Guide 99/43

PFI case study

The re-development of the Cruciform Building

University College London

Contents

Foreword
Executive summary
Background
Procurement process
The agreement
Value for money and risk transfer
Lessons learned
Acronyms and abbreviations

Foreword

This case study was produced with the assistance of Jack Foster, Director of Finance, and Allan Mitchell, Assistant Director of Estates and Facilities, both of University College London, and Sheldon Taylor of Ernst and Young.

Readers should note that the redevelopment of the Cruciform Building was a response to the 1992 Tomlinson Inquiry, and the project received directly targeted public funds. As a result, both the Department for Education and Employment and the HEFCE became involved in the approval process. This is not standard practice for PFI projects in the higher education sector: higher education institutions are normally free to enter into contractual agreements without further consultation.

Executive summary

1. The PFI project undertaken by University College London reached financial close on 5 December 1997. Its origins lay in the 1992 Tomlinson Inquiry, which led to the creation of an enlarged medical school and a need for additional teaching and research space. The 90-year-old Cruciform Building in nearby Gower Street, designed by Sir Alfred Waterhouse, offered an ideal location.

2. The Cruciform development did not begin life as a PFI project. Having identified an opportunity for PFI to add value to the development, the college moved forward to financial close in little more than a year. The data bank built up prior to this decision undoubtedly speeded up the process.

3. With its advisers, the college short-listed six consortia for the project. The selected consortium consisted of Jarvis Construction (UK) Ltd, Jarvis (Facilities Management) Ltd, Rotch Property Group Ltd and a lender (later Newcourt Capital Inc. of Canada). Following wide-ranging negotiations, a full business case was submitted in early May 1997, and approval received in July.

4. In common with most PFI transactions, the project agreement between the private sector and the college established the main parameters of the scheme. However, the project’s complexity and the variety of interests involved produced a number of supporting documents, particularly with termination issues.

5. The college believes that the project offers a number of lessons to higher education institutions contemplating similar schemes. Extensive pre-planning allowed it to minimise the effect of the unforeseen, while stark deadlines drove negotiations at a rapid pace. Above all, by demonstrating genuine commitment, both to the preferred partner (with whom the college had a strong existing relationship) and to the project itself, the college lessened the chances of stagnation, and was able to resolve the problems that presented themselves along the way.

Background

The Private Finance Initiative in higher education

6. The Private Finance Initiative (PFI) is a procurement methodology with three main features:

• it focuses on the procurement of a service, rather than the provision of the assets required to deliver the service
• the risks associated with the project are borne by whichever party to the deal can manage them most effectively, so as to achieve best value for money
• the transfer of risk is reflected in the payment method.

7. The Higher Education Funding Council for England (HEFCE) expects higher education institutions to base their investment decisions on seeking best value for money, and encourages them to consider PFI, and other innovative forms of procurement, as ways of achieving this. However, the HEFCE recognises the status of institutions as independent bodies, and there is normally no requirement on them to test for a PFI solution. This reflects current Government policy. Furthermore, there is normally no requirement on institutions to seek approval from the HEFCE, the Department for Education and Employment (DfEE) or the Treasury.

8. However, when this particular project was undertaken there was a requirement to test PFI because, as a Tomlinson project, it was to receive specific public funding.

University College London

9. Founded in 1825, University College London (UCL), the original University of London, is the oldest higher education institution in England after Oxford and Cambridge. With an annual turnover of £320 million, 15,000 students and a staff of 6,000, it is now one of the UK’s largest institutions and a major centre of excellence for teaching and research.

10. The Medical School’s tradition of achievement and innovation goes back to its foundation in 1828, as the first university faculty in London with a hospital for teaching purposes. This success is continuously enhanced by outstanding academics – including three Nobel Laureates – and advances in research and teaching. The present development of the Cruciform Building, previously University College Hospital, is set to extend this distinguished reputation. It will accommodate the Wolfson Institute for Biomedical Research, allow some 300 scientists to develop novel treatments for chronic degenerative diseases, and provide state of the art facilities for training the clinicians of the future.

The needs

11. The merger of the UCL Medical School, the Royal Free Hospital School of Medicine [RFHSM] and the postgraduate Institutes of Child Health, Neurology and Ophthalmology (and latterly the Eastman Dental Institute), followed the 1992 Tomlinson Inquiry. The mergers would result in the creation of an enlarged medical school within the college, the Royal Free and University College London Medical School.

12. As well as teaching accommodation, the college needed additional research space. It had previously identified a need for an additional 25,000 m2 of space, of which approximately 10,000 m2 would be required for the merging institutions and for planned expansion of research. The merger process began with the submission of an initial case for capital funding in November 1994. This demonstrated the need for new teaching facilities for the combined medical course, and the financial viability of a new research institute of approximately 300 staff. The college refined and amended the submission in discussion with the HEFCE and submitted an outline business case in October 1995.

13. The college’s strategic objectives outlined in the October 1995 business case were to:

14. Good progress was made and formal mergers accomplished with the Institute of Child Health, Institute of Neurology and Institute of Ophthalmology by 1 August 1997. However, the most significant of the merger proposals, with the RFHMS, depended on the successful outcome of the Cruciform Building redevelopment.

Early steps

15. An important part of the college’s PFI process was the way in which it progressed the early stages of this project under a traditional tendering route. This brought significant benefits to the PFI negotiations.

16. The college established a working group consisting of the Provost (the principal academic and administrative officer of the college), the Dean of UCL Medical School, the Dean of the Royal Free Hospital School of Medicine, the Dean of the UCL Faculty of Life Sciences, and Professor Salvador Moncada, who would be joining UCL to lead the new research initiative. They were joined by the Director of Planning and Resources, the Director of Financial Accounting and Mergers, and the Assistant Bursar.

17. Though the group considered a range of options, the preferred route was to purchase and redevelop the Cruciform Building, previously occupied by University College Hospital - the only site available and capable of meeting the requirements. Additionally, its Gower Street location was ideal. The college proposed to redevelop the Cruciform Building completely to provide 8,750 m2 net of teaching and research space while leaving the facade largely unchanged.

18. A Grade II listed building designed by Sir Alfred Waterhouse and completed in 1906, the Cruciform Building had undergone many alterations, mostly subject to ‘Crown Immunity’ and thus avoiding planning or listed building criteria. English Heritage indicated that it would be necessary to repair and overhaul the exterior of the building and remove some of the non-conforming additions so that the building was visually more in keeping with its origins. There was also a need to upgrade the main fabric and core services of the Cruciform Building, and undertake alterations to the existing buildings of the two medical schools at UCL, the Royal Free Hospital and Whittington Hospital sites. As capital values were low, these latter alterations were eventually undertaken through conventional procurement.

19. The college’s initial bid for HEFCE funding under Tomlinson was made in November 1994. It proposed that approximately half of the Cruciform Building (phase 1) should be developed, using HEFCE grant funding. This would provide state of the art facilities for medical education, allow teaching of a significantly enlarged intake of medical students (following merger with the RFHSM) on a single site, and provide core facilities for the remainder of the building. The development of research facilities (phase 2) depended on the college attracting funding from external sources other than the HEFCE, and was to be let as a separate contract.

20. The college developed the business case further during 1995, working closely with the HEFCE to ensure that final proposals conformed to guidelines issued for Tomlinson funding. Following approval in principle to a grant of £14.5 million, it appointed a design team through notices in the Official Journal of the European Communities (OJEC). It developed an overall design strategy together with detailed designs for phase 1 of the redevelopment (the repairs and alterations to the fabric of the building, the core services and the fitting out of the teaching areas) and began a traditional procurement process. This project timetable anticipated the letting of tenders at the beginning of 1996, with completion of the teaching areas by September 1998.

21. It was decided that the redevelopment, procured with public funds under Tomlinson, should be subjected to PFI testing. As a result, both the HEFCE and DfEE became involved in the approval process.

22. The college submitted an outline business case (OBC) in October 1995 on the basis of traditional procurement. The development of designs for phase 1 of the project therefore continued, with the intention of keeping to the original timetable of letting contracts in 1996 with completion of works in time to use the building for teaching at the beginning of the 1998-99 academic session. In December 1995 the HEFCE formally confirmed its approval of the college’s OBC and project funding of £14.5 million, but the PFI test requirement remained subject to further discussion with the Treasury.

23. Following submission of the October 1995 OBC, the college:

• purchased the Cruciform Building from the Secretary of State for Health and undertook certain decontamination works using a combination of charitable and HEFCE funds
• began alterations to the UCL, Royal Free and Whittington sites.

24. In February 1996, the college was notified of the decision requiring it to PFI test the Cruciform element only, of the various developments. At this point, the design for phase 1 of the project was almost complete and the college had begun the tendering process. Despite the decision to PFI test, the college was allowed to progress PFI in parallel with the traditional procurement route (‘dual track’) to avoid unnecessary delays should the case for PFI not be proven. In October 1996 a notice was placed in OJEC inviting tenders for the entire building contract under PFI, following legal advice on EC procurement-related issues.

25. In the event, the ‘dual tracking’ approach was not feasible since contractors were not prepared to commit resources to two tendering routes simultaneously. All the same, the preparation of what became the Public Sector Comparator (PSC) from a set of fully worked up tender documents had the advantage of:

• better informing the PFI tenderers of the college’s requirements
• considerably shortening the PFI tendering period
• enabling the speedy evaluation of the PFI tenders against the PSC
• allowing contractors to be short-listed and interviewed under the initial OJEC notice for traditional tendering which assisted the same process under PFI.

Procurement process

Introduction

26. The Cruciform Building did not begin life as a PFI project. The contract for its refurbishment was relatively small (around £30 million) and, with a 90-year-old listed site, the options for innovation and commercialisation were limited.

27. After entering the initial stages of the PFI process the college identified an opportunity for PFI to demonstrate value for money. In little over a year from the OJEC notice, financial close involving private sector partners was reached on 5 December 1997. The refurbished research and teaching space is scheduled for occupancy in September 1999, in time for the academic session 1999-2000 and the intake for the enlarged Royal Free and University College Medical School.

28. The key stages in the procurement process are set out below.

Project team

29. A core negotiation team was assembled by the college prior to the OJEC advertisement and then maintained until financial close. The core negotiation team’s brief was to:

• report to and keep updated the senior college management, including the Provost, senior scientists and administrators at all key points during the process
• maintain contact with the HEFCE and DfEE throughout the process and involve them in negotiation and approvals as required
• maintain contact with all private grant funders and ensure their participation
• maintain contact and solicit advice from the Private Finance Panel Executive (PFPE) to ensure adherence to best practice in PFI
• liaise and negotiate with the private sector to obtain a transaction which offered best value for money and was affordable
• prepare the appropriate business cases for approval by both the college and the HEFCE prior to contract execution.

30. The core negotiation team consisted of:

• Jack Foster, Director of Finance, University College London
• Allan Mitchell, Assistant Bursar, University College London
• Pauline Williams, Project Officer, University College London
• legal advisers, Berwin Leighton
• financial advisers, Ernst & Young.

31. The public sector comparator scheme had been drawn up by a design team consisting of HLM (Architects), Davis Langdon Everest (Quantity Surveyors), Oscar Faber (Services Engineers) and Allott & Lomax (Structural Engineers) together with various members of the college’s estates and finance staff. It formed the basis of the invitation to negotiate (ITN) and was adopted by the bidding consortia. The design team continued to work on the project during the negotiating period but under the aegis of the preferred bidder. Input into the project’s requirements (down to the level of room data sheets) was completed for the teaching areas well in advance of the PFI procurement process. As collection of the detailed requirements for the research areas began before the PFI process and was largely completed by the time bids were received, very little primary collection of information was required during the process. This speeded up the PFI process, clarified both parties’ understanding of the design output specifications, and enabled the college rapidly to ascertain value for money for each element of the PFI tender.

Approval structure

32. The approval structure for the PFI contract was:

Approval of the OBC

Submission of the full business case (FBC)

Approval of the FBC

Approval of the transaction

33. From the outset, the core negotiation team identified the approval process as a potential major obstacle in achieving delivery on time. Therefore, frequent contact with the stakeholders was maintained throughout the procurement process. These meetings included:

The HEFCE

Formal letters or submissions at all key stages, such as:

• selection of short-listed tenders

• selection of preferred bidder

• consultation on evaluation of ITN responses

• FBC submission.

Grant funds

All private funders were updated on a fortnightly basis as negotiations proceeded.

The Provost

Senior management were updated on a weekly/fortnightly basis.

Pre-qualification and short-listing

34. Following wide consultation with the HEFCE, PFPE and DfEE, an advertisement for the project PFI was placed in the OJEC in October 1996. The college also contacted each of the respondents to the traditional development inviting them to bid under the PFI.

35. An information memorandum outlining the project and requesting brief details of each consortium was distributed to the initial 15 respondents to the advertisement. Six consortia were short-listed for interview. Each of the parties included leading contractors, facilities management providers and financiers. They were invited to outline their respective experience and knowledge of PFI to an evaluation team comprising the core negotiating team and representatives of the PFPE and the HEFCE. The objective was to reduce to three the number of tenderers who would be issued with the ITN document. At the second stage of this initial short-listing process each consortium was evaluated on the various criteria set out in the information memorandum. These included:

• financial stability of the candidates
• quality and experience of key personnel responsible for carrying out the project
• overall ability of the candidates to undertake the project
• strength and track record of the candidates in relevant education sector or similar projects
• understanding of the PFI principles of value for money and transfer of risk, and how they would apply in a project of this nature
• track record of achieving value for money and transfer of risk for the client on design and build contracts.

36. Only four consortia accepted the offer of an interview and the evaluation team was able to select unanimously the three short-listed tenderers.

Invitation to negotiate and tender period

37. The college spent much time and effort creating a set of output specifications suitable for inclusion in the ITN. This process was shortened by the information already collated under the traditional tendering regime, and fully involved the HEFCE and the PFPE. Tenderers were invited to base their compliant tender on the use of the Cruciform Building and then offer variants outlining different service delivery strategies.

38. Although technical designs and drawings (and all data supporting this work) which the college had prepared for the traditional procurement were made available to the consortia, no effort was made to require or recommend their use. Tenderers were also given access to the site surveys, user briefs and all supporting documentation to ensure that the college’s requirements were fully communicated and understood.

39. During the tender period (approximately three months) several meetings were held with the three consortia, and the college clarified issues as they arose. In February 1997 all three consortia submitted their tenders for the provision of teaching and research facilities.

Tender evaluation and selection

40. The tender documents were evaluated by the college’s technical team and professional advisers, and a range of clarification questions and issues were prepared for each tender. The tenderers were asked to clarify and revise their bids on the basis of the identified omissions, errors and clarification questions. The revised bids (the ‘Best and Final Offer’) were then finally compared.

41. All three tenderers offered essentially the same facility and services over the 25-year term of the transaction. Possible non-financial benefits which might accrue as a result of the service configuration in individual schemes were not considered a major factor in the selection decision (these were, however, important in June when the FBC was submitted).

42. The college prepared a detailed analysis of the PSC and PFI tenders which compared:

• capital costs
• financing and other fee expenses
• services cost
• life cycle costs
• availability fee
• total unitary charge.

43. This exercise enabled the college to investigate variations in the tenders and to probe areas where a tender seemed to be over- or under-priced, or widely at variance with the PSC. Its overall driving comparison was the net present value (NPV) of the unitary charge after making adjustments for comparability. The college then further evaluated the three tenders and the PSC by applying a risk weighting. A detailed analysis of risks was prepared, applicable to seven components of the project and with over 100 specific risk assessment areas. Risk adjustments to each of the tenders ranged from 4.9 per cent of the PSC to a low of 0.8 per cent, reflecting the different risk profiles of the PFI tenders.

44. An NPV analysis of the PSC and consortium tenders, adjusted for risk and assuming a capital contribution by the college during the construction period, showed the lowest PFI tender could provide better value for money than the PSC. However, this was dependent on an analysis of sensitivities which demonstrated that this would only come about if the lowest tendering consortium delivered on the proposed capital savings, and at that point these were largely estimated. There was enough evidence, therefore, for the college to recommend formally to the HEFCE and the DfEE that clarification work be carried out on both the PSC and lowest tender during a further one month period. During this period, the second tenderer was held in reserve, but the third tenderer was stood down immediately.

45. During the one month extension period, the college gained an assurance that the preferred partner was able to deliver the majority of the capital savings, and negotiations continued as the FBC was developed. The reserve tenderer was stood down when it became clear that the significant deal points had been resolved - approximately six weeks after final evaluation of the ITN.

46. The consortium selected as preferred partner comprised the following:

• Jarvis Construction (UK) Ltd
• Jarvis (Facilities Management) Ltd
• Rotch Property Group Ltd
• a lender (later to become Newcourt Capital Inc of Canada).

Negotiations with the preferred bidder

47. April 1997 was largely devoted to discussions with the main contractor to establish the significant risks associated with the building contract, which would enable the consortium to agree the guaranteed maximum price. At that time, the outline timetable required the FBC to be submitted to the HEFCE in early May, with financial close envisaged for July 1997. The first draft of the FBC was actually submitted on 9 May. All parties recognised the very tight schedule for completion, but were confident that, with hard work and co-operation, financial close could be reached within the proposed timetable. In fact, gross under-estimation of the complexity of the transaction, and delays caused by the inability of some of the parties in the transaction to commit to the negotiating process, meant that financial close was not achieved until 5 December 1997.

48. During the intervening seven months the college and preferred partner:

• formally answered 70 questions on the FBC in writing
• finalised the project layouts and programme
• released enabling works to ensure the contract schedule would be met
• cleared the ambiguity over planning consents and listed building status
• developed legal documentation which finally encompassed over 3,000 pages
• developed a versatile and accurate financial model
• spent more than 15,000 man hours in closing the transaction.

49. The college maintained the continuity of its core negotiating team throughout the transaction. In addition the HEFCE, DfEE and the grant funders were regularly updated with developments.

50. FBC approval was formally received in July, although the college was told informally of the favourable outlook in June. During the June-July period detailed discussions took place with the consortium on heads of terms, but in the absence of formal HEFCE and DfEE approval to the college’s FBC the consortium was reluctant to commit to sizeable legal fees. Consequently, the college underwrote fees to an agreed limit to maintain the impetus of negotiations.

51. The consortium also failed to commit formally to a funder at an early stage and, as a consequence, the funder’s involvement early in the negotiations (assuming their willingness to commit at such an early stage) was not possible. During May, the consortium partners introduced a new funder.

52. Amid wide-ranging negotiations, specific issues of the UCL transaction included:

• the absence of a lease structure for the underlying asset
• the inability of the college to pledge the Cruciform Building to the funder
• dealing with external grant funders (the HEFCE and charitable institutions) during development
• meeting the deadline of the beginning of the academic year - autumn 1999
• negotiating the heads of terms
• analysing the corporation tax and VAT arrangements for both the college and the special purpose vehicle (SPV) established by the consortium as the private sector contracting body
• obtaining an order from the Charity Commission relating to the college’s powers to enter into the transaction
• completing the detailed design (phase 2)
• defining an outline project agreement
• maintaining relationships with external charitable sponsors outside the PFI negotiations.

Full business case

53. Through evaluating bids in response to the ITN, the college produced a comprehensive document (already considered by the HEFCE in March) which provided relevant financial summaries of each consortium’s bid. This included the primary schedules such as the NPV of cash flows and sensitivity analyses. During the negotiating period leading to submission of the final FBC in June, only minor changes were required to the original financial summaries produced in March. The PSC was amended to reflect the impossibility of delivering the refurbished building for teaching in September 1999. As a result, the programme envisaged research space becoming available first. When risk associated with the project was taken into account, in particular risk relating to the need to have completed the refurbishment works in time for the beginning of a new academic session, the PFI route demonstrated better value for money. The college was also concerned that no formal arrangement had been struck with the lender. If negotiations with the consortium were going to delay the September 1999 completion date, it was clear that a traditional tender route should be pursued.

54. The FBC also referred to the VAT situation. In the PSC the college was able to avoid VAT on certain elements of the construction relating to use of charitable funds and the listed building status for non-business use. Under PFI this benefit disappeared since the SPV, a commercial organisation, would be contracting the building works.

Financial close

55. After many months of negotiation the college reached financial close on 5 December 1997. It needed to demonstrate to the governing body, through its Chairman and Treasurer, and to the HEFCE that the FBC remained valid, in that the assumptions made for PFI had not significantly changed. The deal was struck on the telephone with the funders in Canada, and the gilt rate applicable agreed at the last minute of close of the London market.

The agreement

Contractual overview

56. As in most PFI transactions, the project agreement between the private sector and the college establishes the majority of ruling criteria for the service delivery and the subsequent payment. However, due to the complexity of the transaction and the many different interests involved, a number of other separate but supporting documents also proved necessary. A schematic diagram of the legal relationship is shown in Figure 1.

Project agreement

57. The project agreement determines the services required of the SPV, the programme and schedule for delivery, the method of determining the service payment to be paid by the college, and the consequences for non-performance by either party.

Building design and commissioning

58. The project agreement cross-references the design to the detailed drawings and specifications produced under the traditional tendering route and adopted by the consortium, subject to value engineering. The agreement also sets out the arrangements for commissioning the completed building.

Payment mechanism

59. Payments for services were broken down into two distinct categories: availability (relating to capital costs of the building including financing); and services (primarily facilities management services including contribution to a sinking fund for maintenance of the building fabric and plant renewal).

Availability

60. The college is required to pay 100 per cent of the availability fee if the Cruciform Building is completely available. However, in the event of areas being unavailable and the college being unable to occupy the space in question, then the payment is reduced according to the level of non-availability, based on a weighting attributed to the importance of the area affected.

Services

61. The facilities management (FM) services are monitored monthly and quarterly against a performance standard. This measurement is then applied to a system of weightings (after a buffer zone) which allows for up to 30 per cent abatement of the fee. Service payments may also be reduced pro-rata to areas unavailable. Repeated poor performance in respect of services can lead to self-help remedies for the college arising from termination of the FM services sub-contract. Included within the service payment is an element relating to building maintenance and plant replacement. The level of payments agreed from the outset continue to be paid by the college into a sinking fund which can be drawn upon by the FM services provider. Notwithstanding poor performance in other areas, payments into the sinking fund remain unabated.

Construction

62. The project agreement specifies the programme for delivery of the refurbished building, including reference to equipment requirements and an outline planned maintenance schedule. Payments during the construction period, reflecting grant funding received by the college, are specified and scheduled on the basis of the value of work to date certified by independent quantity surveyors. Commencement of services (that is, asset delivery) is required by a specific date, and failure to deliver will result in penalties. If delivery is not achieved prior to a ‘college long stop date’, the college is able to terminate the agreement subject to an exit payment. If a ‘default long stop date’ is reached, the college or the lender is also able to terminate the agreement, again with payment of a termination sum by the college but with the SPV forfeiting their investment in the project.

Termination

63. Termination events proved to be the most significant issue arising during the negotiations and became pivotal to nearly all of the separate legal documents. Much effort was incurred in developing a table of termination events and the notice periods for which certain events could be called. The termination events summary continued to be referred to and developed well into the finalisation of the project agreements, particularly since it was necessary to ‘dove-tail’ the dates for triggering termination with those within the financial documents. The project agreement specifies the different possible termination events (for example, SPV default, college default, force majeure, uninsured event) with the relevant compensation arrangements for payments to the SPV, lenders and/or the college. It should be noted that the lender’s exposure on the deal is highly protected, and discussions with them centred around their rate of return; their investment in the project is backed by security over assets and the college’s covenant. On the other hand, the consortium members who are shareholders in the SPV stand to lose their entire investment. In view of the special arrangements entered into between the college and the Wellcome Trust, a termination event occurs after 10 years if the building has not been occupied for whatever reason.

Performance-related payments

64. Negotiations took place on a method to establish the expected quality of performance in the hard and soft service provision and the actual management of these services. A matrix of criteria was agreed between the parties to form the basis of regular monitoring. Financial penalties on a remission profile were agreed but opportunities to recoup deductions were also given to offer incentives to the service providers to improve declining performance. In parallel the college negotiated a methodology for monitoring services and communicating between the parties on a day-to-day basis.

Other major negotiating issues

65. The following were major factors during the negotiation of the project agreements:

a. Insurance. The college spent substantial time identifying insurance issues and the responsibilities falling on each of the parties to the agreement. Even as financial close approached, requests for policy amendments were being received, requiring last minute action on the part of insurers and resulting in legal argument over what was or was not acceptable to provide comfort to the consortium partners.

b. Performance bond. The contractor obtained a bond for 10 per cent of the construction value. In the event of failure to complete the contract, the SPV would have the financial ability to complete the project. It was considered prudent to negotiate for this bond to be assigned to funders or UCL in the event of the SPV’s inability to perform (since the contractor forms part of the SPV).

c. Changes by the client. The college needed a mechanism for introducing client changes during the construction, contract and operational phases. For the construction phase, it agreed the speed of response necessary for both the consortium and the college to assess the effects on capital, unitary charges and the programme itself. For the operational phase, the consortium was to be the first point of contact to cost and carry out negotiations, but the college negotiated a reserve position whereby it could introduce alternative contractors if it thought it was not achieving value for money. The subsequent introduction of five client change orders with agreed financial impact and without delay to the programme bore out the benefit of this clause.

d. Delivery delays. Delivery was tied to the start of the 1999-2000 academic year. If the delivery were to miss this deadline, the school could not relocate until the Christmas break. The private sector accepted this risk and is subject to penalty payments if delay occurs, over and above the existing agreement whereby payment is made when the project is complete.

Construction documentation

66. This is the design and build contract between the SPV and the contractor, Jarvis Construction (UK) Ltd. It includes contractor warranty and parent company guarantee.

Facilities management agreement

67. This contract is between the SPV and the FM service provider. Its terms enable the service provider to participate in the design development of the scheme and produce a planned maintenance schedule for the completed facility. It lays down procedures for market testing of the subcontract services to ensure value for money, and procedures for variations to service provision. It includes all the relevant project agreement provisions ‘back to back’ (such as availability, performance-related payments, sinking fund arrangements, quality assurance and dispute resolution).

Accounts agreement

68. The SPV is required to maintain several bank accounts to provide, for example, for liabilities relating to the sinking fund and debt service reserve. The agreement specifies the required account, when and for what purposes deposit and withdrawals may be made, and whether the lender and/or the college have a charge over the accounts.

Direct agreement

69. The direct agreement is between the lender and the college. It establishes the relationship between them when there is a default event and termination becomes an option. It also stipulates the covenants and pledges which the college must maintain over the life of the loan.

Event of default

70. If, as a result of SPV default, the project is about to be terminated, then the lender may ‘step in’ for a period of time to rectify the problem. This extension period allows the lender an opportunity to ‘protect’ their loan when problems arise, without the immediate threat of termination by the college. However, failure on the part of the lender to rectify the default event leaves the college with the option of terminating the contract.

Covenants and restrictions

71. The agreement requires the college to provide security in respect of the lender’s investment in the project (‘lender’s liabilities’). In this case, because of restrictions imposed by the charitable benefactors on the Cruciform Building, alternative security was offered by the college. If the college fails to meet its payment obligations under the contract, the lender has recourse to those assets secured under the agreement but the Cruciform Building remains with the college.

72. The college is also required to adhere to its financial memorandum with the HEFCE.

Loan agreement

73. The loan agreement, between the lender and the SPV, specifies the loan amount, how and when the funds are released during the construction period, terms of the loan (that is, interest rate, term, amortisation schedule, etc) and the covenants of the SPV during the term of the agreement. The agreement was entered into ‘back to back’ with the project agreement and direct agreement.

Other documentation

74. The entire project documentation consisted of multiple documents which in total covered a conference room table and took several hours to execute. Primary points were that:

• the SPV simultaneously entered into agreements with its subcontractors to mirror the requirements of the project agreement

• the college obtained direct agreements with the primary service providers to ensure service performance in the event of default by the SPV

• financing of the construction was from a mixture of grants, equity and lender funding. Therefore assurance was obtained that the funds would be available from all parties and then a joint mechanism for releasing funds was agreed

• subsidiary agreements were required for the college’s charitable sponsors

• the shareholders formalised their relationship through a shareholders’ agreement

• the college required undertakings from parent companies relating to the involvement of their subsidiaries.

75. To the extent that the project was not a typical PFI arrangement, a certain degree of innovation was required in the preparation of legal documentation.

Value for money and risk transfer

Introduction

76. The FBC included a value for money analysis which incorporated an NPV analysis of the tender, a risk analysis, and a sensitivity analysis of primary drivers. The analysis, in line with Treasury guidelines and HEFCE practice, yielded various unusual aspects and procedures.

NPV analysis

Comparing indexation

77. The three PFI tenders and the PSC were based on varying indices applicable to the college’s payments. Most of the payment profiles were RPI-linked but some involved fixed increases or flat payment profiles. As the NPV comparison following Treasury guidelines is based on a real discount rate of 6 per cent and excludes inflation, some of the payment profiles needed adjustment. This variance in payment profiles was sensitive to projected inflation rates and variations in the discount rate. Although this did not change the NPV ranking, there were fluctuations in the absolute rate.

Comparing proposals

78. The college undertook a detailed breakdown of each tenderer’s development budget, operating costs and cyclical expenditure before finalising comparison of tenders. The summary analysis (which included any allowances it thought necessary) was then presented to tenderers for review and comment. This clarification process gave confidence that each tender had been understood and that analysis had been conducted on a like-for-like basis.

Risk transfer

Risk of time

79. Although the September 1999 deadline was fixed, it was difficult to allocate value to each monthly delay. The college managed to attribute a value to this area of risk by evaluating the cost of potential temporary accommodation for staff, the effect of delays in the new medical curriculum on undergraduates, the wide damage to institutional reputation, the loss in potential research income due to the flight of prominent researchers, and a potential research income loss in committed grant funding. These prospective costs were aggregated and then attributed to a four month period (representing the first term of the academic session). The college then applied the monthly figure, including a probability weighting, in comparing PFI with the PSC.

Risk of PFI failure

80. At the start of negotiations, the college believed that there was a good chance the PFI tender would fail to deliver due to the relationship with the funder. However, as the process continued, the lender became more involved, long-term interest rates declined and the college was able to reduce this risk to an insignificant figure for the final FBC submission.

Lessons learned

State of readiness

Feasibility

81. Since the college had completed extensive work in preparation for traditional procurement, the feasibility had already been established. The project was affordable, grants were secured and planning issues basically in hand. This meant that during the procurement process, external surprises were minimal. At all times the college believed that the project would be achieved.

Design parameters

82. The college spent a great deal of time preparing a thorough design brief through consultation with departments prior to procurement. By removing the need to refer to the end user during negotiations, it not only saved time but avoided the user commissioning work with the private sector.

Public sector comparator

83. The college completed extensive capital life cycle and facility management costing in the process leading up to traditional procurement. This gave it confidence when negotiating, enabled it to respond quickly to the tenderers and validated the full business case.

Negotiating team

84. The college committed substantial in-house senior resources to the project. The team, whose members had an existing rapport with the stakeholders, was given sufficient scope to engage the private sector in meaningful negotiation that was rarely overruled. The college supported this team with advisers who were fully briefed on the transaction and experienced in negotiating similar contracts.

Stakeholders

85. The college began the project with all stakeholders fully committed to the project, strengthening this attachment by consistent reporting, contact and involvement in decision making.

Programme

Deadline

86. The September 1999 deadline had to be met or the PFI transaction would die. In addition, should delivery be delayed the academic calendar would prevent the school from relocating until the Christmas break. An awareness of this timeframe drove negotiations, quickened the pace and kept the project from stagnating.

Enabling works

87. The college demonstrated genuine commitment to the project by releasing a demolition package prior to close. This confirmed its eagerness to complete the project to time and maintain a realistic programme for the contractor. At the same time the consortium was not allowed to gain too strong a foothold on the project, and it would have been possible to revert to traditional procurement.

Alternative route

Traditional procurement

88. The college always had an alternative to PFI - a traditional construction contract. The private sector understood this and realised that a true limit existed on the total annual service payment. Overall, this maintained the commerciality of the negotiations and helped to secure value for money.

Alternative funding

89. Some way into the negotiations, the consortium introduced a potential second funder as competition to the preferred lender. Although this caused a great deal of anxiety and strain for all of the PFI consortium members (and the college), this action may well have provided a better balance to the transaction by encouraging the lenders to act reasonably in their demands.

Financial model

90. Early in the negotiation process the college realised that the financial model was insufficiently flexible for sensitivity analysis and the assessment of technical items such as tax and accounting policy. In the event these issues took over two months to resolve. Although the consortium held its price, this difficulty created a certain amount of tension, if not suspicion. This drawback could have been overcome had the college allocated the appropriate resources earlier in the process.

Full business case approval and submission

91. The college separated the FBC into two submissions. First it presented a preliminary FBC which recommended the selection of the preferred tenderers, established where the transaction would finalise and reconfirmed the value engineering. The second and final FBC required fairly minor changes and was primarily built on appendices previously approved.

92. Obtaining approval to elements of the first submission gave the college and consortium confidence that the transaction was in line with all stakeholders’ expectations. This provided more clarity in negotiations and gave the consortium more confidence to commit to negotiation costs.

93. The college prepared both the first and second FBC submissions within two weeks. Although this was very painful at the time, it avoided a long drawn out period of writing and rewriting which is characteristic of many business cases. It should also be noted that all questions were answered within three days, in writing.

Organisation of the negotiations

94. Negotiations were frequently delayed after the lender became engaged because points previously settled between the college and the private sector partner were unacceptable to the later stakeholder. This could have been avoided if the lender had been engaged earlier in the process. This is a common problem in PFI.

95. Lack of availability on the part of the lender’s in-house solicitor often produced a delay in negotiation of difficult issues. The college eventually overcame this problem through a more consistent and intense involvement from the in-house lawyer. The college understands that the lender has recognised this problem by the appointment of a locally based legal adviser.

96. Negotiating meetings frequently involved 15-20 people at any one time. Eventually all parties began to delegate negotiating roles downward to much smaller groups. These smaller negotiating teams then assembled larger issues for the principals to resolve in a senior forum. This strategy saved time and allowed for some rapport to be generated between all parties.

Termination

97. The importance of termination events and related notice periods should not be under-estimated, and were continually referred to during negotiation of the project agreement and other agreements. Although a termination grid document was drawn up mid-way during negotiations, it would perhaps have been more helpful to have this at an even earlier stage to assist each party with its understanding of the nature of this PFI project and to identify key negotiation points. Even towards finalisation of the document, it was necessary to re-visit the termination grid in order to co-ordinate the various notice periods arising under the different agreements.

The college’s powers

98. There was some doubt as to whether the college was within its powers, as a chartered body, to enter into the transaction. Given the understandable caution of the lender and their desire for adequate guarantees, the college requested a Section 26 Order from the Charity Commission. The issue of powers should have been considered at a much earlier stage to avoid major concerns for all parties and to smooth the negotiations.

Security

99. Negotiations might have been shorter if there had been clarity at the outset about the need for collateral on the loan and what sort of property might have been acceptable.

Legal opinions

100. Early in the legal negotiations, the issue of who was to provide legal opinions, and what legal opinion was required by whom, should have been resolved. In the event, there was a last minute frantic ‘legal opinion negotiation’. This was resolved, but its absence might well have threatened the entire deal.

Preferred partner

101. The preferred partner that the college selected was by far the best placed of the three tenderers. Added to the excellent existing relationship, this advantage meant that the college genuinely wanted the transaction to succeed. Throughout, it had confidence in its preferred partner and believed that the economics offered the best value for money.

Conclusion

102. Where existing standard forms of contract were used, negotiations were less protracted. Work on providing such outline forms of documentation and guidance on principal issues affecting PFI are currently at an advanced stage of production. The college would, undoubtedly, have found this of use, to shorten what turned out to be a lengthy process.

Acronyms and abbreviations