Supplementary memorandum submitted by
the National Audit Office
Questions 142-147 (Mr Bacon) and Question 149
(Mr Williams): Transport for London (TfL) Bond Financing Proposals
INTRODUCTION
Mr Bacon asked Mr O'Toole for documents, in
London Underground, prior to his arrival, containing advice given—by
TfL—to the Department on preferences for bond financing.
Mr Callaghan (Q 148) referred to London Underground publishing,
in 2002, two reports called the Final Assessment Report
and the Update to the Final Assessment Report both of which
summarised TfL's submissions at some length. Mr Williams asked
the NAO to do a brief critique of London Underground's summary
for the Committee.
OVERVIEW
All parties agree that the effective modernisation
of London Underground requires:
— the assurance of stable funding over
many years (and all recognise that such stability cannot be guaranteed
with conventional public finance);
— effective risk management of the
physical upgrading work (which, under all financing options would
be carried out by a range of private sector contractors).
In both TfL's approach and the PPPs', stable
funding is provided because the greater part of the money needed
is raised at the outset. The direct costs of finance would differ:
TfL's approach would be cheaper than in the PPPs.
The approaches to risk management are radically
different. TfL would rely on strong public sector management of
the private sector contractors. In the PPP, the private sector
would have higher exposure to risk than in a conventional project
but greater responsibility for managing those risks. Because private
sector money would be at risk, there would be incentives for the
private sector to manage the risks effectively.
If private sector risk management as in the
PPPs is more likely to be effective than the public sector contract
management proposed by TfL, then the extra cost of the PPP finance
might be outweighed by the benefits of superior control of the
risks. The Government took that view and ruled out public sector
bond options accordingly. The TfL public sector bond option would
also have been unique in the field of local government finance
in the UK.
The bond financing proposals run up against
two opposing philosophies. TfL proposed lower cost finance within
a contract framework controlled by the public sector to counteract
private sector self-interest. London Underground (when owned by
London Regional Transport) considered that higher cost finance
included transfer of sufficient risk and project control to the
private sector so that their self-interest would incentivise them
to deliver performance that would more than offset the extra financing
costs.
TfL's bond financing proposals
In February 2002, TfL proposed raising £2
billion to £3 billion in part by bonds secured on the London
Underground fare box ("Asset Backed Bonds") supplemented
by bonds based on TfL's corporate good quality investment rating
(which it expected to be rated as AA or high investment quality).
These proposals formed part of a different public sector approach
to modernising the Tube which was set out in the Mayor's Transport
Strategy published in April 2001.
Several advantages were claimed for this approach
to modernising the Tube which TfL claimed had been used to turn
around systems in Boston, Philadelphia, New York City and Toronto:
— gives local government the financial
means and full responsibility;
— provides stable funding, largely
independent of annual grant negotiations (with capital funding
segregated from operating budgets);
— management skills, including certain
project skills, are retained in house and hold private sector
contractors (handling large-scale resource intensive projects)
to task; and
— lower borrowing cost than PPP, as
further described below.
TfL's advisers recognised that raising the total
amount would need five to seven years of Government financial
backing for the bonds through grant provisions. In response to
the objection that bonds would not lead to the transfer of risk
to private sector shareholders, TfL argued that risk transfer
could instead be achieved through rigorous contracts with the
suppliers of infrastructure improvements and services, without
having to pay an equity risk premium across the entire project.
TfL also argued that their retained project risk under the PPP
was equivalent to the project risk under their preferred structure.
On this basis TfL estimated that its proposals
would lead to a saving in financing costs of £36 million
to £84 million per year. About one-third of the saving would
arise because of higher PPP borrowing cost (rated at a low investment
quality or BBB) and the other two-thirds from saving the cost
of remunerating PPP equity at around 20% per year. A further unquantified
potential interest cost saving could arise if bonds were only
issued in tranches of £500 million as and when funds were
needed (but there would be a risk of lack of market interest or
of the necessary support from central government at the time).
Transport for London estimates of financing
costs:
| Cost of financing (excluding fees)
| February 2002
(TfL report) |
4 April 2003
(Metronet close) |
| Estimated average cost of PPP debt and equity
| 6.7% to 8% | 7.15% per year
|
| Cost of AA rated corporate bond financing |
5.2% to 5.5% | 5.29% per year |
| Differential in cost of finance | 1.2% to 2.8%
| 1.86% per year |
| Assumed amount of financing | £3,000 million
| £5,000 million |
| Annual cash differential | £36 million
to £84 million
| £93 million |
| | |
Source: Estimates based on publicly available (Bloomberg)
market data. The calculation based on the market rates and the
terms achieved by Metronet in April 2003 is referenced in the
NAO Report "London Underground PPP: Were they good deals?"
paragraph 2.43 (page 26).
London Underground's findings
London Underground generally did not dispute the market-based
pricing that TfL put forward, but did not expect Treasury approval
"recognising that this is not the way public infrastructure
projects have historically been financed in the UK". As a
related matter, London Underground advised their Board not to
assume, under a radically changed approach, the same level of
grant support as that available for a PPP that met policy objectives.
Of the contract structure, London Underground said that it
would not deliver the PPP policy objectives on risk and contract
responsibility, noting "Some of the contracting structures
proposed by TfL, requiring equity contributions from suppliers
or parent company guarantees, are not standard ways of doing business
with the UK public sector, and may add to the time needed to put
contracts in place." Adding to timing problems, a new procurement
competition would have been required.
Publicly, the advisers to TfL and London Underground showed
themselves unable to agree on a common set of assumptions for
making undistorted bond financed price comparisons. Little common
ground exists, and Ernst & Young, [37]as
further discussed below, largely discarded London Underground's
bond comparison (see NAO Report "London Underground PPP:
Were they good deals?" para 2.42 on page 26).
London Underground's May 2002 "Update to the Final Assessment
Report" gave a high level summary of TfL's objections to
London Underground's findings. In addition to maintaining that
the PPP was not value for money, TfL challenged the findings about
the extent of delay and the availability of grant.
NAO analysis of TfL's proposals and London Underground's findings
Part 1 of the NAO Report shows that in discussion of the
design of the PPP there was analysis of public operations combined
with private funding (paras 1.9-1.10, 1.17). This analysis showed
that such an option would only have been possible with Government
backing, which the Government was not prepared to offer, in part
because the Government wanted private sector direction of the
expenditure in the belief that it would thereby be more effective.
Paras 2.41-2.44 discuss the possible savings from TfL's approach
(updated in April 2003 to represent £5 billion of bonds at
5.29%) compared with Metronet's actual financing structure. Ernst
& Young, as para 2.42 Reports, considered that any such saving
should be treated with caution in project selection. In particular
if the cost of capital (5.29%) is lower than the discount rate
at which TfL make the comparison (8.65%), discounting the cash
flows appears to reduce the public sector cost. This would imply
that the more money that is borrowed by the public sector, the
more value for money would improve. It is partly for this reason
that Ernst & Young, correctly in London Underground's view,
dismissed the relevance of bond financing comparisons, as they
felt that in all circumstances an element of distortion, which
they termed arbitrage, would occur.
Arguably the risks to a project under public sector management
might outweigh the benefit from cheaper finance—and have
certainly done so in previous projects. Instead of letting a lower
public sector borrowing cost act as a form of arbitrage or subsidy
to a project, Ernst & Young asked how much the public sector
might save in the costs of carrying out such a project if annual
grant funding were to be replaced by long term stable funding.
Ernst & Young then estimated the potential saving from such
stable funding at some £825 million. On London Underground's
analysis the Committed Finance Offer bids for the PPP remained
value for money after reducing the estimated public sector cost
by this amount of saving.
Para 2.43 identifies that the debt and equity financing structure
agreed under the PPP costs £90 million more annually than
the estimated cost for TfL issued bonds, two thirds of which is
related to the cost of equity and is dependent on performance
under the PPP. Para 2.44 estimates the annual extra cost of the
private debt borrowing as £30 million. This debt cost is
generally payable regardless of performance. The extra cost has
to be considered against an assessment of the benefit of risk
transfer and scrutiny of the project by private lenders.
Bond issues were not available to London Underground because
the Government attached priority to bringing in private sector
responsibility in the manner envisaged by the PPP. The scope for
improved value for money compared to the option adopted would
depend on public sector efficiency under the bond option compared
to private sector efficiency under the PPP option.
Questions 79-85 (Mr Bacon) and Questions 166-168 (Mr Trickett):
transaction costs incurred by London Underground and the PPP bidders
The additional information is provided by the National Audit
Office in the form of three tables, together with notes and commentary,
including the Department for Transport's commentary (where indicated).
1. The first table[38]
provides, in one place as requested by Mr Bacon (Question 80),
a chart of transaction costs with further breakdown on the composition
of bidders' costs. The commentary following the first table describes
the scope of work undertaken by the lawyers acting for London
Underground and the scope of work undertaken by Price Waterhouse
(subsequently PricewaterhouseCoopers) as financial advisers.
2. The second table[39]
provides supplementary information on additional firms used by
London Underground. The named firms account for some £18
million out of an aggregate expenditure of £26.5 million
over six years, during the corporate reorganisation that preceded
the PPPs and during the procurement of the three concessions.
A further £4.5 million was not identified to specific firms,
covering items such as the "Project Office Costs" and
"External Contractors", and the balance of £4 million
was spread across an additional 50 to 60 organisations and individuals.
3. The third table[40]
provides supplementary information on the unsuccessful bidders'
costs, naming the principal firms of lawyers and consultants involved.
The supplementary information is cross-referenced to figures
in the Comptroller and Auditor General's Report "London Underground
PPP: Were they good deals?" (HC 645). In that Report Figure
12 sets out the external advisers' costs and Figures 13 and 13a
on pages 30 and 31 chart the progression of London Underground's
costs over time.
Question 166 (Mr Trickett): He asked for an explanation
of why Tube Lines, bidding for one contract, had been reimbursed
more than Metronet which had bid for two contracts. The reasons
why Tube Lines incurred greater costs (at £134 million) than
Metronet (at £116 million) have been provided by the Department.
Tube Lines' bid was the most advanced during the later stages
of the transaction process. As the contracts were negotiated,
developed and completed earlier they incurred greater costs in
agreeing contract principles than Metronet. This work was incorporated
into the two remaining contracts thus reducing transaction costs
on the Metronet bids where agreed terms could be incorporated
from the Tube Lines deal. Further financial savings were also
available from having a single bidder for the two final contracts
which avoided some inevitable duplication in costs had separate
bidders been selected.
There are two main components within the bids as paid where
Tube Lines costs are substantially larger than Metronet. The first
of these is the reimbursement of £7 million that Tube Lines
received for their unsuccessful bid for the Sub Surface contract.
Metronet received no such fees, being successful in both its bids.
The other substantial difference is in pre-contract operational
activities and the creation of business plans ("Transition
Team resources"). These costs were some £21.5 million
for Tube Lines, compared to £2.4 million for Metronet, reflecting
the different make-up of the consortia. Tube Lines were required
to prepare supply contracts prior to the close of the deals and
built up resources capable of letting the necessary contracts
to deliver their performance obligations during their negotiations.
This required Tube Lines to undertake greater preparation than
Metronet to ensure that these arrangements were satisfactory before
the PPP deal could be completed. Metronet, on the other hand,
included their supply chain within the consortium, and were not
required to undertake a similar level of preparation before deal
close because contracts had already been arranged with the supply
chain.
This does not necessarily mean that the costs for Metronet
were less, because the Metronet supply chain will have incurred
substantial costs pricing their respective obligations and in
particular performing the necessary due diligence to be able to
offer fixed prices. But those costs would be included in the overall
fixed prices offered to Metronet by their suppliers. In this way
those costs are also contained within the Infrastructure Service
Charge paid by LUL, although not separately identifiable.
Without the PPPs, as the Accounting Officer told the Committee,
London Underground would have had to negotiate separately potentially
a very large number of other contracts. In most other contracts
bidders' costs are not separately identified but included within
the deal price.
Table 1
BREAKDOWN OF TRANSACTION COSTS SUMMARISED IN "LONDON
UNDERGROUND PPP:WERE THEY GOOD DEALS?" (HC 645)
| £ million | LUL
| Tube Lines
(JNP) | Metronet
(BCV
& SSL)
|
Freshfields legal advice & drafting—just over 50% in period 11
PricewaterhouseCoopers—nearly 70% in period 11
Ove Arup—75% before invitations to bid
PA Consulting & Andersen—over 90% pre bids2
Other firms (see Table 2)
LUL internal costs—60% in period 11
+ add back effect of discounting (see Figure 12 in NAO Report)
| 29.2
21.4
6.0
26.3
26.5
61.0
10.0
| | |
Lawyers for Bidders (Tube Lines Periods 1:2 £1.9 million: £11.7 million)
(Metronet both BCV and SSL bids for periods 1:2 £3.3 million: £10.7 million)
| | 13.6
| 14.0
|
| Advisers (including sums due on completion)
| | 13.1 | 16.4
|
| Lawyers for banks and funding bodies |
| 7.8 | 6.2 |
| Banks technical advisor, modelling, etc |
| 4.4 | 2.5 |
| Other 3rd party advisers' costs to date—tax, audit, VAT
| | 4.2 | 3.4
|
| Balance of 3rd party costs forecast for remaining period to close
| | 5.4 | 3.0
|
| Bid team resources (Metronet reduced for cap3)
| | 16.6 | 14.3
|
| Transition team resources4 |
| 21.5 | 2.4 |
| Project Office expenses | |
1.4 | 3.2 |
| Tube Lines unsuccessful SSL bid (see Table 3)
| | 7.0 | n/a
|
| Payments to sponsors (Metronet both BCV and SSL)
| | 39.0 | 50.6
|
| Unsuccessful bidders (see Table 3) |
25.0 | | |
| Total (£455 million) | 205.4
| 134 | 116 |
| | |
|
Source:
London Transport Board approval November 2002 and NAO analysis.
1 The first time period closes with selection of Preferred
Bidders (May and September 2001). Detailed break-down over time
in Figure 13 of the NAO Report.
2 Fees to Arthur Andersen and PA Consulting principally covered
reorganising London Underground into three Infracos and one operating
company.
3 Further Metronet bid costs after November 2002 were 50%
reimbursable subject to a £3 million cap—applied to
the bid team resources line. Final legal fees took Metronet's
out-turn to £117.9 million after completing their financing
in April 2003.
4 "Transition team" activities include a early operational
activities during the extended period before closing the transactions
and the creation of Business Plan.
COMMENTS ON
TABLE 1
Legal advice and drafting:
Work by Freshfields for London Regional Transport and the
Department for Transport is further described as follows:
— providing input to the overall design of the PPP
contracts including evaluation of the options and ensuring the
appropriate provisions were included in the GLA Act 1999 eg in
relation to tax, protection of railway assets, and the role of
the PPP Arbiter;
— supporting LUL in reorganising into four separate
businesses and the preparation of information to support the bid
process;
— development of PPP documentation in parallel with
design of the Infracos and LUL (and reworking the documentation
as elements of the design changed or developed);
— supporting negotiations with Railtrack;
— restructuring of the Northern Line Train Services
Contracts in preparation for transfer to Infraco JNP and supporting
protracted negotiations with Alstom;
— supporting LUL's negotiations with PFI contractors
in relation to transfer of their contracts to TfL;
— supporting the tender process and LUL's negotiations
with all bidders;
— advising on judicial reviews; and
— advising on the competition aspects of the transaction
including state aid clearance.
Financial advice and negotiating:
Work by Price Waterhouse, and subsequently by PricewaterhouseCoopers
included the following:
— review with London Underground of the procurement
options open to London Underground and detailed presentation of
alternative models to Government;
— providing input to the overall design of the PPP;
— development of the PPP structure, including design
of the Infracos and LUL interfaces;
— supporting LUL in reorganising into four separate
businesses and the preparation of information to support the bid
process;
— development of the London Underground output requirements
and performance measurement regime;
— continual updates on LUL performance and determining
benchmark levels of performance;
— review of the PFI arrangements and their interfaces
with the PPP;
— supporting negotiations with Railtrack;
— supporting the tender process and LUL's negotiations
with all bidders;
— financial review of bids in conjunction with LUL
bid evaluation team; and
— assisting LUL with financial close and interest
rate risk management.
Table 2
OTHER FIRMS USED BY LONDON UNDERGROUND
| Company | Spending levels
| Activity |
| Up to £100,000 |
| |
Arnold & Boston | £20,000
| Architects |
| Symonds | Over £20,000
| Project management—Windsor House |
| "Mystery shopper" | £30,000
| Setting up and testing ambience measurement
|
| Cazaley | £30,000 |
Construction work—Windsor House |
| Dearle & Henderson | £50,000
| Cost management (asset knowledge) |
| Aslan | Over £50,000 |
Support on engineering standards work |
| Kvaerner | Over £50,000
| Engineering support (asset knowledge) |
| Eurica consulting | £60,000
| Consulting |
| Touchstone Renard | £60,000
| Technical review of asset data |
| Soil Mechanics | Over £80,000
| Survey information |
Between £100,000 and £1 million
| | |
Harris & Porter | Over £100,000
| Project management—accommodation |
| Haswell | Over £100,000
| Bid evaluation—technical support |
| Kennedy & Donkin | Over £100,000
| Engineering support (asset knowledge) |
| Mellish & Lynch | Over £100,000
| Quantity Surveyor—support on cost databases
|
| Osprey | Over £100,000
| Project management |
| Steer Davies Gleave | Over £100,000
| Customer survey |
| TFPL Ltd. | Over £100,000
| Reorganisation of deeds room, etc |
| William Dick Partnership | Over £100,000
| Commercial managers—Windsor House |
| WS Atkins | Over £100,000
| Eng support (asset knowledge) |
| Bacon & Woodrow | £150,000
| Pensions services and advice |
| Nigel Rose and Partners | Over £150,000
| Bid evaluation—quantity surveyors |
| Decision Focus | Over £170,000
| Modelling early PPP structural options |
| Slaughter and May | Over £200,000
| Legal services |
Royal Bank of Canada/
Ernst & Young
| Over £200,000 | Advice on interest rate hedging strategy
|
| Opteama | Over £200,000
| Engineering consultants (process mapping) |
| Metro Consulting | Over £250,000
| Bid evaluation—train operations |
| Computercenter Limited | Over £300,000
| Computers and accessories |
| ITNET | Over £400,000 |
IT services |
| Cap Gemini | Over £400,000
| Information technology services |
| AEA (and Risk Solutions) | Over £500,000
| Design/test safety management structure |
| Marsh (formerly Sedgwick) | £520,000
| Insurance services and advice |
| Hobley Nesbit | Over £570,000
| Commercial/management support to H&H |
| PML Ltd. | Over £700,000
| Bid preparation and checking |
| Field, Fisher, Waterhouse | £750,000
| Legal advice to LRT on transfer to TfL |
Over £1 million |
| |
Bank and legal fees | Over £1 million
| PPP interface with Northern Line PFI deal |
| Richard Ellis | Over £1 million
| Property services and advice |
| ICL (and others) | Over £1 million
| IT services |
| Overbury | Over £1.2 million
| Office refurbishment |
Pugh Roberts
(to end March 2001) |
£2,200,000 | Dynamic Simulation Modelling
|
| KPMG | £2,400,000 |
Audit and review of Public Sector Comparator methodology
|
| Hornagold & Hills | Over £2.5 million
| Project management and related services |
| | |
Source:
Project records provided to the NAO support the stated estimates.
Precise figures could be compiled, but only after confirming the
amounts with each individual firm to eliminate errors in data
interpretation, such as possible double counting.
Table 3
UNSUCCESSFUL BIDDERS COSTS
| £ million |
Tuberail (JNP)
|
LINC (BCV and SSL) | Tube Lines SSL Group
(SSL)
|
| | |
|
| Internal costs | 4.9 | 10.9
| 5.4 |
| External costs—legal | 3.7
| 2.0 | 1.2 |
| External costs—other | 2.4
| 6.9 | 1.5 |
| Total Bid | 11
| 19.8 | 8.1 |
| Total Paid | 11 | 14
| 7 |
| | |
|
Internal Costs for all bidders
Staff Costs, including accommodation, clerical support, expenses
and printing.
Tuberail
External Costs
Includes Asset Management, Financial, Legal, Technical, Public
Relations, Human Resources, Risk Assessment and other Due Diligence.
Lead firms include Linklaters, Allen & Overy, Arthur
Anderson, S&P and Moody's.
Amount Paid
Following a claim by Tuberail in September 2002, the London
Transport Board agreed to full payment of the £11 million
requested. This was based on accepting that Tuberail had diligently
followed London Underground's instructions during a prolonged
bidding process.
LINC
External Costs
Lead firms include Schroder Salomon Smith Barney (part of
Citigroup), Slaughter and May, Clifford Chance, Ernst & Young,
Symonds Group, Steer Davis Gleave and CJ Associates.
Amount Paid
£7 million was paid for each unsuccessful bid making
a total of £14 million.
Sub Surface Lines (Tube Lines Group)
External Costs
Includes LSA (public relations), AON (insurance advice),
APM Ltd (asset management consultancy), Capita (technical advice),
Chapman & Company (Property advice), Claire Howard Consultancies
(Property), KPMG (Tax), Lovells (legal advice), Norton Rose (lenders'
legal advice), PKF (audit of financial model), Scott Wilson Kirkpatrick
(lenders' technical advisor), VST Comeco Rail Ltd (technical advice).
Amount Paid
As Tube Lines owed Tuberail £7 million for their unsuccessful
bid for Infraco JNP, Metronet paid the £7 million they owed
Tube Lines direct to Tuberail.
37
The Department retained Ernst & Young to provide independent
advice to the Secretary of State. Back
38
See Ev 36-39 Back
39
See Ev 37-38 Back
40
20 See Ev 38-39 Back
| 
