Train procurement takes a different track

Direct state intervention to procure rolling stock is changing the model of how the rail industry orders new passenger trains. 21CR reviews how the market has evolved since privatisation to its current state.

20 years ago the railways were privatised: However, unlike other important milestones in rail history, this event was not marked by a one-off ‘big bang’ – there was no overnight transformation from an old order to a newer one  – as had been the case with the 1948 Nationalisation or the 1923 Grouping. Instead, rail privatisation was characterised by a steady succession of smaller bangs of varying sizes spread out over several years as the industry was sold off on a piecemeal basis.

While there has been much comment and analysis about Network Rail’s role and on other issues, rolling stock procurement has received far less attention. Train operating companies (TOCs) do not own their rolling stock but lease it from one of the three rolling stock companies (ROSCOs). For some TOCs  rolling stock charges can be greater than the track access charges they pay. However, track access charges are tightly controlled and require the approval of the Rail Regulator, whereas rolling stock leasing charges are not.

When the industry was privatised in 1994, it was anticipated that rolling stock provision could be left to the market, for the TOCs and ROSCOs to best negotiate. Events have taken a different course, and the creeping micro-management of the rail industry now appears to spreading to the rolling stock sector. It’s common knowledge that the DfT has long been dissatisfied with the leasing charges TOCS pay for their trains – and as paymaster responsible for funding the passenger franchises – their are legitimate grounds for concern.

The problem is that there is now not one elephant in the room, but three. These take the take form of the three publically-procured mega rolling stock contracts – Intercity Express (IEP), Thameslink and the Crossrail projects. In terms of size and complexity these contracts dwarf everything that has gone before, and they provide the government with the pretext to intervene in the market in a way never that was contemplated at privatisation.

While the DfT is keen to emphasise that these three contracts are in response to ‘exceptional’ cases, they do set an important precedent; one thing is certain though; the size of these orders alone will substantially reduce the market share of the leasing market currently enjoyed by the three ROSCOs and change it in the long term.

Pie-charts showing: (l) age of existing fleet; and (r) its ownership.
Breakdown:(l) by age of existing fleet and (r) by ownership at the end of CP4 (Control Period 4, March 2014).Total passenger fleet  consisted of 12,647 vehicles.
Breakdown by vehicle type: (l) 2014 and (r) in 2043
Breakdown by vehicle type: (l) 2014 and (r) in 2043.


Over 11,000 passenger vehicles were divided (roughly equally) between the three ROSCOs – Angel, Eversholt and Porterbrook – at privatisation. The more numerous fleet types were split between the ROSCOs, but the smaller-sized fleets tended to go to single leasing companies.  Nearly 6,000 new carriages have been delivered since 1994 and they now make up nearly half the total fleet.

Angel Trains is the market leader with a 33% market share, followed by Porterbrook (31%) and Eversholt (27%). Angel is the only company to supply all 19 franchised TOCs (and the two open access operators, Hull Trains and Grand Central, as well).

However, the three ROSCOs do not enjoy a complete monopoly: (bank-owned) Voyager Leasing supplies 59 ‘Voyager’ sets to Arriva Cross Country and another 18 units to Virgin West Coast; QW Leasing supplies 57 class 378s to London Overground; and Macquarie Leasing supplies 30 Class 379s for Abellio Greater Anglia’s Stansted services. BAA owns both the (open access) Heathrow Express and Heathrow Connect stock.

However, all these classes represent new (i.e. post privatisation) stock and remain with the first user. Most other stock remains under the ownership of the three ROSCOs, the main exceptions being First Great Western (which acquired some HST power cars and coaches from Porterbrook), and Arriva Trains Wales (which bought carriages for its loco-hauled services).

Average rolling stock age increased from 15 years (in 2007/08) to 18.5 years (in 2012/13), reversing a previous downward trend. This figure masks huge variations between operators: First TransPennine has an average age of less than eight years as against over 24 for Northern Trains.

All three ROSCOs have been resold – several times over  – since they were set up and they have realised princely sums for their former owners:  Angel Trains was originally bought for £696m in 1996 and sold for £3.6bn in 2008; Eversholt went for £518m (sold for £2.1bn in 2008); Porterbrook for £528m (sold for £1.4bn in 2010).

The book value of the assets upon privatisation was estimated at £2bn. The government of the day said it represented good value for money, but arch-critic Christian Wolmar dubbed it, “a flawed sale … the greatest scandal of privatisation in terms of loss to the taxpayer.”

DfT intervention

DfT intervention is nothing new: The Department already can influence the type of rolling stock a train operator may use, as well as determining the cascade process.  However, procurement of rolling stock itself is a new departure, though not entirely unexpected. The House of Commons Transport Committee reported in July 2008 that: “It is entirely appropriate that strategic decisions about rolling stock procurement and specification should be taken centrally. Given the level of fragmentation of the industry, there is no other way to ensure sensible use of taxpayers’ money for long term investments.”

Indeed, the Department even set up a company called Diesel Trains Ltd in 2008 to purchase 202 diesel sets, but abandoned the project when GWML electrification got the green light in the following year.

Basically, the current DfT plan is to establish mini-ROSCOs to lease out the IEP and Thameslink stock to the train operators.

Hitachi-led consortium Agility Trains will supply the IEP stock, and Siemens-led consortium Cross London Trains will supply the new Thameslink stock.

The IEP stock (designated classes 800 and 801) is to replace the ageing HSTs on the East Coast and Great Western main lines: 369 new vehicles are for GWML delivery in 2017/18 (costing £2.5bn); and 497 for ECML delivery in 2018/19 (costing £2.2bn). The Thameslink order is for 1,140 vehicles (Class 700) for 2015/18 (costing £1.8bn).

These costs (at 2014 prices) also include depots and servicing, etc.

This intervention has been the source of some concern in the industry. Presenting evidence to The National Audit Office (Procuring New Trains, July 2014), the DfTsaid: “The Department’s stated policy is that it will leave future train procurement to the industry, although it reserves the right to intervene where scale and complexity make it necessary.” It added that the Thameslink procurement was particularly complex as it involved three different train operators (which since September have been combined).

Under the terms of the contracts, “the rolling stock owners will only be paid once the train operator receives the trains ready for use. Payments will be reduced if they fail to meet demanding performance and reliability criteria (which are not features of standard rolling stock contracts).  Moreover, the rolling stock leasing companies will still own the trains after the end of the contracts, so they will have an incentive to maintain them to a high standard. This differs from standard PFI contracts where the asset reverts to public ownership when the contract ends.”

The rolling stock for Crossrail is also being directly procured, but it is being funded entirely by the public sector as well (private finance being initially considered but later discarded). This puts it in a class of its own. Bombardier won the £1bn contract to supply 650 (Class 345) vehicles for delivery in 2017/18. FormerTransport Minister, Stephen Hammond, explained the rationale: “The decision reflects the unique circumstances that apply to Crossrail. As a new route that is currently under construction it has no inherited train fleet and without new trains the service cannot open. TfL and the Government believe this decision to be the right one.”

Rolling stock charges

TOCs fork out almost as much in train leasing rentals to the rolling stock companies as they do in track access charges to Network Rail. The latest ORR statistics (GB Rail Industry Financial Information 2012-13, published in April) tells an interesting tale: Total operating expenditure (for all 19 TOCs) for the last financial year came to £9.529bn, of which £1.460bn (15.3%) was spent on rolling charges, as against £1.679bn (17.6%) paid out in track access charges.

But these are only average figures – they mask huge variations between the operators: In six out of 19 cases – Virgin Trains, Arriva Cross Country, First TransPennine Express, Abellio Greater Anglia, c2c and London Overground – the TOCS paid out considerably more in rolling stock rentals than they did in track access charges. And all (bar Greater Anglia) possess exclusively modern (i.e. post-privatisation) fleets.

Virgin came out top: The company paid £302m in rolling stock charges, nearly twice its track access bill of £165m. This is almost six times the £53m train rental charged to rival East Coast Trains (which uses older BR stock). In fact, Virgin contributed more than one fifth of the total rolling stock bill (paid out by all 19 TOCs).

This seems to bear out the criticism that new trains are disproportionately expensive to hire.

When the industry was privatised in 1994, there was no train leasing market so BR and Hambros Bank devised something called ‘indifference pricing’ to kick start the system. The objective was to equalise the cost of hiring stock; the higher costs of newer trains would be offset by lower operating and maintenance costs. The costs of leasing out older stock was retrospectively  determined by this notional concept and adjusted to even out the difference so train operators would be ‘indifferent’ to the cost of the stock they hired. That was the theory, but things turned out differently.

Dissatisfaction with some features of the rolling stock market prompted DfT complaints to the ORR in 2006 alleging ‘prevention, restriction and distortion of competition.’  The case was taken up by the Competition Commission (Rolling Stock Market Leasing Investigation, 2009): Its recommendations included longer franchises (12/15 years or more); greater choice of rolling stock for franchise bidders; and removing the non discrimination codes (brought in when Porterbrook temporarily passed under Stagecoach control).

However, the Commission ruled out regulation, much to the DfT’s chagrin. It found  capital rentals charges for new stock generally tended to be 15/20% higher than for used vehicles. Switching was expensive: “A TOC proposing to introduce new rolling stock also faces substantial switching costs, including short-term lease premiums for incumbent rolling stock, staff training costs and route acceptance costs. There are also risks of delays and initial faults… A TOC proposing introduction of new rolling stock faces all these costs and risks in the context of franchises of around seven years in length but it will only reap a small proportion of the benefit.”

Rentals are high, partly because there is little spare capacity due to passenger growth: The off-lease stock fluctuates between 1.2% and 4.5% of the total passenger stock, but these tend to be older mark 2 and 3 locomotive hauled coaches.

The TOCs have little choice of the stock they use, and if that was not bad enough: “the majority of current passenger rolling stock leases contain ‘hell or high water’ clauses that require TOCs to make payments irrespective of the reliability of the rolling stock under all but exceptional circumstances.”

Alas, some of the most useful and interesting findings in the Commission’s report were removed on grounds of commercial confidentiality. Obtaining more detailed information about rolling stock leasing tends to be either very difficult or non existent.

The erratic production runs on the manufacturing side has also inflated costs that have been picked up by train companies: “The manufacturers represented by the Railway Industry Association suggest that up to 20% of procurement costs could have been saved between 1988 and 2010 if there had been continuity of orders,” says Network Rail (Passenger RUS, 2011) referring to the stop-go nature of the industry.

And ATOC (Rail Value for Money: Rolling Stock Whole Life Costs, 2011) notes that the ROSCOs have only limited powers: “Since their sale by their former bank owners over the last few years, they have generally taken on significant levels of debt which we understand are subject to banking covenants which set significant limits on lease pricing.”

TOCS pay almost as much leasing rentals to rolling stock companies as they do in track access charges to Network Rail.
TOCS pay almost as much leasing rentals to rolling stock companies as they do in track access charges to Network Rail.

Future Growth

What does the future hold?  In February, the cross-industry Rolling Stock Strategy Steering Group (RSSSG) published its updated long term forecasts for passenger rolling stock (Long Term Passenger Strategy for the Rail Industry). This body brings together Network Rail, the three ROSCOs and representatives from the TOCs:

“The total passenger fleet is forecast to grow by between 53% and 93% over the next 30 years, and the proportion that will use electric traction is forecast to rise from 69% today to over 90% in all scenarios,” it says. Between 13,000 and 19,000 new trains will be needed. Over 3,000 new vehicles will be delivered under Control Period 5 (2014-2019) but most are for the IEP, Thameslink and Crossrail projects already mentioned. Under CP5 production will be tripled to 12 vehicles a week (as against four under CP4).

RSSSG does not address the cascade issue – of what will become of the older stock when it is displaced. This issue is still to be resolved.

The good news is that news trains are on the way; the bad news is that we may have to wait longer for them.


657/Oct 14




Leave a Reply

Your email address will not be published. Required fields are marked *