Delivering affordable housing in the Vauxhall Nine Elms Opportunity Area

Transcription

1 Delivering affordable housing in the Vauxhall Nine Elms Opportunity Area Prepared for London Boroughs of Wandsworth, London Borough of Lambeth and Greater London Authority September 2014

2 Contents 1 Executive Summary 3 2 Introduction 5 3 Analysis of future affordable housing levels 10 4 Appraisal models and affordable housing delivery Current and alternative approaches to delivering affordable housing The role of review mechanisms Difficulties faced by developers and RPs in delivering housing in NEV OA NEV schemes and potential GLA investment Conclusions and recommendations 7069 Appendices Appendix 1 - Planning status of NEV OA sites Appendix 2 - Development appraisal year 1 Appendix 3 - Shared ownership income analysis Appendix 4 - Guaranteed income model assumptions Anthony Lee MRTPI MRICS Senior Director BNP Paribas Real Estate 5 Aldermanbury Square London EC2V 7BP anthony.lee@bnpparibas.com realestate.bnpparibas.co.uk BNP Paribas Real Estate

3 1 Executive Summary This report considers approaches to the delivery of affordable housing in the Nine Elms Vauxhall Opportunity Area ( NEV OA ) where at least 20,000 new homes and circa 500,000 square metres of commercial space are to be provided over the next twenty years. Developments in the area are typically high density and high specification, reflective of the target markets that scheme are aimed at. The volume of additional residential units and commercial floorspace in the area requires significant investment in new transport infrastructure, which developments will help to fund. This additional call on development value limits the amount of affordable housing that can be secured. Scheme viability is a crucial factor in determining the amounts of affordable housing that can be secured in developments in the NEV OA. It is vital that scheme viability is considered on an appropriate basis and, in particular, that benchmark land values are based on existing use values plus an appropriate and site-specific premium, not market values. Our viability testing on a sample scheme applies growth factors to sales values and inflation on costs as well as changes to other key variables. This analysis indicates that improvements in viability that have been seen provide opportunities to increase the levels and types of affordable housing delivered through the opportunities in the area. The analysis also identifies the significant opportunities and additionality that could be derived from exploring and developing options for off-site provision and gifting of land for affordable housing development. In combination higher values and development momentum and a more creative approach has the potential to assist in the delivery of affordable housing targets and indeed for the two boroughs to revisit affordable housing targets to provide greater clarity as to what can be delivered across the NEV OA. Appraisals are essentially snapshots of development proposals at a particular point in time and the results can change considerably as a result of changes to key inputs. Review mechanisms could help to reflect the dynamic nature of scheme viability by offering the councils an opportunity to reconsider viability over the course of a development. Developer concerns regarding the impact of reviews on fundability could be mitigated through capping the amounts that might be payable under viability reviews. Markets are inherently cyclical and counter-cyclical measures may require significant public funding in a high value area such as the NEV OA. However, affordable housing requirements can be changed during periods of recession to help address viability issues, with the quid pro quo being enhanced provision when markets recover. Measures could also include supporting developers who build out schemes for rent initially, with sale when the market recovers. This support could fund the additional costs of capital until the units are eventually sold. Delivering affordable housing in the NEV OA is complex and difficult due to the high density nature of planning developments. When combined with the councils affordability criteria, the amounts payable by Registered Providers are low, which squeezes viability. There is an inevitable tension between aiming to meet local housing need and delivering target percentages of affordable housing. As a result of this tension, compromises will often be required. Options to improve the proportion of units that can be secured as affordable include reviewing the mix of housing required, direct delivery by developers of some forms of low cost housing and provision of free land by developers to RPs and the Council in the lower value portions of their sites. As an alternative to on-site delivery, the councils could secure a significantly greater quantum of affordable housing by taking commuted sums and using 3

4 these to procure affordable units elsewhere. The analysis identifies that it is likely to be easier to deliver affordable rent and larger family homes through offsite provision in the quanta required to meet local demand. Off-site provision could include funding estate redevelopment programmes on the councils own land holdings and developers/landowners purchasing and indeed developing sites preferably in the local area to demonstrate a continued commitment to place making in the NEV OA. Flexible tenure arrangements offer a potential solution to mitigating risk which may enable RPs to offer more to developers for completed units, thus improving viability and the percentage of housing that is targeted to meet local demand. Whilst such arrangements do not have a particular national or regional policy locus flexible tenure arrangements, if accommodated, have the potential to secure a level of income and deal with risk and this report recommends that the GLA and the boroughs explore such options. Service charges are a significant issue in the NEV OA, both in terms of the impact on occupiers but also the impact on viability, as they are taken from the gross rents that RPs collect. Other issues associated with intermediate housing affordability could be resolved through different forms of development, including smaller space standards. 4

5 2 Introduction 2.1 Background The London Plan identifies the Nine Elms Vauxhall area ( NEV ) as an Opportunity Area ( OA ) of 195 hectares. The Vauxhall Nine Elms Battersea Opportunity Area Planning Framework ( VNEB OAPF ) produced by the Greater London Authority ( GLA ) with the support of Wandsworth and Lambeth Councils, indicated that the area had the capacity to deliver: 16,000 new homes; 300,000 square metres of employment space at Battersea Power Station (60,000 square metres of retail; 160,000 square metres of offices and 80,000 square metres of other employment uses); and 200,000 square metres of mixed employment use on other sites. The boundary of the NEV OA is shown in Figure 2.1. The 2013 Update to phasing of development and infrastructure requirements in the Nine Elms Vauxhall Area study indicates that the number of dwellings and non-residential floorspace will increase from the numbers anticipated in the OAPF. Developers/landowners anticipate delivering 18,276 residential units (an increase of 2,276 units) and 608,612 square metres of non-residential floorspace (an increase of 108,612 square metres). The draft Further Alterations to the London Plan (January 2014) increases the minimum number of residential units to be provided in the NEV OA to 20,000 units. Figure 2.1: NEV OA boundary Sites in the NEV OA are at various stages in the planning process. A full position statement on each site is attached as Appendix 1. This appendix also shows the details of each proposal, including a summary of the contents of the 5

6 Section 106 agreement. Table below summarises the schemes that have secured planning permission. Table 2.1.1: Schemes in NEV OA with planning permission Scheme Date of approval Number of residential units 81 Black Prince Road Allowed on appeal 15/9/09 10 Albert Embankment 17/2/14 No residential Hampton House 21/6/ Prince Consort House 16/10/ Tinworth Street and Vauxhall Walk AKA Spring Mews /8/12 No residential Lord Clyde, 90 Tyers Street 21/3/12 No residential Island Site Vauxhall Cross 20/8/ Bondway & 4-6 South Lambeth Place 18/4/11 No residential Vauxhall Square Cap Gemini 2/7/ Vauxhall Sky Gardens 239 Keybridge House 28/1/ Queensborough House 21/11/13 No residential Sainsbury s Nine Elms 31/5/ South Lambeth Road 1/9/12 No residential Eastbury House 1/9/12 48 Battersea Power Station 23/8/11 3,444 Battersea Gas Holders 18/1/ Marco Polo House 18/1/ Riverlight Deed of variation 15/12/13 Market Towers 487 Sleaford Industrial Estate/Dairy Crest 20/1/ US Embassy 12/10/10 No residential Embassy Gardens 16/2/12 1,978 Royal Mail Depot 5/3/12 1,870 New Covent Garden Market 25/10/ ,491 Total 12, Delivery of affordable housing in NEV OA While developments in the NEV OA typically attract high residential sales values, there is also pressure on development viability as a result of the need to fund new infrastructure to support the new communities that will be established 1 New planning application submitted on NCGM site on 22/5/2014 (Ref: 2014/2810) for 3,019 residential units 6

7 in the area. This includes the construction of an extension to the Northern Line ( NLE ) from Kennington through a new station adjacent to Sainsbury s on Wandsworth Road and onwards to a new station at Battersea Power Station. In addition, developments will need to contribute towards non-nle transport improvements, highways works, education, health and emergency services. Contributions from developments towards provision of this social infrastructure will reduce the amount of development value that could have potentially been used to deliver affordable housing. It is important to stress, however, that without this supporting social infrastructure, the developments would be unacceptable in planning terms and the number of new units would need to be scaled back considerably Affordable housing policy requirements in the NEV OA The two boroughs take very different approaches to seeking affordable housing contributions from developments in the NEV OA. Lambeth apply the same policy in the NEV OA as applies to other sites throughout the Borough, i.e. 40% of habitable rooms on sites capable of providing 10 or more units if no grant is available and 50% where grant is available. In Wandsworth, developments are required to provide a minimum of 15% affordable housing which is lower than the Council s policy for the rest of the Borough. Outside the NEV area of Wandsworth, schemes are required to provide a minimum of 33% affordable housing and up to 50%, subject to viability. However, even with this reduced quantum of affordable housing, the Development Infrastructure Funding Study ( DIFS ), which supports the OAPF, identified that the income generated by the tariff would not meet the funding requirements identified to support the OA. The schemes which were submitted for planning prior to 2014 were generating negative residual land values, indicating that adopting the 15% minimum affordable housing target has secured delivery that might not otherwise have been provided. The councils policies are supported by viability studies which were undertaken in The studies indicate that the policy targets will not be achievable on every site in the NEV OA. However, setting a target based on the least viable scenario would result in sites that could have provided more affordable housing not doing so. Revising the viability studies will not assist the councils in securing additional affordable housing. The councils need to maintain a firm position of how they will test viability to ensure that developers do not over-pay for sites, thus maximising the amount of value uplift generated by developments to provide affordable housing. The rationale for the lower affordable housing target in Wandsworth is set out in the DIFS. The DIFS explains that the financial contribution towards infrastructure will limit the extent to which affordable housing can be provided. The DIFS suggests that developer contributions will be an important source of funding to deliver the infrastructure that the NEV OA requires. Rather than basing the tariff on the costs to be funded, it determines the amounts that developments can viably support, reflecting other policy requirements (affordable housing, and Code for Sustainable Homes). The DIFS concludes that a tariff of 40,000 per unit is viable on riverside sites and 20,000 per unit in all other parts of the NEV OA, in combination with 15% affordable. The tariff would fall to 25,000 and 15,000 respectively if 40% affordable housing is provided. A higher tariff is proposed between 2016 to 2031, as shown in Table and

8 Table : recommended tariffs in DIFS: 15% affordable housing Land use tariff per unit tariff per sqm tariff per unit Residential (area 1) 40, , Residential (area 2) 20, , Office Mixed use retail Retail Hotel Table 4.1.2: recommended tariffs in DIFS: 40% affordable housing Land use tariff per unit tariff per sqm tariff per unit Residential (area 1) 25, , Residential (area 2) 15, , Office Mixed use retail Retail Hotel tariff per sqm tariff per sqm The tariff has been subsumed into Wandsworth s Community Infrastructure Levy ( CIL ) which the Council adopted in November Wandsworth have a commitment to reviewing their CIL Charging Schedule in 2015, including whether an increase in the level of charges in Nine Elms can be justified, in line with the DIFS. Lambeth have adopted a similar approach and have included a CIL rate commensurate with the DIFS tariff in their CIL charging schedule, which was examined in February Aims of this study The councils and the GLA have commissioned this study to assist them in decision making on the requirement for investment to support the delivery of affordable housing in the NEV OA to maximise delivery against targets. The report is structured as follows: Section 3 analyses potential future affordable housing levels and provides a set of metrics that can be used in negotiations with developers on individual developments; Section 4 examines development appraisal models; their role in negotiations on affordable housing levels; and explores their effectiveness and whether alternative approaches are required; Section 5 considers current and alternative approaches to the delivery of affordable housing in the NEV OA; Section 6 analyses the role that viability review mechanisms current play in maximising affordable housing in the area and how they could be used in the future; 8

9 Section 7 considers the role of Registered Providers and developers in the delivery of affordable housing and how this role might change to assist the councils and the GLA to maximise delivery of affordable housing; Section 8 considers how the GLA might direct some of its resources to delivery of schemes in the NEV OA; Section 9 sets out our conclusions and recommendations. 9

10 3 Analysis of future affordable housing levels In this section, we consider current forms of development being brought forward in the NEV OA and potential new opportunities that might arise in the future. We then consider how developers and the authorities establish appraisal parameters when establishing an appropriate and viable quantum of affordable housing on individual schemes. A key element of this negotiation is the land value against which a scheme residual is benchmarked. Finally, we consider how levels of affordable might change in the future. The two factors that the councils can seek to influence are the benchmark land value that developers use for their viability assessments and the price that Registered Providers can pay developers for completed units. With regards to the first factor, the greater the uplift from the benchmark land value to scheme value, the more scope there will be to deliver affordable housing. Or putting this another way, the less the developer pays the landowner, the greater the value available to fund affordable housing. With regards to the second factor, the price RPs can pay is directly linked to the tenure mix, affordability criteria and certainty as to future income streams. 3.1 Types of development in NEV OA Schemes that are currently on-site in the NEV OA are being developed by large national developers, including Berkeley Homes (Riverlight, Hampton House and Ebury House), Ballymore (Embassy Gardens), Telford Homes (Parliament House), Fraser Properties (Vauxhall Sky Gardens) and Special Purpose Vehicles (Battersea Power Station Development Company, which was formed by S P Setia, Sime Darby Employees Provident Fund; and Sainsbury s working in partnership with Barratt Development and L&Q to redevelop their store). The larger developments contain a mix of residential and commercial floorspace, including retail and leisure. The housing element is predominantly built for sale to owner occupiers and investors. There is a small amount of private rented sector ( PRS ) housing in the recent Christies Warehouse proposals include an element of PRS. Developments in the area are almost without exception high density, flatted schemes which will add significant pressure to infrastructure, as noted in the previous section. Many of the proposed developments that have already secured consent, or where proposals have recently been submitted for planning, include tall towers some of which exceed 30 storeys. This form of development presents challenges to the developer and selected RP in accommodating a mix of tenures, both in terms of physically accommodating separate cores and entrances and managing the impact of service charges. There is currently significant appetite for new development in the area, with the first phase of Battersea Power Station selling off plan within 1 week of the launch. Other sites have also sold very quickly, including Riverlight and Parliament House. Developers are prioritising sales to UK purchasers, but overseas sales are still important. Developers have not disclosed how many units have been sold to overseas buyers, but we understand that a significant proportion of units in schemes on-site have been sold overseas. How these units will be used by overseas investors remains to be determined; inevitably a majority of owners will rent out their properties while a minority will occupy them for part of the year or to an even lesser extent simply keep them empty and hold the properties as an investment and secure a return through capital growth rather than income. 10

11 There is strong pent-up demand from the domestic market which has recently been boosted by the Government s Help to Buy scheme, which helps many people who were previously excluded from the market by high deposit requirements. While this programme was initially intended to operate for a short period of time, the Treasury recently announced that it will be retained until at least This type of intervention is often difficult to wind back without having a significant impact on the market, so it may become a permanent feature of the housing market. However, it is not clear what impact this scheme will have in NEV. At present the maximum price of properties which can be bought through the scheme is currently 600,000 and many of the units marketed to date are priced above this figure. Affordable housing content of the schemes on site is at around 15% in Wandsworth and typically between 20% to 40% in Lambeth 2. Affordable housing is being delivered as a mix of affordable rent and shared ownership, both of which are being provided in line with the councils respective affordability criteria. 3.2 Potential new opportunities The majority of larger sites in the NEV OA have already secured planning permission from their respective planning authority, including inter alia Battersea Power Station, Ballymore Gardens, Riverlight, New Covent Garden Market (although a new planning application is to be submitted), Market Towers, Sainsbury s, Royal Mail, Vauxhall Sky Gardens, Sleaford Street, Marco Polo House, Hampton House, Parliament House, and Eastbury House. Most of the larger applications were in outline or hybrid and are likely to be re-visited which may affect the delivery of affordable housing. Given that market conditions have improved significantly since some of the larger permissions were granted, it is likely that any reconsideration of these schemes is more likely to result in an improved delivery of affordable housing than for there to be an adverse impact. The larger outline schemes also provide an opportunity for planning authorities to consider how they should most appropriately deliver affordable housing in a high value, high density areas. Many of the smaller sites do not yet have planning permission, so there is scope for new types of provision to be brought forward, subject of course to cooperation from the landowners. With regards to physical form, it is unlikely that developers will move away from the high density, flatted model that the consented schemes have established. However, there is clearly scope for the authorities to take a strategic approach across the remaining sites and potentially some of the consented schemes, which may be amended. For example, there may be potential for securing greater amounts of affordable housing by pooling requirements from higher value sites on some of the lower value parts of the NEV OA. 3.3 Approach to affordable rent In respect to the Affordable Rent tenure, both councils have undertaken assessments of the rent levels that would be affordable to current and future tenants, in the context of government reforms to the welfare system. The proposed rents are deemed to be the maximum that can reasonably be charged while having regard to the Local Housing Allowance and the Universal Credit (which will limit the total amount of benefits a household can claim to 26,000) and the incomes of households at whom the Affordable Rent tenure will be targeted. 2 Lower levels of affordable housing have been agreed, but normally where other benefits have been secured (e.g. the provision of a school at the Keybridge House site). 11

12 One important characteristic of the Affordable Rent tenure is that the rent must be fully inclusive of service charges. Therefore the sum of the rent and service charge must be within the levels of rent that the boroughs consider affordable in the context of the factors in the preceding paragraph. In high specification, high value developments, service charges can be substantial (more than 50 per week in some instances). This severely limits the rent that RPs are able to charge tenants while conforming to appropriate affordability levels. This has the potential of placing substantial downward pressure on the value that RPs can pay to developers for completed Affordable Rent units in new developments. For example, in LB Wandsworth, RPs are required to provide 50% of units at no more than 50% of market rent, with a blended percentage not exceeding 65% of market rents. The Council has set a maximum rent of 190 per week for a 2 bed Affordable Rent unit. If the service charge were disregarded, a 70 square metre (750 sq ft) flat might generate a capital value of circa 127,500 ( 1,830 per square metre/ 170 per sq ft). However, if a 35 per week service charge is levied, the capital value of the unit would fall to circa 97,500 ( 1,399 per square metre/ 130 per sq ft). For context, if the same flat were to be provided as a social rented unit at target rents its value would be in the region of 86,250 ( 1,238 per square metre/ 115 per sq ft). This example serves to demonstrate the difficulties that might be experienced in delivering Affordable Rent units in high value schemes where high service charges are likely. There is a similar impact on the value of shared ownership units, where the effect of service charges upon affordability will reduce the amount of initial equity that can be sold; or will reduce the percentage rent charged on the retained equity. Affordability criteria and service charges both have a significant impact on the amounts RPs can pay developers for new stock, with a consequential impact on the percentage of affordable housing that can be viably provided. In later sections, we consider how these issues might be resolved. 3.4 Metrics for negotiations on scheme viability Both councils policy requirements for affordable housing are framed as targets which are subject to site-specific circumstances, including the viability of development. When a developer is unable to meet the policy targets in full, the onus is upon them to demonstrate why the scheme cannot do so by submitting a Viability Appraisal with the planning application. Both councils procure external valuation advice to validate appraisals submitted by applicants and this process frequently results in a change in the level of affordable housing provision. There are various appraisal models available to test the viability of developments. These models all share similar characteristics and can produce results in different ways. The main options are as follows: A residual land value as an output; or Land cost as an input, with profit as the output. The various inputs to an appraisal are summarised in Figures and Clearly one of the additional key factors is time and flows of income and cost, which we consider later. 12

13 Figure 3.4.1: Appraisal model (residual land value) GROSS DEVELOPMENT VALUE ( GDV ) Private flat sales values Receipt from Registered Provider Car parking sales Ground rents Investment value of commercial floorspace A LESS DEVELOPMENT COSTS Base build costs Contingencies Professional fees Marketing Finance DEVELOPER S PROFIT B C RESIDUAL LAND VALUE = A (B + C) Figure 3.4.2: Appraisal model (profit as output) GDV Private flat sales values Receipt from Registered Provider Car parking sales Ground rents Investment value of commercial floorspace A LESS DEVELOPMENT COSTS Site value Base build costs Contingencies Professional fees Marketing Finance SITE VALUE B C PROFIT = A (B + C) Evidencing inputs to a development appraisal Developments have unique characteristics that should be reflected in the inputs to a development appraisal. For example, sales values of individual units will 13

14 be determined by aspect, location, height and internal specification, while build costs will be influenced by design, specification, ground conditions and so on. When preparing a development appraisal, a valuer would have regard to scheme-specific characteristics so that the result (in terms of residual land value) is reflective of these characteristics. Inputs to an appraisal reflect the current day situation and circumstances may change very quickly. For example, sales values can change in response to changes in demand (up or down) over short periods of time. Although the impact of changes to inputs can be tested through sensitivity analyses, the base position will always be rooted in today s market conditions. Evidence provided in support of a development appraisal therefore has a short shelf-life and any user of an appraisal should have regard to the need to collect new and updated evidence if the viability of a scheme is to be re-visited. When assessing the viability of a development proposal, we would expect to see the following evidence produced by the Applicant: Sales values: a unit-by-unit pricing schedule, showing how aspect, height, specification and location have been considered. This pricing schedule should be supported by an analysis of comparable sales within the vicinity (if schemes have recently been sold) or other relevant developments that share similar characteristics. Sales rates: the speed at which units in a development are sold is an important factor in determining viability. Off-plan sales which result in completion of a sale when a unit reaches practical completion will improve the overall cashflow profile of the development. In other words, the sooner a unit is sold, the sooner the developer receives payment and this reduces finance costs. We would expect to see a sales phasing plan that is reflective of the sales rates achieved by other comparable developments, having regard to site-specific factors that may result in a different rate of sale. Receipt from affordable housing Registered Provider ( RP ): developers will typically sell the affordable housing units to an RP which will take responsibility for selling equity stakes in shared ownership units and letting the rented units. We would expect the developer to either produce a valuation of the units, based on anticipated rental income and/or the value of equity stakes sold, or provide offers from RPs for the units available. Commercial floorspace: many schemes coming forward in the NEV area include an element of commercial floorspace, typically retail, office or leisure uses. Appraising the residual value of these elements is similar, except that the method for arriving at a capital value is based on capitalising the expected rental income. Developers therefore need to evidence both the rental income and also investment yields, both of which can be demonstrated through comparable lettings and investment sales. Build costs: a cost plan for the proposed development, reflecting schemespecific characteristics, including design, ground conditions, access issues and site constraints, below ground issues (such as Underground lines, services etc). We would typically assess the overall build costs against benchmark data and recommend that the costs are reviewed by a Cost Consultant if they are significantly higher than the benchmarks. Professional fees: developments typically require professional inputs from a group of specialists, ranging from design to rights of light advice. Schemes do not require an identical level of professional inputs, as they will vary in complexity. For example, the structural engineering input to a 40 14

15 storey tower scheme will clearly be greater than would be the case for a 4 storey building. Consequently, professional fees will lie within a range of 8% to 12% (possibly more in very exceptional circumstances). Developers will therefore need to demonstrate why the level of fees used in their appraisal is appropriate to the nature of the scheme under consideration. Developers would normally need to evidence professional fees by providing a breakdown of the total between the different disciplines. Marketing costs: marketing costs include the Selling Agent s fees, but also the cost of show homes, advertising, brochures and overseas marketing activities. Marketing costs typically account for 3% of GDV, but can sometimes be higher in exceptional circumstances. For example, on schemes being sold out over very long periods, the marketing home and other material may require updating and re-branding to reflect changes in customer requirements. Developers should provide a detailed schedule of how the marketing budget is divided between different costs if the total budget exceeds the normal 3% allowance. Finance costs: financing developments became more complex after 2008 due to funders tightening their criteria for lending. In particular, it is now uncommon for banks to fund the entire development cost and unless developers have access to their own equity, they will need to source the balance elsewhere (either through mezzanine finance or external equity). The cost of funds can vary in relation to the type of developer, their perceived longevity and their experience in the type of scheme they are seeking funding for. Funds may also vary in relation to the type of development, with more complex schemes with lengthy build out periods perhaps attracting higher funding costs than simpler schemes. However, the market accepts a blended 7% finance rate (inclusive of arrangement and exit fees). Where a higher rate is used, we would expect to see evidence in the form of funding agreements. Development profit: profits are to an extent scheme-specific but also must have regard to the general stance adopted by banks who might fund the development. Profits have generally been at around 20% of private housing GDV since 2008, with a reduced profit on the affordable housing (6% of cost). The primary purposes of profit are to enable the developer to secure a return on capital and to mitigate against risk (i.e. that the sales values anticipated in the appraisal are not achieved). Sales risk on the affordable housing is low in the NEV area, as there is strong demand from some RPs for new stock. Where a developer seeks to include higher profit levels in an appraisal, we would expect to see a robust rationale in terms of the risks associated with the scheme relative to other comparable schemes in the area Current and future values in the NEV area Chart below summarises the sales values predicted in various schemes in the NEV OA that BNP Paribas Real Estate have assessed on behalf of the two authorities. It is important to stress that the values shown are scheme averages and that there is likely to be a significant variation between per square metre values for units in different parts of a building and/or with different aspects. The councils can update sales values by applying forecasts from the leading agents, but these are of course based on sets of assumptions that may or may not pan out in practice. It is impossible to accurately forecast the future trajectory of sales values with any degree of certainty. We also note that the leading agents do not attempt to forecast values beyond a five year timeframe, underlining the point that long term forecasting is inherently unreliable. 15

16 Table : Values estimated on NEV OA schemes at time of planning application Scheme Borough Date of assessment Vauxhall Bondway 1 Lambeth High St 8 Albert Embankment Sleaford Street Average market values (per sq m) Average affordable rented values (per sqm) Average shared ownership values (per sqm) Lambeth Mar ,099 1,615 2,691 Lambeth Feb ,077 2,799 4,252 Lambeth Feb ,000 to 10,721 2,045 3,046 Wandsworth Jan ,650 1,345 2,960 Riverlight Wandsworth Sept ,216 n/a 3,518 Prince Consort House Vauxhall Sky Gdns Hampton House Lambeth Aug ,683 n/a 3,724 Lambeth July ,690 1,049 3,077 Lambeth Feb ,345 n/a 2,831 Sainsbury s Lambeth Dec ,221 to 7,825 Vauxhall Square 2,378 3,283 Lambeth Nov ,367 1,615 2,637 NGCM Wandsworth May ,427 to 11,722 Market Towers 1,432 2,616 Wandsworth May ,208 n/a 2,691 Royal Mail Wandsworth Feb ,641 to 8,385 Embassy Gdns 915 (note 1) 2,960 Wandsworth Feb ,707 n/a 3,326 Marco Pollo Wandsworth Jan ,086 n/a 2,411 BPS Wandsworth Sep ,988 to 13,918 1,292 2,691 Table summarises the 5 year forecasts of sales values growth from each of the leading agents. This shows a degree of divergence of opinion in terms of the future housing market trajectory in London. Table provides pessimistic and best case scenarios over the next five years based on the forecasts. 16

17 Table : House price forecasts (nominal annual growth) Agent Savills 3 4.5% -0.5% 7.0% 7.0% 4.5% JLL 4 8.0% 8.5% 8.0% 7.0% 5.5% Knight Frank 5 4.0% 0.0% 5.0% 5.0% 5.0% Cluttons % 7.0% 4.0% 3.0% 2.0% Strutt & 6.0% 2.0% 7.0% 7.0% n/a Parker 7 Table : Pessimistic and best case forecasts (nominal annual growth) Pessimistic 4.0% -0.5% 4.0% 2.5% 2.5% Best case 11.0% 8.5% 8.0% 7% 5.5% Clearly some of these scenarios consider the London market as a whole, or in other cases, prime central markets, and are not specifically based on the NEV OA area. The most appropriate method for updating information on sales values is for the councils to periodically collect information on achieved sales values (unit and per square metre values) on various schemes in the NEV OA. Ideally, this information should also include factors that are likely to impact on values, including height, specific location, aspect and specification. Forecasting affordable housing values is to an extent more difficult than attempting to forecast private sales values. There has been considerable flux in the arrangements for delivering affordable housing since 2010, which impact on rent levels; the degree of variability of rent levels; affordability criteria; operating costs; funding costs; void levels; and RPs attitude towards risk. Furthermore, there is continued uncertainty with regards to the levels and application of grant funding that may be applicable to certain types of development Current build costs in the NEV area Schemes in the NEV OA tend to be unique in terms of their design and built form and this typically means that costs will vary (to some degree) between different schemes. When assessing development appraisals on behalf of the councils, we typically compare the build costs on a per square metre basis to benchmark data from the RICS Building Cost Information Service ( BCIS ). Where costs exceed the RICS benchmarks, we recommend that the councils procure specialist advice from a cost consultant. In terms of updating data on build costs, it is possible to apply one of the BCIS cost indices to the base data, which currently forecast five years into the future. This would provide an indication of the likely change in costs, but the optimum way of determining up to date costs would be to require a developer to commission their cost consultant to update their cost plan. This would then need to be reviewed by the Council s cost consultant. 3 Prime London. Equivalent mainstream London forecast is as follows: 2014: 8.5%, 2015: 6.5%, 2016: 4.0%, 2017: 2.0% and 2018: 2.0% 4 Central London Developments forecast. Same forecast applies to Prime Central London 5 Prime central London forecast 6 Greater London forecast. No forecast by year for prime London markets 7 Prime central London 17

18 3.4.4 Target profit levels Target profit levels and how profitability is measured will vary to some degree between different types of scheme and to a lesser extent between developers. Developers are likely to adopt a different approach for schemes which are developed over long periods of time (i.e. ten years or more) than they would for schemes developed over relatively shorter periods (i.e. the typical development period of two to three years). For the latter type of scheme, developers are likely to adopt a profit on value or cost approach, where they simply target a percentage of either scheme value or scheme cost. For long term schemes, there is also a need to reflect the time over which the return is received, so these developers will target a profit based on an Internal Rate of Return ( IRR ). The IRR is simply the discount rate required to achieve a Net Present Value of zero. IRRs can be geared or ungeared meaning that the cost of funds is either included in excluded respectively. At the current time, developers are targeting profits on private housing of 20% of value, which broadly equates to 25% of costs. For larger schemes, developers who have recently secured planning permission for schemes in and around the NEV OA have proceeded with development on the basis of IRRs of 12% to 13%. This also reflects research by Gerald Eve in its viability study for the City of London s Community Infrastructure Levy. Prior to 2008, developers and funders were satisfied to proceed with a development on the assumption of a developer s profit in the region of 15% of private housing GDV, or sometimes even lower. As the recovery gathers strength, it is possible that developer s profits could fall, but it is difficult to predict when this might happen. A lot will depend on the attitudes of banks, which are still working through significant stocks of toxic debts, many of which are associated with property development. It may take an additional five years for the banks to address all these outstanding debts, after which time they may expand their lending on new residential development. With regards to affordable housing, most developers accept that the risk associated with a disposal to an RP is lower than the risk of selling individual private units. Given that the developer will forward sell the affordable housing in a single transaction and that the RP will make staged payments over the construction period, a profit of 6% on cost is generally accepted Development Finance Following the last housing market peak in 2007, funders significantly reduced their exposure to residential development. As a result, they reduced the proportion of overall scheme costs that they would fund to no more than 50%. Funders are typically charging a margin of 6% to 6.5% above the base rate, with cost of funds to developers of around 6.5% to 7%. Developers need to provide their own funds for the balance of the development costs. This may be sourced from internal funding (i.e. retained profit) or through mezzanine finance or equity. These sources of funds attract varying costs, but they may exceed the 7% charged by the banks. Banks may reduce their margins as the risks associated with residential development recede. However, that is likely to be accompanied by an improvement in general economic conditions and an increase in the Bank of England base rate. So whilst the banks margins might fall over time, this is likely to be off-set by an increasing base rate. The net result will be that finance rates are unlikely to fall much below the current 7%. Funders currently require developers to achieve off-plan sales of as much as 30% of units prior to allowing developers to draw down loan funding. This 18

19 means that developers will need to fund the initial costs through equity or mezzanine funding, or from their own internal resources. Over time, it is likely that banks will relax their criteria and remove the pre-condition for off-plan sales. In the NEV OA, many schemes involve tall or large buildings which are very difficult to phase once construction is commenced, it is not possible to stop building and occupy the part that has been completed. This has implications for cashflow there is a long period during which the developer has to commit resources without receiving any income. This introduces a dimension of risk that is unique to these schemes and may impact on a developer s ability to fund such schemes. Terms of finance may therefore differ to reflect the perceived or actual cashflow risks. 3.5 Appropriate benchmark land values and viability The residual land value of a scheme is one half of the equation when testing its viability and ability of deliver affordable housing. The other half of the equation is the benchmark or threshold land value, i.e. the value that will be sufficient to bring the site forward for development. There has been considerable debate over the past few years on what constitutes an appropriate benchmark land value, which to an extent has not been assisted by the National Planning Policy Framework ( NPPF ) which talks in general terms about competitive returns to landowners. The notion of a competitive return is clearly open to considerable variation in interpretation. There are broadly speaking two approaches to determining an appropriate benchmark land value for testing the viability of a development: Existing Use Value plus an appropriate uplift, or an Alternative Use Value; which are the preferred and recommended approaches in both boroughs and Market Value Existing Use Value plus appropriate uplift, or Alternative Use Value An existing use value is literally the value of the site assuming that the existing use continues and there is no change of use or redevelopment. In other words, it reflects the current situation with regards to the income that the existing building generates (or does not generate). If the building were to remain in its existing use, the landowner could continue to receive the income for as long as demand for the building remains. If the existing building is not currently let and there is doubtful future demand, clearly the landowner will be keener to release the site for development to avoid the burden of keeping the building empty (including empty rates, insurance costs, security costs etc). If a development proposal fails to generate a residual land value that at least exceeds the existing use value, then it will not come forward, as the landowner would be better off retaining the existing building and receiving the rental income. If the proposal generates a residual land value that is the same as the existing use value, then the decision to sell will be based on the landowner s assessment of likely future demand for the building, which will clearly include the age and facilities of the building in comparison to others, as well as demand for the particular type of space in the location. It is therefore important to consider the extent to which a premium above existing use value is required to incentivise a sale for development. This premium is likely to range from zero to as much as 30%, but should always be based on site-specific factors and characteristics. 19

20 Taking the existing use value plus an appropriate premium is an objective floor below which the residual land value of a scheme cannot fall if it is to be viable. It is readily understood and can be easily measured and tested. A variant to existing use value is to consider an alternative use value (i.e. a scheme that the landowner might consider in place of a residential scheme). For example, rather than selling for residential development, the landowner could sell the site for a hotel. Alternative use values may be valid providing the proposed use would be acceptable in planning terms and also that the appraisal fully reflects any policy requirements that would attach to such a scheme. Adding an incentive to an alternative use value would be inappropriate, as the landowner does not currently have the benefit of the income that derives from this use; it needs to be considered alongside other competing alternatives that would also require planning consent Market Value Some members of the RICS advocate the use of Market Value as a benchmark land value when testing viability of development proposals. The main difficulty with this measure is the extent to which individual valuers interpret the meaning of this term and how it should be applied to individual sites. The RICS has published guidance on testing viability 8 which at paragraph takes as its starting point the existing (or current) use value of a site: For a development to be financially viable, any uplift from current use value to residual land value that arises when planning permission is granted should be able to meet the cost of planning obligations while ensuring an appropriate Site Value for the landowner and a market risk adjusted return to the developer in delivering that project (the NPPF refers to this as competitive returns respectively). The return to the landowner will be in the form of a land value in excess of current use value. The Guidance goes on to state that it would be inappropriate to assume an uplift based on set percentages given the diversity of individual development sites and that the uplift from current use value should be determined by market value which would include market/transactional evidence. There are considerable dangers in relying upon transactional data, as follows: Transactions are often based on bids that take a view on squeezing planning policy requirements below target levels. This results in prices paid being too high to allow for policy targets to be met. If these transactions are used to set a benchmark land value, the outcome would be unreliable and potentially highly misleading. Historic transactions of housing sites are often based on the receipt of grant funding, which is no longer available. There would be a need to determine whether the developer who built out the comparator sites actually achieved a profit at the equivalent level to the profit adopted in the viability testing. If the developer achieved a suboptimal level of profit, then any benchmarking using these transactions would produce unreliable and misleading results. Developers often build assumptions of growth in sales values into their appraisals, which provides a higher gross development value than would actually be achieved today. Given that viability appraisals are based on 8 Viability in Planning: Advice to Practitioners August

21 current values, using prices paid would result in an inconsistent comparison (i.e. current values against the developer s assumed future values). Using these transactions would produce unreliable and misleading results Impact of benchmark on affordable housing secured Ultimately, the choice of benchmark will be a key determinant in the level of affordable housing that a scheme can provide. The higher the benchmark land value, the lower the amount of value uplift arising from development that the councils can secure to secure for affordable housing provision. This is illustrated in Figure below. The chart compares the benchmark land value (which is represented by the blue bar) to the residual land value generated by two schemes, one of which provides a policy compliant level of affordable housing (the red bar) and one providing a reduced proportion of affordable housing (the green bar). Figure : Impact of benchmark land value on affordable housing In the chart, the existing use value plus margin is shown as the dark blue segment of the bar. If this is adopted as the benchmark land value, then a policy compliant scheme (represented by the red bar) is sufficiently high for the scheme to come forward. However, if the Market Value is adopted as the benchmark (represented by the combined total of the dark and light blue segments), then the policy compliant scheme is no longer viable. This would force the planning authority to agree a scheme with reduced levels of affordable housing (represented by the green bar). Every scheme generates a finite amount of value that must be divided between the various interests (the landowner, developer and local planning authority). In almost all cases, the Developer s interests are secured through inclusion of a profit in the development appraisal. The planning system must therefore mediate between the public interest (in terms of affordable housing and other obligations) and private interests (in terms of land value to the developer). Under the market value approach advocated by the RICS, the bulk of any uplift in land value goes to the landowner, by necessarily squeezing the amount of affordable housing provided. Under the existing use value plus premium approach, the uplift in land value is split between the landowner and the public interest. 21