Graph v Relativity for 70 years unexpired term [2015] UKUT 651 (LC)

IN THE MATTER OF AN APPEAL FROM A DECISION OF THE
UT Neutral citation number: [2015] UKUT 651 (LC)Venue FIRST-TIER TRIBUNAL (PROPERTY CHAMBER)
Between
GraphRelativity for 70 years unexpired termRelativity for 75 years unexpired termCommentsCluttons86.289.65W A Ellis8588Houses onlyJohn D Wood/Gerald Eve879090% HousesCharles Boston8790Flats and houses including some in Greater LondonAverage excluding Cluttons86.389.3
Colin Hurst MRICS of Colin Hurst and Partners for the appellantPaul Harper LLB (Hons) of Collingwoods for the respondents
[1]This is an appeal from the decision of the First-tier Tribunal Property Chamber (Residential Property) (“the F-tT”) dated 14 August 2014 whereby the F-tT decided that the price payable by the appellant to the respondents for the granting of an extended lease (pursuant to the Leasehold Reform, Housing and Urban Development Act 1993) of the top flat at 17 Ormiston Road, London W12 OJR was £39,000.[2]17 Ormiston Road is a Victorian mid-terrace house in a residential road in Shepherd’s Bush. The property was originally one single house but has now been divided into two self-contained dwelling units, the ground floor remaining in the ownership of the respondents who retain the ownership of the freehold and who are the landlords of the appellant. The appellant holds the upper unit, which can conveniently be called the top flat, from the respondents upon a long lease at a low rent and it is in the right of holding this lease that the appellant is entitled to acquire the grant of an extended lease pursuant to the 1993 Act.[3]Various matters were agreed before the F-tT. One matter which was not agreed before the F-tT and which the F-tT therefore determined was the freehold vacant possession value of the flat as compared with the agreed long leasehold value of the flat. The F-tT decided that the freehold should be reflected by a 0.5% uplift on this long leasehold value and there is no appeal against that aspect of the decision.[4]In consequence there are two points which remain in dispute so far as concerns the assessment of the price to be paid. The parties have prepared a statement of agreed facts and issues. Taking first the agreed facts they are these:(1) The valuation date is 2 October 2013.(2) The present lease is for 99 years from 25 December 1986.(3) The years unexpired as at the valuation date were 72.167.(4) The long lease value is £650,000.(5) The freehold value of the flat without a leasehold interest is £653,250.(6) The value to be put upon the capitalisation of the ground rents is £3,085.[5]The following two matters, which are necessary ingredients in the calculation of the price to be paid, remain in dispute and are the subject of the present appeal namely:(1) The deferment rate to be applied when calculating the value as at the valuation date of the reversion. This reversion would confer upon its owner vacant possession of the flat in 72.167 years. The appellant contends for a deferment rate of 5.695%. The respondents contend for 5% (the figure adopted by the F-tT).(2) The relativity, that is to say the percentage of the freehold value (namely £653,250) which represents the value of the unexpired term of 72.167 years as at the valuation date. The appellant contends for a relativity of 96%. The respondents contend for 91.4% (the figure adopted by the F-tT).[6]Before the F-tT Mr Hurst on behalf of the appellant adduced substantial evidence and presented substantial argument upon both the question of the deferment rate and the question of relativity. The evidence presented was not exactly as presented to us on this appeal, but it is accepted by Mr Harper that in substance evidence and argument broadly as extensive and to the same effect as that presented to us was presented to the F-tT.[7]In granting permission to appeal the Deputy President observed that it is arguable that the F-tT did not sufficiently engage with the evidence presented to it and that it gave an inadequate explanation of its reasons for rejecting the appellant’s case. It was ordered that the appeal before the Upper Tribunal should proceed by way of review with a view to a rehearing - i.e. such that there would be a rehearing if it was decided at the review stage that the F-tT’s decision could not stand.[8]At the hearing before us the representation was the same as it was before the F-tT namely:(1) The appellant was represented by Mr Colin Hurst of Colin Hurst and Partners, who acted both as advocate and as expert witness on behalf of the appellant. Mr Hurst is a chartered surveyor (MRICS) who has undertaken surveying and valuation functions in the south east of England for the last 30 years and most frequently in west London. He has substantial other professional experience and qualifications including a degree in soil science from Reading University, a Diploma in surveying from what is now East London University, and a Higher National Certificate in civil engineering.(2) Mr Paul Harper of Collingwoods appeared on behalf of the respondent, both as advocate and as expert witness. Mr Harper has a Law Degree LLB (Hons), he has worked and lived in Shepherd’s Bush and Hammersmith for about 26 years and he commenced employment with Collingwoods in 1988. Collingwoods have been established in Hammersmith since 1932 and they manage residential property portfolios and also undertake sales and lettings of both residential and commercial property. Both Mr Hurst and Mr Harper gave sworn evidence before us. Review stage[9]It is first necessary, at the review stage of this appeal, to consider whether the F-tT’s decision can properly stand such that no rehearing is appropriate. We can take this aspect of the appeal briefly.[10]As already indicated there was presented to the F-tT by Mr Hurst substantial evidence and argument upon both the question of the deferment rate and the question of relativity. The totality of the F-tT’s reasoning in rejecting Mr Hurst’s argument and evidence on the deferment rate is contained in paragraph 6 of the F-tT’s brief decision:
“The Respondents’ representative stated that he felt that the appropriate deferment rate was 5% in line with Sportelli . The Applicant’s valuer argued that the risk of obsolescence merited a 0.25% increase on Sportelli and that a further 0.25% increase was justified by the increased risk of investment in the W12 postcode area and finally that another increase of 0.25% should be added for the increased burden of administration giving a total of 5.75% as an appropriate deferment rate. The Tribunal was not persuaded by the Applicant’s arguments and considered that the differences were amply reflected in the capital values of the property. It therefore determines that the deferment rate in the instant case will remain at 5%.”
As regards the evidence and argument advanced by Mr Hurst upon relativity the F-tT noted that Mr Hurst rejected various published graphs in favour of his own calculations and graphs. The F-tT rejected Mr Hurst’s analysis in the following passage in paragraph 9 of its decision:
“The Tribunal was not persuaded by his arguments and although acknowledging the imperfections in the RICS statistics prefers to follow the conventional analysis as reflected in many previous First-tier and Upper Tribunal decisions.”
The F-tT then considered the various published graphs and decided that the most relevant were the three graphs referred to in paragraph 10 of its decision.[11]We asked Mr Harper whether he contended that the F-tT had given clear and sufficient reasons for its decision in rejecting Mr Hurst’s evidence and argument upon the deferment rate and relativity. Mr Harper (correctly in our view) indicated that he did not wish to advance any such contention and he left it to us to decide whether the F-tT’s decision could stand. We conclude that the F-tT’s analysis of the evidence and arguments presented by Mr Hurst does not constitute clear and sufficient reasons for rejecting them. We therefore conclude that the F-tT’s decision cannot stand and it is necessary that the Upper Tribunal rehears the case upon the evidence now presented by the parties. Rehearing Stage - evidence and argument regarding deferment rate[12]On behalf of the appellant Mr Hurst has clearly given much thought and devoted much work to the preparation of his evidence and arguments in the present case, especially in relation to his analysis regarding relativity rates which we refer to later. Mr Hurst presented the appellant’s case with courtesy. His evidence and arguments were also notable for their frankness and modesty when he said in relation to the important decision of the Lands Tribunal in Earl Cadogan v Sportelli [2007] 1 EGLR 153 (LRA/50/2005) that he did not fully understand the Tribunal’s analysis (page 14 of his statement), which he expanded upon in his evidence so as to say that he was completely baffled by Sportelli and did not understand it.[13]In Sportelli the Upper Tribunal decided that the deferment rate could be expressed by the basic formula DR = RFR - RGR + RP (deferment rate equals risk-free rate minus real growth rate plus risk premium). Mr Hurst made clear that in the present case he did not feel able to comment on the risk free rate or the real growth rate and in consequence he just accepted them. However he observed that in the case of Shepherd’s Bush (and indeed the vast majority of England) the risk premium rate of 4.5% as adopted in Sportelli was in his opinion too low. It was upon the adoption of a 4.5% risk premium in the present case that Mr Hurst concentrated. He presented arguments and evidence that a higher risk premium rate should be adopted for the reasons he gave.[14]Mr Hurst drew attention to the recognition by Carnwath LJ in the Court of Appeal decision in Earl Cadogan v Sportelli [2007] EWCA Civ 1042 at para. 102 which recognised the possibility of evidence being called in other cases, in relation to the relevant property or the particular area in which the relevant property was situated, for the purpose of seeking to justify a deferment rate different from that adopted by the Lands Tribunal in Sportelli which was 4.75% for houses and 5% for flats.[15]As regards the risk premium this was described in paragraph 75 of the Lands Tribunal’s decision in Sportelli as a recognition of the risks of investment in long reversions, namely volatility, illiquidity, deterioration and obsolescence. Mr Hurst referred to Zuckerman v Trustees of the Calthorpe Estates [2009] UKUT 235 (LC) as an example of a case where the Upper Tribunal, upon evidence, had concluded that the deferment rate provisionally indicated for a property by the rates decided in Sportelli were in fact too low. A departure from these rates could be justified upon proper argument and evidence. It was just such argument and evidence which Mr Hurst contended that he was able to present in this appeal.[16]There were three particular topics of concern which Mr Hurst considered needed to be examined in the present case to see whether a departure from the Sportelli rate should be made, namely:(1) risk regarding the property cycle in Shepherd’s Bush as compared with parts of prime central London with which the Sportelli decision was concerned, which for convenience Mr Hurst referred to as Sportelli-land;(2) risk regarding what Mr Hurst described as the urban cycle in Shepherd’s Bush as compared with Sportelli-land;(3) risk of obsolescence in Shepherd’s Bush as compared with Sportelli-land.[17]In summary Mr Hurst’s evidence and arguments involved the following steps:(1) The open market value of the flat takes into account these three identified risks (see paragraph 16 above) but only for a limited time horizon, which Mr Hurst put at 11 years. Accordingly the purchase price does not make any allowance for these risks for years 12 to 72 (72 years being the unexpired term).(2) It is necessary to get a measure of the risk of holding an investment for 72 years as opposed to 11 years. This can be done by reference to the market in gilt-edged securities and shows that an additional 0.75% (approx) of risk is associated with the longer investment.(3) This 0.75% of additional risk can properly be divided into three equal parts to represent the three risks with which Mr Hurst is concerned. These three separate parts of risk are elements of risk within the Sportelli risk premium of 4.5%. Consequently the appropriate way forward is to deduct 0.75% (i.e. 3 x 0.25%) from the Sportelli risk premium of 4.5% so as to leave 3.75%. The question then arises as to how much to add back to this basic 3.75% to reflect these three categories of risk for years 12 to 72 for the flat (which is in Shepherd’s Bush) as opposed to a property in Sportelli-land.(4) As regards the property cycle Mr Hurst did not detect any additional risk for Shepherd’s Bush as opposed to Sportelli-land, so the 0.25% of risk is added back to the basic 3.75% without any alteration.(5) As regards the urban cycle (as to which see further below) Mr Hurst did detect twice as much risk for Shepherd’s Bush as opposed to Sportelli-land, so the figure to add back to the basic 3.75% is not 0.25% but is instead 0.5%.(6) As regards obsolescence Mr Hurst detected substantially more risk for the flat in Shepherd’s Bush as opposed to a flat in Sportelli-land, namely a risk that was 2.78 times as great, so that instead of adding back 0.25% for risk of obsolescence it was necessary to add back 0.695% (i.e. 0.25% x 2.78).(7) The total of all these adjustments, including the additional 0.25% allowed for in Sportelli in respect of the management of flats, lead to a risk premium of 5.445%.(8) Inserting this risk premium into the Sportelli formula (see paragraph 13 above) gives a deferment rate of 5.695%.[18]As regards Step (1) in paragraph 17 above Mr Hurst’s case was as follows. An owner occupier who purchases a dwelling is seeking the utility of occupation for, on average, 11 years. Their ownership of the dwelling will only last a small part of the property cycle. In consequence the allowance made by such a purchaser (which will be reflected in the purchase price) for risks concerning the long term social and financial stability of the area will be based upon more foreseeable factors, namely over an 11 year time horizon, as compared with the longer term factors which will influence an investor who is purchasing a reversion which will not confer vacant possession for 72 years. It was relevant to detect how much less risk the occupier purchaser (whose bid will decide what the open market value of the flat is) would be concerned about as compared with the purchaser of a 72 year reversion. Mr Hurst accepted that the open market value of a flat as at the valuation date would reflect risks such as he had identified in paragraph 16 above, but only over the first 11 years and not over years 12 to 72.[19]As regards step (2) in paragraph 17 above it becomes necessary to seek to quantify these risks with which Mr Hurst is concerned (see paragraph 16 above) so far as these risks are attributable not to the first 11 years of ownership (which is already reflected in the open market value) but to years 12 to 72. Mr Hurst considered that an appropriate way of analysing the additional risk for years 12 to 72 as compared with the first 11 years was to make reference to published figures regarding the yields upon gilt-edged securities. Figures are published for such securities which are stripped of any coupon or interest payment, such that the purchaser is paying £x in return for the guarantee that he will be paid at maturity £100. A comparison of the yield upon a purchase of an 11 year investment as compared with such yield upon a much longer investment would give a measure of the difference in risk between holding an 11 year investment and a much longer investment such as a 72 year investment.[20]Upon this point Mr Hurst recognised that while there were such stripped gilts which would mature in about 11 years there were no such gilts for so long a time period as 72 years. He therefore took the longest gilts available, namely those maturing in December 2055. He did however remark that he had been advised by those knowledgeable in such matters that there was a slight additional value (and therefore lower yield) attached to the very longest such gilts as compared with slightly shorter such gilts, because persons wanting an investment with a maturity as far distant as possible had effectively no choice but to accept the longest maturity on offer even though they might want a longer maturity.[21]Mr Hurst presented material from the United Kingdom Debt Management Office in his appendix 10 and drew attention to the yield percent for 4¼% Treasury Principal Strip 7 December 2055 (namely 3.635%) as compared with the yield on Treasury Coupon Strip 7 June 2024 (namely 2.886%), both figures being as at 2 October 2013. Mr Hurst then drew attention to the fact that there was a difference of 0.749% between these two yields. He considered that this figure of 0.749% represents the increase in risk which the long term investor in a reversion would detect over and above the risk elements already incorporated into open market values.[22]Mr Hurst concluded therefore that 0.749% (i.e. effectively 0.75%) was the element of the Sportelli-land 4.5% risk premium which represented risk from year 12 to year 72. The remaining 3.75% of the risk premium (i.e. the Sportelli figure of 4.5% minus this 0.75% as identified by Mr Hurst) was the risk premium for Sportelli-land with no allowance for the relevant risks between 12 and 72 years.[23]As regards step (3) in paragraph 17 above, it was necessary to attribute a proportion of this 0.75% to each of the three risks identified by Mr Hurst in paragraph 16 above. He concluded that the 0.75% should be divided equally so that 0.25% should be attributed to each of the relevant risks, although he recognised that he had no justification for saying that this 0.75% should be divided into three equal portions rather than divided in some other way.[24]As regards step (4) in paragraph 17 above Mr Hurst examined (his appendix 7) the graphs of average flat sale prices recorded by the Land Registry over the last 20 years as regards both Shepherd’s Bush and Sportelli-land. He recognised that those graphs show that in both locations there has been no down turn in flat values. Accordingly he could not conclude that flats in W12 suffered any more than flats in Sportelli-land from the risk of price fluctuations through the property cycle. Therefore as regards the 0.25% of risk attributed to the property cycle Mr Hurst concluded that no adjustment was needed. In consequence when calculating the risk premium for the flat in the present case the 0.25% for this element should be added back unaltered to the basic 3.75%.[25]As regards step (5) in paragraph 17 above Mr Hurst explained that concern regarding risk relating to the urban cycle arose in the following manner. He explained he was here, by the expression urban cycle, referring to the economic rise and fall and rise again of certain locations within cities. With the passing of time, accommodation may become less suitable for the purpose for which it was constructed such that the original occupier may move out and there may be more intensive forms of occupation leading to an increase of dilapidation and a general lowering of the quality of the area. He made reference to Booth’s Maps of London Poverty (his appendix 6) which were issued in 1889. These maps classified certain localities which would broadly include Sportelli-land in the top category of “Upper Middle Class and Upper Class Wealthy.” Mr Hurst concluded that in Sportelli-land there had been no social decline over the last 130 years and properties there had remained with the class of occupants for which they were built. As regards Ormiston Grove and its neighbouring streets these were shown in Booth’s Maps as being occupied by “Fairly Comfortable Good Ordinary Earners” thus not up to the affluence of Sportelli-land properties. Also, unlike Sportelli-land, Mr Hurst considered that Shepherd’s Bush had experienced an urban cycle. He summarised the history of the area from between the Wars (where there was a mixture of owner occupation, tenanted housing and social housing); into the 1950s (when the area was, as Mr Hurst put it, in a decadent state with sub-divided properties and landlords spending a minimum upon them); into the 1960s (where the image of the area was further dented by the television series Steptoe and Son); into the 1980s and 1990s (when the area started to improve with substantial investment from the Borough Council); and up to 2000 (by when professional people started to buy properties and gentrification was in full flow).[26]In short Mr Hurst contended that Sportelli-land had remained for over 120 years as a location for residencies of the highest class occupied by wealthy individuals whereas the Shepherd’s Bush area had originally been occupied by “Fairly Comfortable Good Ordinary Earners” but had been through a cycle where the quality of the area had greatly fallen and had then risen again. He argued that a potential investor in the 72 year reversion upon the flat would be concerned of the risk of this happening again - or more precisely would be concerned that there was a substantially greater risk of such a cycle happening again in this area as compared with the risk of such a falling in quality occurring in Sportelli-land. Mr Hurst therefore took the 0.25% which he had allocated to the risk of the urban cycle and he doubled this to 0.5% to reflect the greater risk from an adverse urban cycle in Shepherd's Bush as compared with Sportelli-land. Mr Hurst stated that he had no rationale for doubling 0.25% but that some upward adjustment must be made and as the period of the reversion (72 years) was in the order of half of the period over which this urban cycle had occurred in Shepherd’s Bush he considered that this supported a doubling of the 0.25% risk.[27]As regards step (6) in paragraph 17 above Mr Hurst advanced the following evidence and argument in support of a larger allowance for a risk of obsolescence in Shepherd’s Bush as compared with Sportelli-land. He pointed out that if one possesses an expensive item which is always worth repairing, i.e. where the cost of repairing the item will always represent value for money, then the item will always be repaired. But with an inexpensive item, where the cost of repairing it is more than the cost of buying a new one, the item will be thrown away rather than repaired. He contended that the ratio of the costs of maintenance of a building to the capital value of the building gives a measure of the risk of the building deteriorating and becoming obsolete. If the cost of maintenance is low and the capital value is high then the risk of deterioration and obsolescence will be less. In order to get a measure of the difference in risk on this topic between Sportelli-land and Ormiston Grove Mr Hurst examined the ratio of property value to rebuilding costs in Cadogan Square (as a representative of Sportelli-land) and in Ormiston Grove. Taking the sale of 46 Cadogan Square in November 2013 as an example Mr Hurst compared its sale price of £2,538,550 to its estimated rebuilding cost of £352,000 to give a ratio of value to rebuilding cost of 7.21. As regards the subject flat the value was £650,000 and the rebuilding cost estimated at £251,000 which gave a ratio of value to rebuilding cost of 2.59. He then divided 7.21 by 2.59 which gave 2.78 and he concluded that the risk of obsolescence in Ormiston Grove was 2.78 times higher than in Sportelli-land.[28]Consequently the 0.25% of risk attributable to obsolescence should be multiplied by 2.78 so as to reflect the difference in risk of obsolescence in Ormiston Grove as compared with Sportelli-land. This gave 0.695%.[29]Mr Hurst then put all of these matters together as follows. The basic risk premium after having deducted out the 0.75% was 3.75% (i.e. the 4.5% risk premium as identified in Sportelli minus this 0.75%). It was, as a result of the analysis as described above, necessary to add back 0.25% (i.e. an unchanged amount for property cycle) plus 0.5% (to reflect urban cycle risk) plus 0.695% (to reflect obsolescence risk). Accordingly the risk premium rate for the subject property would be 5.195% plus a further 0.25% to give 5.445% (this final 0.25% addition being made, as per the Sportelli case, to reflect the fact that the subject property is a flat rather than a house). Reverting then to the formula regarding deferment rate (see paragraph 13 above) the deferment rate of the subject property is 2.25% minus 2% plus 5.445%. This gives the deferment rate which Mr Hurst adopts of 5.695%.[30]On behalf of the respondents Mr Harper said that, whatever may be the shortcomings of the extent of the reasoning in the F-tT’s decision, the F-tT’s conclusion was correct. He said that he saw no reason to depart from the 5% deferment rate for flats as identified in Sportelli . He had been born and brought up in the Shepherd’s Bush area. It is true that there had been over the last 120 years or so a falling and then a rising again of the standard of the area. However the upturn had occurred and was now well entrenched. Mr Harper saw no significant risk of the cycle going into reverse. It may be that in Sportelli-land there had not been any down turn in the urban cycle for a long time, but there could now be detected risks arising from recent developments (and the possibility of further such developments) regarding stamp duty and the taxing of off-shore companies and the change of the profile of wealthy foreign purchasers who might wish to buy in prime central London. He saw no reason to allow more in the risk premium for the chance of an adverse effect from an urban cycle in Shepherd's Bush as compared with prime central London as dealt with in the Sportelli decision. Also Mr Harper said that he saw no reason to allow any extra risk for obsolescence in Shepherd’s Bush as compared to prime central London. Discussion regarding deferment rate[31]The appellant’s case regarding deferment rate rests upon an argument that, by reason of circumstances present in this appeal but not present in prime central London, a departure is appropriate from the deferment rate for flats of 5% as indicated by the guidance given by the Lands Tribunal in Sportelli . While Mr Hurst is to be commended for the frankness and modesty with which he informed us that he did not understand the decision in Sportelli , this is not a promising starting point from which to mount a reasoned argument that a departure in the present case should be made from the Sportelli rate.[32]We are unable to accept Mr Hurst’s arguments. We consider there to be flaws in several of the steps in his analysis, being the steps summarised in paragraph 17 above.[33]It is the starting point for his analysis, indeed it is his step (1), that there is a distinction between the perception of risk in the mind of an occupier purchaser as compared with that in the mind of an investor purchaser who is purchasing the 72 year reversion. However Mr Hurst’s analysis appears to assume that the purchaser of the reversion will be required to hold it, or will choose to hold it, for the full 72 years as compared with the occupier purchaser who will sell after 11 years. This is a false assumption. The purchaser of the reversion can sell the reversion. As recognised in paragraph 76 of Sportelli , tradeability would be an important component of owning the 72 year reversion. A reason why the volatility of the housing market and the relative liquidity of the investment are significant factors in the assessment of the risk premium is that the investor purchaser would be concerned to possess an asset which was tradeable. If the investment was to be held for the full 72 years these would be less significant factors. Also it could be said that it might be easier rather than more difficult for an investor purchaser of the reversion to sell this investment whenever he wished to do so (be it in 11 years or at some other time) as compared with an occupier purchaser who would have not only the task of selling but also the upheaval of removing to new accommodation.[34]It is an important aspect of Mr Hurst’s analysis that his identified risks are allowed for in the open market value of the flat but only up to year 11 and that the owner occupier who purchases the property will not be concerned (but the purchaser of the 72 year reversion will be concerned) with these risks for years 12 to 72. We do not agree with this analysis. An occupier purchaser who notionally has it in mind (in accordance with Mr Hurst’s argument) that he is purchasing for about 11 years of occupation will be concerned regarding the future prospects, viewed 11 years hence, for the property as this may affect the price he will obtain when in due course he sells. We do not accept that a purchaser occupier will be unconcerned regarding potential risks for the property after year 11.[35]Accordingly we disagree with Mr Hurst’s analysis at his step (1). If we were wrong about that, it next becomes necessary to consider his step (2) which involves the quantification of the additional risk for years 12 to 72.[36]We do not accept that some useful guidance as to the additional level of risk attributable to holding the freehold reversion for years 12 to 72, as compared with the risk for years 1 to 11, can be obtained through the examination of the market in gilt edged securities as performed by Mr Hurst. The investor in the freehold reversion will possess a freehold subject to a lease which will end in 72 years time. On the basis of those parts of the decision in Sportelli which are not challenged by Mr Hurst, regarding risk free rate and real growth rate, such an investor will have an investment which in principle should keep up with inflation and generate 2% growth (the risk of this not happening being allowed for in the risk premium). However the gilt edged investments referred to by Mr Hurst are investments which confer the prospect of getting precisely a certain number of pounds in x years time. There is effectively no risk of not getting that number of pounds. However presumably (although no evidence was called upon this) the yield required by someone paying £x now for a stripped gilt (i.e. with no income) which will pay a return of £100 in, say, 2055 is substantially driven by the view taken regarding long term interest rates and prospects of inflation. Neither of these two matters are of central importance to the purchaser of the 72 year reversion.[37]As regards Mr Hurst’s step (3) if, which we do not accept, it is appropriate to identify 0.75% of the 4.5% risk premium (i.e. as identified in Sportelli ) as reflecting the risks attributable to the years 12 to 72, there are the following further problems.[38]First, as Mr Hurst frankly accepted, there is no rationale for dividing this 0.75% equally between the three categories of risk identified by him.[39]Next, as regards the urban cycle, Mr Hurst had no (or no sustainable) justification for multiplying the 0.25% of risk by 2. Also we accept Mr Harper’s evidence that the Shepherd’s Bush area, while it may have had a dip in the middle of the last century, is now a much improved area. There is no reason for an investor to detect any extra risk (as compared with the like risk for Sportelli-land) that this area of London will deteriorate between years 12 and 72.[40]As regards Mr Hurst’s step (6) regarding obsolescence, there is no evidence before us to suggest that an investor would be more concerned about risk of obsolescence for this flat in Ormiston Grove than for a property in prime central London, or Sportelli-land as Mr Hurst described it. The flat comprises the upper part of a Victorian house which is divided into two units. Nothing has been drawn to our attention regarding the house itself or the area in which it is situated to indicate any additional concern regarding obsolescence. The house has been there for over 100 years. It is not the type of building which suggests some particular cause for concern regarding deterioration or obsolescence.[41]As regards Mr Hurst’s calculation, by reference to a ratio between property value and re-building costs (see paragraph 27 above) we can see that if a property is in an area where maintenance expenditure is not reflected in a stable or rising price, then there may be an additional risk of deterioration and obsolescence (although so far as concerns the investor in the freehold reversion there would be the ability to enforce service charges to raise money to keep the property in accordance with the covenants). However in the present case, as shown by Mr Hurst’s figures, value for money is obtained by spending money in building or rebuilding a property in Ormiston Grove. Adopting Mr Hurst’s figures in his appendix 9, £1 spent on (re)building in Sportelli-land can be said to generate £7.21 of value, whereas £1 spent on (re)building in Ormiston Grove generates £2.59 of value. Accordingly in both areas (re)building work is well worth doing. The risk of necessary (re)building work not being done and the property being allowed to deteriorate and become obsolete is not significantly greater in Ormiston Grove as compared with Sportelli-land. It is a fallacy to say that the risk is 2.78 times as great in Ormiston Grove. In both locations self interest will more or less equally ensure that the property does not deteriorate or become obsolete.[42]For all of those reasons we are unable to accept Mr Hurst’s arguments regarding the deferment rate. We see no reason to depart from the 5% deferment rate as indicated by the Lands Tribunal in Sportelli and as supported by Mr Harper in the present case and as adopted by the F-tT in its decision. Rehearing - evidence and argument regarding relativity[43]For the appellant Mr Hurst rejected the F-tT’s reliance upon the relativity graphs contained in the RICS Research Report “Leasehold Reform: Graphs of Relativity” (October 2009) for four reasons:(i) the data upon which they were based had not been fully disclosed;(ii) the data samples used were unlikely to be random;(iii) the method of construction of the graphs was not explained; and(iv) the statistical confidence limits of the data had not been disclosed.[44]Mr Hurst did not feel that the published graphs reflected “the relativity that I note” in the locality of the appeal property and he did not think that purchasers would refer to them. A mortgage valuer might take account of them but as long as the lender was prepared to loan the required percentage of the price the purchaser would use their savings to make up the difference. A mortgage valuation could form part of the reasoning about relativity but it facilitated the sale rather than quantified the price. There was a risk that the use of published relativity graphs, despite their well publicised faults, would become a self-perpetuating fallacy. Mr Hurst preferred an approach based upon an objective analysis of local sales data. He had therefore produced his own relativity graph for a subset of the London W12 postcode area. In summary Mr Hurst proceeded as follows:(i) He used the Land Registry website to collect details of all the sales of flats in Ormiston Grove and four adjacent residential streets between 1995 to the present.(ii) He excluded from his analysis any flat which had not been sold at least twice since 1 January 1995 or where the lessees owned a share of the freehold or where it was not possible to determine the date of the lease, its term and the term commencement date.(iii) He adjusted each transaction for time to the valuation date (2 October 2013) using two methods: (a) The Land Registry’s online price calculator; and (b) His own bespoke price index created by calculating the running average monthly price of flat sales in the London W12 postcode area and calibrating them to the valuation date which had an average value of £382,596 and an index of 1. Thus the index for July 2006 which had a running average price of £260,836 was £382,596/£260,836 = 1.46681.(iv) The time adjusted data was then plotted on two graphs, one for each of the two indices used. The unexpired lease term (which ranged from 65 to 190 years) was plotted on the x axes and the price was plotted on the y axes.(v) Two “best-fit” lines were shown for each graph; one derived using non-linear regression and the other using linear regression.[45]Mr Hurst acknowledged that for each property differences between one sale price and the next might be due to one or more of the following factors:(i) an increase or decrease in flat values;(ii) an increase or decrease in the unexpired term of the lease;(iii) the deterioration or improvement of the flat;(iv) the deterioration or improvement of the neighbourhood; and(v) a sale at above or below market value for non-market related reasons.[46]By using price indexation to the valuation date based upon flat sales in the local area, Mr Hurst considered that factors (i) and (iv) had been addressed and “effectively removed” as influences. He made no adjustment for factor (v) because he did not have knowledge of the circumstances of each sale but he thought that sales at non-market values would be so infrequent as not to affect the results. Mr Hurst considered that by using many data points he had derived a best fit line where flats that had deteriorated between sales would cancel out those which had been improved. That just left factor (ii), changes in price due to the changes in the unexpired term of the lease, which Mr Hurst said “is the relativity curve”. He went on:
“Those graphs are effectively relativity graphs for the flats situated within the locality of this property.”
[47]Mr Hurst concluded that the graphs showed a nearly horizontal trend line across a range of unexpired lease terms which did not noticeably decline down to and including the unexpired term of 72 years in this appeal.[48]Mr Hurst referred to the decision of the Leasehold Valuation Tribunal (“LVT”) in Frances v Maher [2005] LON/NL/2469/04 in which the Tribunal found that the relativity of a flat in Oaklands Grove (a street adjacent to Ormiston Grove) with an unexpired term of 68 years was 94%. Mr Hurst considered that “for the subject property with 72 years unexpired and with the graphs that I have produced the relativity should be greater than 94%.” Mr Hurst supported this opinion by expressing the year’s purchase for 72 years at 5.6% (which we understand to be a rounded approximation of Mr Hurst’s deferment rate) as a percentage of the year’s purchase into perpetuity at 5.6%. The result was 98%. The rationale for this calculation was said to be that the purchaser of a long lease at a low ground rent was effectively acquiring a “huge profit rent” and that a purchaser would value that profit rent in the way he suggested. Therefore one could derive a relativity from the ratio of the year’s purchase for the length of the subject lease (72 years) and that of an equivalent freehold (perpetuity).[49]Mr Hurst said that “allowing for some margin of error” he thought that the appropriate relativity in this appeal was 96%. He did not consider that it was necessary to adjust the relativity of a lease with an unexpired term of 72 years for the benefit of the Act.[50]Mr Harper said that he had been unable to find relevant transactional evidence and had therefore relied upon the published graphs of relativity contained in the RICS Research Report. He relied on the three graphs that the F-tT had adopted, namely Cluttons’ Prime Central London Graph for Flats, Nesbitt & Co’s graph which comprised predominantly flats (90%) predominantly located in Greater London and the outer suburbs; and Andrew Pridell Associates Limited’s graph for flats located predominantly in the South East and suburban London. The relativity for an unexpired term of 72.167 years was 88%, 92.25% and 94% respectively, giving an average of 91.4%.[51]Before the F-tT Mr Harper said he had also relied upon the Beckett and Kay Mortgage-Dependant Graph (2013: second revision). But he now considered that in Shepherd’s Bush although the majority of residential property continued to be funded by a mortgage, nearly half of purchasers bought with cash. That being so Mr Harper no longer relied on the Beckett and Kay graph.[52]Mr Harper preferred the Cluttons’ graph to other prime central London graphs because it was comprised entirely from data about flats including those sold in Maida Vale which was close to the appeal property.[53]Mr Harper said that he always referred to the published relativity graphs as a starting point in negotiations. He said that everyone knew about them and used them. He acknowledged that no one graph was “on all fours” with the Shepherd’s Bush area that one had to do the best one could with what was available. He said “until something better comes along we have nothing else.” Using the graphs saved the parties to negotiations both time and money.[54]Mr Harper said that it was “virtually impossible” to make any allowance for the benefit of the Act in the real world and he had not made any adjustment in respect of it. Discussion regarding relativity[55]Mr Hurst clearly spent a lot of time and effort producing what he described as relativity graphs for a subset of the London W12 postcode area. But they are not relativity graphs and cannot be used to obtain the relativity of the lease in this appeal. The expression “relativity” is defined in paragraph 2.1 of the RICS Research Report:
“In the context of statutory valuations under leasehold reform legislation, “leasehold relativity” is the value of a dwelling held on an existing lease at any given unexpired term divided by the value of the same dwelling in possession to the freeholder, expressed as a percentage.”
[56]Mr Hurst produced data for sales of leases with a range of unexpired terms from 65 to 190 years but he produced no evidence of freehold values. Consequently the necessary comparison between leasehold and freehold values could not be, and was not, made. Mr Hurst’s graphs purport to show that there is very little variation in the time adjusted value of leases with unexpired terms in the selected range. They say nothing explicit about what the relativity should be; only that, by implication, the relativity - whatever it is - will be constant over the said range. Mr Hurst acknowledged, in answer to the Tribunal’s questions on this point, that the lack of any information about freehold values meant that his analysis was “in that sense … deficient.”[57]We do not find Mr Hurst’s analysis of flat transactions in the local area to be of assistance in determining the relativity and it is not necessary for us to comment upon it in detail. We confine our observations to two points: (i) Mr Hurst uses two indices to adjust the transactions for time. He said that he had no preference between them. In our opinion prices should not be indexed over long periods, in this case up to 20 years. In a very recent decision, London Sephardi Trust v John Lyon’s Charity [2015] UKUT 619 (LC) the Tribunal said at [92]:
“It is clear from the above authorities that the Tribunal does not generally favour the indexation of property prices over as long a period as that used by [the respondent’s valuer] (3 years and 7 months), particularly over periods where the market is volatile or showing extreme price movements.”
The problems of indexing over long periods were well illustrated by Mr Hurst’s analysis of three historic sales of the appeal property in 1995, 1998 and 2001. Using his index the time adjusted prices were, in round terms, £592,000, £629,000 and £589,000 respectively. Using the Land Registry online price calculator the figures were £649,000, £673,000 and £679,000 respectively. Not only are there substantial differences in the calculated values (£90,000 in one case) but also one index shows a decline in value as the unexpired term reduces while the other shows an increase. That type of discrepancy does not assist the Tribunal. (ii) Mr Hurst assumes that the large number of transactions in his analysis will ensure that the best fit regression line will take account of differences in the state of repair of individual properties, with flats that have deteriorated cancelling out those that have been improved. He put forward a second argument on the point which was that if the tone of prices had increased over time as a result of the general improvement of the area then improvements to individual flats would be reflected in their time adjusted prices. We are not persuaded by this aspect of Mr Hurst’s evidence which at times we found difficult to follow. Neither argument addresses how, for any given pair of transactions of a flat, it is possible to tell whether the price change is due, at least in part, to the specific improvement or deterioration of the property. In time adjusted terms two sales of a flat 10 years apart may show a similar price; but that may be due to the effect of particular improvements (e.g. a new bathroom or kitchen) offsetting the effect of a shortening lease length. It seems to us necessary that a more specific analysis of the five factors identified by Mr Hurst (see paragraph 45 above), and a more detailed knowledge of the individual properties, would be required before the results could be considered robust and reliable.[58]Mr Hurst relies upon Frances v Maher in which the leasehold valuation tribunal determined that the relativity of a lease with an unexpired term of 67.91 years was 94%. In Arrowdale Limited v Coniston Court (North) Hove Limited [2007] RVR 39 the Lands Tribunal, George Bartlett QC and Mr N J Rose FRICS, said about leasehold valuation tribunal decisions as evidence at 45:
“37. …In our judgment leasehold valuation tribunal decisions on relativity are not inadmissible, but the mere percentage figure adopted in a particular case is of no evidential value. The reason for this is that each Tribunal decision is dependent on the evidence before it, and thus, in order to determine how much weight should be attached to the figure adopted in a decision, it would be necessary to investigate what evidence the leasehold valuation tribunal had before it and how it had treated it. Such a process of investigation is potentially lengthy, and it is inherently undesirable that leasehold valuation tribunal hearings should resolve themselves into rehearings of earlier determinations. 38. It is certainly understandable that valuers negotiating the settlement of an enfranchisement claim should have regard to leasehold valuation tribunal decisions on relativity, since these might seem to them to be the best guide of the likely outcome if they were unable to reach agreement, even though, as Mr Pridell said, the decisions are disparate and fail to show any established pattern. But the decisions themselves can constitute no useful evidence in subsequent proceedings.”
[59]For the reasons given in Arrowdale we place no weight on the leasehold valuation tribunal’s decision in Frances v Maher .[60]The third method for calculating the relativity used by Mr Hurst and described by him “as a further test” was the comparison of the ratio of the year’s purchase for 72 years at 5.6% to the year’s purchase in perpetuity at 5.6%. This gave a relativity of 98%.[61]The rationale that Mr Hurst gave for this approach was that the purchase of a lease on a nominal ground rent was, in principle, the acquisition of a substantial profit rent which could be capitalised at the deferment rate appropriate for the length of the lease. It was (presumably) not necessary to ascertain the level of the profit rent (assuming this remained constant) but only to compare the ratio of the year’s purchase for the unexpired term of the subject lease to the year’s purchase in perpetuity at the same rate. A similar, but somewhat more sophisticated, approach using dual rate year’s purchases was adopted by the appellant in Nailrile v Earl Cadogan [2009] RVR 95 at paragraphs151 to 164. That approach was rejected for the reasons given at paragraphs 202 to 209. Those reasons are relevant in the present appeal and three points are of particular importance:(i) In Nailrile the appellant developed a theoretical model of relativity which was then compared with an analysis of market transactions to see whether the valuer “could ‘find an echo’ in the sales data”. That seems to us to be a crucial requirement of any such model. If a valuer conjectures a theoretical solution to the problem of determining relativity, its predictions must be tested against market transactions to see whether the model gives results that are observed in practice. Mr Hurst, unlike the appellant’s valuer in Nailrile , made no attempt to test his model against empirical data. It was simply an untested conjecture and, as such, has no validity.(ii) Mr Hurst assumes that the rate at which to capitalise a hypothetical profit rent should be the deferment rate. In Sportelli the Tribunal said at [8]:
“Market evidence should be more readily available for [capitalisation rates], and in any event such rates, applying as they do to an element of static value, are determined by different criteria from those that are relevant to the deferment rate.”
In Nailrile the appellant’s valuer gave three reasons for his choice of a remunerative rate, which he described as a capitalisation rate and not a yield: “Firstly, it had to be instinctively plausible; secondly, it had to correspond within reasonable bounds to capitalisation rates found in the real world on other kinds of investment; and finally, he had to find an echo in the real world (in evidence of transactions).” (116 [154]). Mr Hurst made no such attempt to rationalise his choice of a remunerative rate; he simply adopted the deferment rate. In our opinion he was wrong to do so. (iii) Mr Hurst’s theory assumes that the notional profit rent is constant. In Nailrile this profit rent was described as “the annual flow of benefits” and it was held (at paragraph 207) that this would not necessarily remain constant as the length of lease reduced and the security of tenure declined. In our opinion Mr Hurst’s theoretical model is not reliable and we give it no weight.[62]We derive no assistance from Mr Hurst’s analysis of transactions, his reliance upon Frances v Maher or his theoretical model. We therefore consider Mr Harper’s use of graphs which he justifies by reference to Arrowdale at 45[39]:
“If no assistance is to be derived from earlier leasehold valuation tribunal decisions for the reasons we have just given, the same will go for settlements that have themselves been based on such decisions. In such circumstances, in our view, it is necessary for the tribunal to do the best it can with any evidence of transactions that can usefully be applied, even though such transactions take place in the real world rather than the no-Act world. Regard can also be had to graphs of relativity…”
[63]Before us Mr Harper relies upon the F-tT’s use of three of the graphs contained in the RICS Research Report: those produced by Cluttons, Nesbitt & Co and Andrew Pridell Associates Limited. He no longer relies upon Beckett and Kay’s mortgage-dependent graph since he now says that approximately half of the purchases in the locality of the appeal property are funded in cash. The effect of abandoning that graph is to increase the average relativity in favour of the appellant.[64]Both parties acknowledge that the appeal property is not in, but is close to, prime central London. Cluttons PCL Graph has the benefit of a separate graph for flats and covers geographical areas such as Maida Vale that are close to the appeal property. But the nature of the data is settlement evidence which the Tribunal said in Arrowdale was of no assistance. We have therefore compared in the following table the Cluttons PCL Graph with those other PCL graphs which were, at least in part, based on transactional evidence. Graph Relativity for 70 years unexpired term Relativity for 75 years unexpired term Comments Cluttons 86.2 89.65 W A Ellis 85 88 Houses only John D Wood/Gerald Eve 87 90 90% Houses Charles Boston 87 90 Flats and houses including some in Greater London Average excluding Cluttons 86.3 89.3 In our opinion it is reasonable and representative to use the Cluttons PCL graph when assessing the relativity of the appeal property.[65]Mr Harper also relies upon two graphs for properties outside prime central London. The Nesbitt & Co graph has the advantages of being based predominantly on data for flats (90%) in Greater London and the outer suburbs but the disadvantages of being based on settlements where the compiler has acted mainly for landlords (80%). The Andrew Pridell Associates’ graph is 100% based on data from flats and includes some transactional evidence. The compiler worked mainly for tenants (90%).[66]To help overcome the problem of the use of settlements in the construction of Nesbitt & Co’s graph we have also used the South East Leasehold graph which was rejected by the F-tT because it primarily comprised data from Bromley and Beckenham. But the South East Leasehold graph has the advantage of being based on transactions of flats with an even balance between acting for tenants and landlords.[67]The average value of the Nesbitt & Co, Andrew Pridell Associates and South East Leasehold graphs for an unexpired term of 72.167 years is 93.25%. The relativity derived from, the Cluttons PCL graph is 87.70%. Given that the appeal property is outside prime central London we consider it reasonable to produce a weighted average of these figures in the ratio of two (outer London) to one (PCL). This gives a weighted average relativity of 91.4% which is the same figure as that determined by the F-tT.[68]The parties accept that the relativity should have been applied to the agreed freehold vacant possession value of £653,250 rather than the agreed long leasehold value of £650,000. Neither party considered it to be appropriate or possible to make an allowance for the benefit of the Act for a lease of this length and, in the absence of any evidence to the contrary, we make no such allowance. Determination[69]We determine the deferment rate to be 5% and the relativity to be 91.4%. The appeal is therefore dismissed.[70]We determine that the premium payable by the tenant for the grant of a new lease shall be £37,545 (see APPENDIX 1 attached). Dated: 30 November 2015 His Honour Judge Huskinson A J Trott FRICS APPENDIX 1

Cited in 4 later judgments