), and these do need to be fully considered to ensure that the calculation of these lease rentals are based on sound property investment and valuation principles.

In my previous reviews of the waterfront leases and rent determinations carried out by WAG, I have provided considerable evidence in relation to the impact of property location and type on the rate of retum that can be generated from that property and the importance of applying consistent calculation methods and inputs across these various markets to ensure accuracy and faimess in the calculation of rents .. These views have been based on published property investment papers and property investment theory.

The current submission by WAG again highlights a number of issues that need to be addressed, if a fair market determination for wetland rents is to be achieved. The issues and concems raised by WAG include the rent formula, the quality and extent of the data used in such calculations, selecting the appropriate discount rate and determining land value and the rate of retum that should be adopted in a fair rental determination formula.

Of particular concern from a property market investment and analysis perspective is the use of incomplete data sets to deterrnine market rates of return and the use of differing market parameters to determine a rent for a very specific and unique property class. I agree with the approach by WAG that all data used to deterrnine a rent for wetland should be as comparable to that land as possible; this includes location, size and tenure.

WAG has also raised legitimate concerns in respect to the discount rate that should be applied when determining wetland rents. The rate of return for residential property is based on the income that is received for the occupation of the house. The house generates the income, with the land facilitating the ability of the land owner to build the house. The actual rent paid is based on the location of the property (dwelling and land) and more importantly the size, age, inclusions and condition of the house. In any location a larger, well maintained and appointed house will always achieve a higher rent than a srnaller house or a sirnilar sized house that is not as well appointed. These principles need to be applied to the wetland situations. WAG correctly state that wetland cannot generate an income, and any improvements on the wetland are actually an ongoing expense and the cost of maintenance and depreciation should be included in the calculation of the discount factor.

I concur with WAG that any rent determination formula for wetland has to be based on comparable data and "like" with "like". Any calculations based on a srnall percentage of selected data cannot be considered reliable or accurate. The strongest aspect of the WAG analysis, from rny perspective, is the fact that the market analysis and the proposed formula construction is based on the complete set of wetland rental data in the study areas and not a set or selected sarnple of properties. A full data set allows a more accurate analysis of the data and any data sorting can be applied in a logical way to minimise errors. This is not possible if only a small sample is used in the analysis.

The WAG review document provided as part of this submission review also included a five (5) step formula that they have developed as a preferred model for the calculation of wetland rents. From a property investment and property market analysis perspective, this is considered to be a more reliable and fairer approach to calculating the rate of return, property value and subsequent wetland rent than the current scheme. Although the most accurate approach would be to calculate the rate of return and SLV for each individual wetland lease, it is acknowledged that this is not feasible. However, the methodology put forward in the WAG submission is a very sound approach and provides a greater level of fairness than the current system and those that I have seen presented as part of this overall review.

In conclusion, I fully endorse the issues raised by WAG in their November 2011 submission to IPART and I also endorse the market analysis approach used to develop the proposed alternate rent determination formula.

Please do not hesitate to contact me to provide further clarification of this endorsement of your review document to IPART.

Yours faithfully

Professor Alfred C Eves. B.Bus (Banking and Finance), Grad Dip Ed (Technical), Master of Commerce(Hons), PhD, Dip App Sc (Ag), Ass Dip Val, AAPI (CPE), FRICS, MUDIA Professor in Property Economics Queensland University of Technology

MICHAEL CHAPMAN BA LLB Maritime Consultant & Lawyer

Suite 19 Princes Street Marina, 16 Princes Street, NewportNSW 2106

Chairman I PART PO 80x0290 OVB Post Office NSW 1230

Dear Sir

RE Review of method for determining rents for domestic waterfront tenancies in NSW

This is in response to the draft recommendations in section 8 of IPART 's draft report of October 2011

Disclosure The writer is i. a committee member of Waterfront Action Group . ii. advisor to in excess of 80 domestic tenants of NSW Maritime and in excess of 30 domestic tenants

of Department of Lands

CONCERNS APPURTENANT TO REVIEW ---IMPLEMENTATION

Based on experience since the 2005 IPART Recommendations were adopted by the NSW Government, I have concern that NSW Maritime (now Roads and Maritime Services) may repeat its sins of the past and while claiming to implement IPART Recommendations, do the opposite in the rental provisions of lease imposed on its tenu~e holders ..

I. The question of how the initial rent is to be set has not been ,,!ddressed; at present Maritime.charges rent up to 62% above what IPART is currently recommending in its draft report; how is this to be redified and the initial rent established for 2012?

2. Rent is initially set at a stated figure (in Item 5(a) of Annexure B) of the Maritime leasa, and Clause 3.7 deals with how that initial rent figure may be changed. Item 5(b) of Annexure 8 also deals with the matter. The problem seems to be in the definition of "Rent Formula!! I which does not include any reference to the Rate of Return and in fact is not the IPART formula. The "Rent Formula" is defined as "the Wetland Rate ($/m2) x the Rent Area (m2) ...... ", which is some other formula, which includes a wetland rate that might .be set quite arbitrarily.

3. The current Maritime lease states [in 3.7(a)] "The Lessor may vary the Rent Formula to implement a recommendation of the Independent Pricing and Regulatory Tribunal.. .. " does not require the Lessor to do anything.

4. The words Dn 3.7(b)] "The Lessor must vary the Rate of Return each year. .... " may provide no real effect on the rent, because the Lease does not specify how the Rate of Return is applied or how any change in the Rate of Return is to be applied. In particular, if the Lessor chooses not to vary the Rent Formula under 3.7(a), then changing the Rate of Return under 3.7(b) will be of no effect.

5. Maritime occupancy holders are left without any rights of appeal to a tribunal or court. That would allow the current situation to continue, whereby CLD occupancy holders can appeal to the Local Land Board orland the Land and Environment Court (and then to higher courts), but NSW Maritime occupancy holders have no such opportunity. WAG believes that as well as providing a normal legal review process, having a right of appeal will improve the quality of NSW Maritime decisions and processes.

G. IPART recommended long term tenures consistent with the period to amortise the investment in a typical jetty and suggested 20 year term as part of a bundle of market rights in return for paying market rent; Maritime presenlly offers 3 year terms with monthly holder-over; this is contrary to IPART's Recommendations of 2005

I would appreCiate your consideration of the matters raised in regarqs to the lease because it is the vehicle by which the IPART Recommendations, if adopted, are deliverable. Ii may well be that these fall outside the· scope of the current Review, but unless recommendations are applied accurately and as intended, the IPART' Recommendations may be thwarted and of little benefit, in the hands of a monopoly lessor whose decisions and administration are not appellable.

Yours faithfully

,; , MICHAEL CHAPMAN 7 November 2011

PO Dox 180 Freshwater NSW 2096 chappo@bigpond,net.au .0408164361 p/f9999 0531 ADN 16 927 842 774 Liability limited by a scheme approved under Professional Standards

HONES LAHO~

Our Ref: BKH:KH:09363 Your Ref: Brian Folbigg

14 November 2011

Waterfront Action Group PO Box 639 SPIT JUNCTION NSW 2088

Attention: Brian Folbigg, Phillip Altman, George Robert Citer, Barney Remond and Michael Chapman

Dear Sir

Domestic leases issued by NSW Maritime Premises:

Comments in regard to leases presented and the I Part draft report

Upon examining leases presently submitted and recommendations of IPART, we note the following consideration and expansion.

comparing those with the matters that required further

• The relevant clause in all leases so far submitted is the rent review formula or rate of return expressed at paragraph 3.7.

• Upon examining that clause we note the following clause which continues to cause concern:

o 3.7(a). There is no obligation on the lessor to conduct a review. The clause states "The lessor may conduct a review of and vary the rent formula at any time, provided that at least 4 years has elapsed since the last rent review ... ".

o As a fair and equitable arrangement between parties we would recommend that there be an entitlement on the part of the lessee to request the review in the event that the lessor does not undertake its review after the 4 years has elapsed. There must then be the mechanics for an enforceable obligation to review.

o 3.7(a) requires that the lessor must notify in the event that it has changed the rent formula and this notification must be 60 days of the rent formula review. Our submissions in May pointed out that the term expressed in the lease placed no fetter and no conditions upon an absolute right to replace the rent formula to any formula that Maritime requires. This concern still remains and must be addressed.

o We note that the precincts are to be amalgamated from a total of 61 into 20 regional precincts. Some consideration needs to be given to those precincts

Suite 2, Level 4, 54 Miller Street North Sydney NSW 2060 PO Box 1989 North Sydney NSW 2059. OX 10534 North Sydney

T: 0299293031 F: 0299297071 www.honeslahood.com.au

B K Hones & J Hones & W M La Hood trading as Hones La Hood ABN 21 030200477 liability limited by a scheme approved under Professional Standards Legislation

Waterfront Action Group Page 2 14 November 2011

that abut another to thereby provide some moderation between discrepancies that could exist within close proximity of one property to another in a different precinct.

o It is essential that there be a right of appeal against the Precinct Statutory Land Value (PSLV) as well as the right to challenge the Rate of Return.

o While there is a formula for the calculation of Wetland Rate, there is no indication of how one calculates other areas leased (i.e. reclaimed land). This needs to be addressed for the clarity of all parties.

This letter is a summary of the major concems but is not an exhaustive examination of all lease matters which could be further improved.

Yours faithfully HONES LA HOOD

Brian Hones Senior Partner [email protected]

Date post:	30-Sep-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times

EXECUTIVE SUMMARY WATERFRONT TENANCIES IN NSW. · separately valued wetland occupancies in Sydney...

Documents