House of Commons - Public Accounts - Forty-Seventh Report

Select Committee on Public Accounts Forty-Seventh Report

THE PRIVATE FINANCE INITIATIVE: THE FIRST FOUR DESIGN, BUILD, FINANCE AND OPERATE ROADS CONTRACTS

ASSESSING VALUE FOR MONEY

Public sector comparators and discount rates

30. To assess the value for money offered by the PFI procurement, the Agency prepared public sector comparators for each of the four projects. The calculations were based on estimates of the construction and operational and maintenance costs of the roads over the lives of the contracts. The Agency also quantified the value of the risks transferred and incorporated them into the public sector comparator.[24]

31. We asked the Agency how robust was their methodology and how reliable were their figures as a basis for making comparisons with conventionally procured public sector comparators. The Agency told us that they had sought to base their analysis on a reasonable judgement, avoiding spurious precision, and that they were satisfied that it was a reasonable basis for testing the market and testing value for money. The quantification of the value of risk transfer had been assessed by the Agency's transaction teams, which comprised professionally qualified people with many years' experience of road schemes. A risk consultant had also checked the assessments independently.[25]

32. We asked the Agency whether they had constructed a public sector comparator based on a design and build contract rather than the conventional type of contract. The Agency told us that they had the experience and expertise within their evaluation teams to take account of a design and build approach but comparators were not prepared on that basis because, with the exception of part of the A1(M), the roads were originally scheduled to be procured using the conventional approach. They told us that they now estimate that had a design and build comparator been used on the M1-A1 project it would have been between £339 million and £333 million compared with the £344 million that was used as a comparator.[26]

33. As part of their analysis of public sector comparators, the Agency had had to take account of timing differences in future payments which they would have to make. They had done this by discounting future cash flows at a rate of eight per cent in accordance with advice which they had received from the Treasury. Table 4 shows that the Agency's assessment of the four projects indicated that savings offered by the PFI approach across the four projects would total around £168 million compared with the conventionally financed alternatives, although one of the projects—the A69—would be some £5 million cheaper using conventional arrangements.[27]

Table 4: The Highways Agency's assessment of the expected cost of the winning bids against the public sector comparator (£ millions net present value discounted at eight per cent)

	M1-A1	A1(M)	A419/A417	A69	Net saving
Expected cost of winning bid	232	154	112	62

Public sector comparator	344	204	123	57
Saving/ (additional cost)	112	50	11	(5)	168

Source: Comptroller and Auditor General's Report, Figure 14

34. The Treasury's own published guidance on investment appraisal which was current at the time stated that generally, and for all privately financed projects in particular, a six per cent discount rate was appropriate. The guidance stated that one of the exceptions to the use of the six per cent figure was in the appraisal of publicly financed roads, for which a discount rate of eight per cent was prescribed.[28]

35. Table 5 shows that using six per cent reduces the expected savings by some £69 million to £99 million. For example the A69 contract using private finance is likely to cost some £12 million more than conventional procurement compared with £5 million more using the eight per cent rate. The A419/A417 project which, using the eight per cent discount rate, indicated savings of £11 million over conventional procurement is expected to cost £3 million more than conventional procurement if the six per cent rate is used.[29]

36. We asked the Treasury what was the justification for the use of discounting and how the discount rate was determined. The Treasury told us that discounting was a standard method which needed to be applied because £100 now was worth more than £100 in five years' time. The discount rate itself was a matter of judgement but was based on the rate of return on gilts adjusted for tax and for systematic risk, and was currently in the range of five to six per cent.[30]

37. We therefore asked the Treasury why they had advised the use of eight per cent rather than six per cent. The Treasury told us that the reason dated back to 1989. The eight per cent rate had been introduced for the appraisal of publicly financed roads because at that time it was the required rate of return for nationalised industries and other central government bodies selling their output in competitive markets and was applied in investment decisions by British Rail. It had been felt appropriate to apply the same rate to roads to ensure that decisions on investment in road and rail were based on comparable information. The Treasury said that they appreciated that there was a case for using six per cent when private finance was introduced into roads procurement but a judgement had been taken to continue with eight per cent.[31]

Table 5: The Highways Agency's assessment of the expected cost of the winning bids against the public sector comparator (£ millions net present value discounted at six per cent)

	M1-A1	A1(M)	A419/A417	A69	Net saving

Expected cost of winning bid	288	192	140	78

Public sector comparator	372	222	137	66
Saving/ (additional cost)	84	30	(3)	(12)	99

Source: Comptroller and Auditor General's Report, Figure 14

38. The Treasury told us that, since the privatisation of British Rail, the argument for using eight per cent to facilitate fair comparison for investment decisions between road and rail no longer applied and the Treasury's new guidance recommended using six per cent.[32]

39. We asked the Treasury whether their decision to advise the Agency to use the eight per cent discount rate was related to the fact that the higher rate would make the privately financed options appear more attractive relative to the public sector comparators. The Treasury told us that, while it was true that the higher discount rate had that effect, the eight per cent figure had not been suggested as a way of artificially biassing the assessment.[33]

40. We asked the Treasury whether even the use of a six per cent discount rate was appropriate. They told us that six per cent was at the top end of the range and accepted that the level of savings provided by the private finance route was dependent on the discount rate used but that had a different rate been used bidders would have constructed their bids differently.[34] The Treasury, however, acknowledged that with hindsight they should have carried out sensitivity tests to assess the impact of different discount rates on the projects but that they had opted for the simple solution to use the eight per cent rate.[35]

41. Given the uncertainties surrounding the most appropriate discount rate to use in these assessments, we asked the Agency whether the methodology used to create the comparators was sufficiently robust to enable decisions on whether to proceed with the PFI route or an alternative to be made. The Agency told us that they felt it was a reasonable basis on which to test the market and value for money.[36]

42. We asked the Agency whether the additional costs involved with the A69 project represented the cost of developing a market in operating roads. The Agency told us that Ministers had accepted the potential that all of the deals would not offer good value for money on the basis that they were pursuing the wider policy objective of establishing a road operating industry.[37] The Treasury also confirmed that they were prepared to proceed with the A69 project as loss leader on the basis that it represented a learning curve from which they would be able to assess how PFI was going to work across a wide range of road procurement.[38]

43. We therefore asked the Agency whether they agreed that had the six per cent discount rate been used, both the A69 and the A419/A417 deals would not represent value for money. The Agency told us that Ministers were involved in the decision and were aware that on an arithmetical basis these projects were not value for money but that it was reasonable to go ahead given the wider objective of generating a road operating industry.[39]

Risk transfer

44. In common with other PFI contracts one of the Highways Agency's objectives in awarding these contracts was the achievement of value for money. In the absence of substantial risk taking by the private sector it is very unlikely that a privately financed road project can ever be better value than the same project financed in the traditional way by the Government. That is because the private sector expects to have to pay higher costs of capital than the Government faces in the market for gilt-edged stock. Without risk-taking by the private sector, for example to reduce the likelihood of the Agency paying for construction cost increases, the use of private finance can bring no benefits to offset the higher cost of finance.[40]

45. To achieve the best value for money it is necessary for individual risks to be borne by those parties best able to manage them. In general, the Agency succeeded in allocating the risks appropriately within the four contracts. Responsibility for construction and maintenance risks rests with the operators who have direct control over such matters, whereas certain legislative risks which could have a direct impact on operators and are subject to changes in Government policy are mitigated.[41] Generally bidders have indicated they were broadly satisfied with the risk allocation achieved in the contracts.[42]

46. We asked the Agency whether the experience with these first four road projects had confirmed the Agency's assessments of the benefits of the private finance approach. The Agency told us that they were very pleased with progress to date. In the case of the A1(M) transferring construction risk to the operator had saved some £40 million in compensation for unforeseen ground conditions which the contractor would have claimed for under a conventional procurement contract. Also the new construction elements of both the A69 and A419/A417 contracts had been completed ahead of schedule bringing the economic benefits of the projects forward. In the case of the A419/A417 these additional benefits amount to some £3.4 million .[43]

47. The use of shadow tolls appeared to place a risk on the private sector which the private sector would not be well placed to manage. This seemed likely to have reduced the value for money achieved by these roads.[44]

48. We asked the Agency why they had chosen this approach. They told us that they adopted shadow tolls as the payment mechanism because Ministers wished to familiarise the industry with the potential for real tolls and the role of being a road operator. They agreed that the effect was to create new risks for the private sector.[45]

49. We asked the Treasury whether shadow tolls introduced an unnecessary element of risk into the pricing mechanism which in effect raised the cost of all bids. The Treasury told us that in their view there may have been additional risk created by using shadow tolls which may have had the effect of raising the cost of all bids. They said that any such cost needed to be set against the benefits of acquiring important information about the industry's readiness to think in terms of real tolls. They accepted that the decision to base charges on shadow tolls was a policy-driven rather than a cost-minimising decision.[46]

50. There is also a risk to the Agency arising from the use of shadow tolls as they are in no better a position to manage traffic volume.[47] We asked the Agency whether they had estimated how much they might have to pay the private sector if traffic volumes grew unexpectedly quickly. The Agency told us that they had limited their liability to pay in the event of very high traffic growth by requiring consortia to set the highest tolling band at zero.[48]

Conclusions

51. The assessment of whether these four road projects are likely to offer better value for money than the conventionally procured alternatives rests on complex calculations. It also requires the exercise of judgement to define the costs of the conventional alternatives, to evaluate the benefits of transferring risks to the private sector, and to take account of differences in timing of payments of public money. We expect departments to carry out such assessments in a way which is sufficiently robust to support their decisions and which avoids spurious precision.

52. The value of these projects hinged on the trade-off between low early costs and high future costs. Yet at the time the Highways Agency were taking decisions on the contracts, they did not examine the sensitivity of their evaluations to the way they had quantified this trade-off, that is the time value of money. The standard method which they used for quantifying this gave results which depended very much on a key assumption, the discount rate. But they were not aware of the fact that a small change in the discount rate would have a large impact on the evaluations of these schemes. We find this omission surprising, especially given that this first tranche of roads was intended to be a prototype for the policy of privately financed roads.

53. The Treasury's approach to evaluating the time value of money in this case puzzles us. Their published guidance pointed clearly to the use of a discount rate of no more than six per cent for comparing privately financed projects with their conventional alternatives, yet they advised the Highways Agency to use a figure of eight per cent. We recognise that prior to the privatisation of the railways the Treasury's guidance required the use of a discount rate of eight per cent in appraising publicly financed roads, and that following railway privatisation the Treasury now recommend using a six per cent discount rate. We fail to understand, however, why the change in the ownership of the railways has anything at all to do with evaluating the alternative ways of financing these four roads.

54. This case has shown that the use of a discount rate which is too high tends to overstate the benefits from the private finance options. Even a discount rate of six per cent is regarded by the Treasury at the upper end of the range of valid figures. We therefore recommend that the Treasury consider further the appropriateness of the six per cent discount rate to assess the value for money of PFI projects.

55. With one major exception, the Agency placed the risks inherent in these contracts with the parties best able to manage them and thus promoted good value for money. The exception was the decision to use traffic volumes as a basis for payments to operators. This was done to test the ability of the private sector to bear one of the risks involved in running tolled roads, but it created a new financial risk which bears both on the public sector and on the private sector and which neither party is able to manage effectively. Bidders can be expected to have included a premium in their pricing for taking this risk which is likely to have reduced the value for money offered by these contracts. Departments should be wary of creating new risks when seeking to transfer risks to the private sector.

56. Although the four roads, taken together, may be expected to deliver savings of £99 million using the 6 per cent discount rate, the decisions to proceed with the A69 and A419/A417 projects as PFI contracts are likely to cost the taxpayer some £15 million more than the conventional approach. This amount represents the cost of the decision to give more weight to the objective of encouraging the establishment of a United Kingdom road operating industry than to achieving value for money from these particular roads.



© Parliamentary copyright 1998	Prepared 2 July 1998