House of Commons - Public Accounts

Select Committee on Public Accounts Minutes of Evidence

Letter and supplementary memorandum submitted by the Department for Transport

At the Committee of Public Accounts hearing on Wednesday 23rd June, the Committee requested a number of notes which I am pleased to enclose.[23]

I thought it might also be helpful to the Committee if I set out why I consider the PPP contracts do represent good value for money. As the National Audit Office (NAO) Report notes,[24] the Department's objectives for the Public Private Partnership were to obtain private sector investment and expertise to modernise the Tube, while guaranteeing value for money for passengers and taxpayers and safeguarding the public interest, in particular safety.

London Underground's (LU) infrastructure is widely acknowledged to be a large and complex asset base that has suffered from many years of under investment and is in need of modernisation and additional capacity. This is highly challenging in engineering terms as it requires existing services to just under one billion passengers a year to be maintained, while carrying out improved day to day maintenance and the capacity enhancements and upgrades needed, all within very limited engineering hours.

The PPP bidders had to deliver better value for money against the public sector alternatives, as measured by the public sector comparator. These were tested against a bond option, even though it was not available to LU, as well as more conventional funding. The public sector comparator provided a range of prices over the 30 year contract period. LU's conclusions depend on the Infracos delivering the expected level of performance with sufficient economy and efficiency to offset the higher private sector borrowing costs. As the NAO Report explains[25] "The costs [of higher borrowing] will be covered if the PPP delivers about one third of the performance benefits considered in bid evaluation.". It should also be noted that the premiums charged by lenders are in line with the rating agencies' independent assessment of correct market rates. Consequently, the lenders' premiums reflect the real levels of risk faced on the non-guaranteed portion of the finance provided.

To provide assurance that the PSC developed by London Transport was fair; the Department commissioned Ernst & Young to review the overall robustness of the value for money assessment. This included reviewing the methodology, the assessment of risk retained by LU in the PPP contracts and that transferred to the private sector, consideration of financial and non-financial factors in the overall value for money assessment, and the overall robustness of the conclusions reached. Ernst and Young concluded that the methodology adopted for assessing value for money "had been robust and appropriate". KPMG, LU's own auditors, also undertook a methodology review of the PSC and confirmed that HM Treasury's guidance had been adhered to, as well as auditing the financial model and related sensitivities.

The bids for the PPP contracts offered higher performance than LU in some areas, as well as additional enhancements. In order to deliver the returns quoted for the shareholders the Infracos will need to have much stronger control over costs and full project planning than LU had previously managed to ensure work is completed to time and budget. The Central Line upgrade of the 1990s was six years late, failed to meet capability targets and was 31% over budget. If just one of these failings were repeated by an Infraco under the PPP contracts for an upgrade, then its rate of return could easily be halved for the full 30 years of the PPP contract. If all three aspects were to be repeated on a line upgrade, then the Infraco would make very low levels of profits for the life of the contract, and these would only be available if the Infraco delivered all its other upgrades successfully. Project and integration risk has genuinely been transferred to the private sector with the Infracos heavily incentivised to manage it well.

Across the 30-year contract, there are periodic reviews and the independent PPP Arbiter provides assurance on value for money by determining at the outset of each review period the economic and efficient price for delivering the services required by London Underground. Only if the Infracos achieve these levels of service and deliver them within the price set by the Arbiter will they achieve their target levels of return on equity.

NAO comment that there is only limited assurance that the price is reasonable, largely because there is uncertainty about what that final price will be. However, detailed analysis showed the PPP to be cheaper than the alternatives in delivering like-for-like outputs, while the Arbiter ensures prices remain reasonable in the long term. Consequently, while there is uncertainty about the long-term price, I am confident that it will offer better value for money than the alternatives. Greater price certainty would have increased the cost because bidders, faced with a fixed and inflexible contract, would have based their cost estimates on a worst case scenario to ensure a return. This would have reduced value for money, as acknowledged by the NAO.[26]

Finally, although large, the costs of the transactions are not excessive given that they were for 30-year contracts to maintain and upgrade large and complex infrastructure. The negotiations took longer than anticipated as the NAO Report acknowledges and this increased the costs for all parties. Without the PPP, LU would have had to negotiate separately a potentially very large number of other contracts to provide the maintenance, modernisation and capacity enhancements that the PPP will deliver. These would all have had their own transaction costs and it would have left the risk of integrating and co-ordinating all the workstreams entirely with LU.

David Rowlands

Permanent Secretary

9 July 2004

Questions 36-38 (Jon Trickett): Supplementary evidence on inadequate and inconsistent information from the Infracos.

I understand that Transport for London (TfL) will be providing further information on this issue.[27]

Questions 57-71 (Mr Bacon) and Question 72 (Mr Williams): Extra costs of private sector borrowing: Breakdown of total debt costs—private and estimated public costs and basis for £450 million difference.

The private sector's cost of borrowing on senior loans is made up from the addition of the underlying interest rate and the margin charged by its financiers to reflect their perception of the credit risk of the borrower. The blended margin (including fees) of the three Infracos, averaged across their 11 separate finance facilities, is 1.35% a year and the underlying blended interest rate is 5% a year, giving a total cost of finance of 6.35% a year.

The total amount of the three Infracos' debt service (interest, fees and principal repayment) over the 30 years is £4.3 billion (using Government's 3.5% real discount rate[28]) or £9.7 billion, expressed in nominal or money of the day terms.

If the PPP had been funded using public sector finance, raised by way of bonds, then providers of finance will price their bonds not on the risks inherent in the company, but the risk of default by the bond guarantor. In the example given in the National Audit Office (NAO) Report "Were they good deals? " (paragraph 2.36) the NAO considers the reduced cost if the bonds were issued directly by central Government, who attract the highest credit rating; AAA. The underlying interest rate on these bonds is the gilt rate on which no additional margin is paid. So the cost of finance would be cheaper by a combination of the avoided margin and lower interest cost; around 1.6% a year cheaper overall, giving a total cost of finance of 4.75% a year.

Funding London Underground's (LU) investment programme in this way (using the Infracos debt profile) would lead to the cost of funds being lower by around £445 million over the life of the PPP (at 3.5% real discount rate) (using the NAO's calculation methodology) ie a total debt service cost of £3.855 billion or an average of £14.8 million per year lower than the cost of debt raised by the private sector. The high level analysis by the NAO estimated a similar discounted sum of £450 million over the 30 years ie a total debt service cost of £3.85 billion at an average of £15 million per year lower.

If the PPP investment programme had been financed using bonds guaranteed by TfL, then the cost of finance would have reflected their lower credit rating, ie AA, which trades at a cost approximately 0.6% a year higher than Government gilts. Therefore, TfL's total cost of finance would be lower than the Infracos by about 1.0% per year and total cost of finance would equal 5.35% per year. Cost of funds would be lower than that achieved by the private sector by a total of £210 million over the 30 years at an average of £7 million per year (at 3.5% real discount rate) ie a total debt service cost of £4.09 billion.

The Department's main objectives for carrying out the PPP were to bring in and incentivise private sector expertise, achieve clear risk transfer, maintain the infrastructure on a whole life basis and all with the support of stable funding. Requiring the private sector to raise its own finance and to have facilities in place to allow it to invest and absorb risk is a fundamental part of meeting those objectives.

Therefore, the Department considers that the key question is whether the higher cost of private sector debt—of which LU were fully aware at the time of the PPP process—is outweighed through the competition, introduction of private sector management skills and incentives to deliver performance that the PPP is designed to achieve. Under the PPP the higher cost of finance will be outweighed if the Infracos deliver the service requirements and cost controls anticipated under the contracts.

It should be noted that any private sector lender that has not been guaranteed by Government will impose a number of disciplines on its borrower, including requiring repeated technical due diligence, sensitivity analysis, reporting requirements, monitoring and control clauses, and regular reviews of performance. These disciplines are an important factor in ensuring that borrowers meet their expected performance levels and deliver the anticipated efficiencies.

Analysis showed that the benefits of the PPP option over a public sector bond option exceeded the £445 million savings available under gilts. However, any bond finance option would have entailed issuance by TfL. Consequently, there is an even stronger case for the PPP compared to TfL issued bonds which would have been £235 million more expensive than gilts, yet without introducing any new incentives or disciplines to improve performance.

To put this key question in context, the NAO calculates in paragraph 2.44 of the "Were they good deals?" Report that the incremental costs of finance would be mitigated if the PPP delivers about one third of the performance benefits anticipated. As another example, the additional cost of finance of the Infracos of £7.0—£14.8 million per year above the Government or TfL alternatives represents around 0.7—1.48% of the year one Infrastructure Service Charge. If those sorts of levels of efficiencies are delivered by the private sector, then the incremental cost of debt will have been outweighed.

Questions 79-85 (Mr Bacon) and Questions 166-168 (Jon Trickett): Transaction costs—Detailed breakdown of transaction costs, including all consultants paid by London Underground, difference in legal and engineering consultancy costs, differences on amounts paid to successful bidders.

We are currently assisting the NAO who are compiling the additional information requested.[29]

Questions 97-105 (Mr Williams) and Question 168 (Jon Trickett): Convergence of consortia and London Underground's cost estimates in the public sector comparator.

As a matter of policy, public sector comparators, which are the public sector's estimate of delivering the same level of service as the PPP/PFI scheme under consideration, are not made public during the competition.

This is because the public sector is looking to obtain best value through competition. There is a justifiable concern that disclosure of the public sector comparator would lead to bids only marginally below that value, so that competition would have been dampened.

The PPP public sector comparator is now fully disclosed in LU's Final Assessment Report. This report, which is being provided to the Committee by TfL, analysed the full background to the transaction, the alternatives available to LU for bringing investment to the Tube, and TfL's alternative proposals. In reviewing whether the PPP was value for money, LU considered in depth both the quantitative analysis, comparing the bids to the relevant PSC, as well as the qualitative arguments, why the PPP offered better value over public sector alternatives.

During the PPP procurement, the public sector comparator was regularly updated to reflect improving asset and performance data, a better analysis of underlying risks, and LU's changing performance requirements. Similarly the private sector bids were being updated on the basis of the same updated data and its own due diligence.

The chart included by the NAO in Appendix 3, figure 3.2, which compares movements in the 30 year public sector comparator with movements in the bidders' Infrastructure Service Charge in the first seven and a half years, shows that over time the relative rate of growth of those figures converged, ie that by the end of the bidding period they had grown by about the same amount since March 2000. This reflects the fact that they were responding to the same changes in underlying data and performance requirements.

While over time the respective costs grew at the same rate, the private sector bids were growing from a lower level. Throughout the procurement process, the total bid costs were at the bottom of, or below, the range of the public sector comparator. The extensive quantitative analysis conducted by LU therefore suggested throughout that the deals represented value for money. This, combined with the qualitative analysis documented in the Final Assessment Report, led LU to consider that the PPP represented the best overall value for money from a range of possible alternatives.

Questions 142-149 (Mr Bacon and Mr Williams): Transport for London preference for Bond Financing—Advice/views provided by TfL to Department for Transport. Summary of this based on London Underground's Final Assessment Report and Update.

I understand that TfL will be providing to the Committee and the NAO copies of LU's Final Assessment Report and the Update to the Final Assessment Report.[30] The NAO will separately be providing a summary of the issues raised by TfL on bond finance.[31]

Questions 164-165 (Jon Cruddas): Staff transfer, new recruits to Infracos & pensions position.

The Committee asked whether new staff of the Infracos were able to join the LRT Pension Scheme. There is nothing in the trust deed and rules that prevents new recruits joining. But it is for the relevant employers to decide whether membership should be offered as part of their recruitment packages. Only ex LRT employees have a statutory entitlement to continued membership of the LRT Scheme.

We understand that Tube Lines has not offered membership of the Scheme to new recruits since January 2003, but instead they are entitled to join the Tube Lines Pensions Scheme. Metronet is also understood to be considering alternative arrangements.

Questions 169-173 (Jon Trickett): Selection of preferred bidder, premature end of bidding process and increase in costs—did bidders estimate their costs after selection and what were they.

As the NAO Report "Were they good deals" points out in paragraph 2.19, the policy of maintaining reserve bidders had drawbacks, but no realistic alternatives were available. The option of pursuing a full reserve bid would have only doubled the bid costs with little value for money and would not have been attractive to those reserve consortia selected. As Figure 14 in the Report illustrates, the cost increases for Tube Lines are significant after the selection of preferred bidders in May 2001.

The appointment of preferred bidders was necessary to ensure that the selected consortium and their backers retained confidence in the bidding process. The consortium were reluctant to commit to detailed negotiations and preparation unless selected as a preferred bidder.

On the issue of recovering costs, successful bidders in any tendering process have to recoup their costs through the prices they charge for their services. In the PPP these are not an upfront cost to the public sector. Instead they are recovered through a component of the monthly Infrastructure Service Charge. While the bidders did not provide an exact estimate of their costs after selection, the scale of these costs was known since they largely relate to items such as legal and financial transaction costs which any preferred bidder would have had to incur prior to concluding the contract. As such, they were part of the cost that was evaluated as being better value for money than a public sector alternative. So the £114 million and £134 million paid to Metronet and Tube Lines respectively were included within the PPP process and pricing before the contracts were closed.

Questions 174-181 (Mr Jenkins) and Questions 182-186 (Mr Williams): Tube Lines Refinancing—Scope for further refinancing and possible gains.

I understand that TfL will be providing further information on this issue.[32]

Questions 6-11 and Question 186 (Mr Williams): Arbiter—decisions (binding & contractual position), penalty for non-compliance.

How are the Arbiter's decisions binding ie through the contracts or through another mechanism?

Section 229 of the Greater London Authority Act 1999 gives the parties to a PPP contract the ability to refer matters specified in the PPP agreement to the Arbiter for a direction. Where a matter is referred to the Arbiter he is required to make a determination which shall be final and binding. While there is no appeal from the Arbiter's decision, his decision would be susceptible to judicial review.

What is the penalty for non-compliance with a decision of the Arbiter?

The Arbiter has no sanction to impose any penalty on the parties to the PPP Agreement for non-compliance with his decision under the Greater London Authority Act 1999.

However, where the Arbiter's determination takes effect as a term of the PPP Agreement (see above) it is enforced in the same manner as any other provision of the agreement. The matter must initially be dealt with under the Dispute Resolution procedures contained in the PPP agreements which can result in the dispute being referred to Adjudicator and ultimately the Courts.

Questions 125-130 (Mr Jenkins) and Question 186 (Mr Williams): Note on fares over 25 years and balance of subsidy & London Underground income.

I understand that TfL will be providing further information on this issue.[33]