Return to Archive -by date - by topic.

INFORMED SOURCES August 2006

DfT nukes the ROSCOs

When a minister won't debate and DfT claims train rentals aren't ‘fair' my antennae start twitching

Here I am at Brush Traction on 6 July, sheltering from the rain, interviewing Transport Minister Derek Twigg who has just been inspecting the first series production re-engineered IC125 power car.

‘Minister', I start, ‘that power car costs£3000 per month to rent, Angel have just spent over £300,000 on reengineering, so what on earth are you playing at going to ORR about the leasing market'?

‘I've said all I can on that' he replies, ‘you can read our statement on the web-site', adding ‘it's business as usual, investment in the industry'.

Try again. ‘But they're not going to invest if there's all this uncertainty around, I just can't understand why.'

Mr Twigg cuts in, ‘I can very confidently say investment will continue to take place, there is record investment going on in the industry as you know, it's business as usual, this is an example of the investment that's taking place'.

That was just eight days after DfT Rail referred the issue of the capital rentals charged by the Rolling Stock companies (ROSCOs) for the inherited British Rail fleets to the Office of Rail Regulation. In railspeak, the ex-BR traction and rolling stock is known as the MOLA fleet after the Master Operating Lease Agreement under which it was sold.

Indifferent

Those responsible for privatising traction and rolling stock thought long term. While creating artificial lease rentals for everything from Class 101 diesel multiple units to Networker EMUs, they also had to ensure that the rentals would not price new trains out of the market, blocking investment.

Bankers Hambros and leasing specialists GATX came up with indifference pricing, which was first explained in this column back in December 1993. What indifference pricing did was, first estimate the total annual cost of ownership – financial, engineering and operational - for a notional new train; next take the actual cost of ownership of its BR equivalent; and finally, give the BR train a capital lease rental which made the cost of ownership roughly the same.

Note the distinction between the capital element of the lease rental and the total rental, which includes allowance for engineering costs, such as heavy overhauls, paid for by the ROSCO. DfT Rail assures me that its complaint to the ORR is concerned solely with the lack of transparency in the capital lease rentals.

When the ROSCOs were sold first time round, potential buyers based their offers on the discounted net present value of the income from the initial 10 year leases running from April 1994. To this they added an allowance for expected income from the newer train, some of which would remain in service to 2020 and even longer.

Risk

All three ROSCOs were sold to management buyouts, because the major banks regarded the business as too risky. But as the industry realised that the Labour party's threats to reverse privatisation were mere rhetoric, the ROSCOs were sold on, making a number of my chums millionaires.

By the time the present owners were brave enough to buy, the guaranteed cash flow of the initial 10 year leases had been securitised. This time, the value had to be based on the discounted Net Present Values of the long term lease rentals from the MOLA fleet.

This was a different sort of risk to the first time round. You looked at the current rentals to 2004, already spoken for financially, then worked out what the rentals would be when leases were renewed and for how long the trains would remain on lease. This determined your base offer, plus an allowance for the long run profit to be made from buying new trains to lease.

ROSCOs are all about operating leases and the acceptance of risk. A ROSCO buys a new train and leases it for a set period: the longer the period, the lower the rate of return charged on the capital invested in the train.

But if you buy a train and lease it for the whole of its life, it becomes a finance lease, the risk goes and the tax treatment changes. Some trains, the new Hitachi electric multiple units for Channel tunnel Rail Link domestic services for example, are so route specific that they might as well be acquired directly by Government on a finance lease.

But with an operating lease, setting that initial rental is clever stuff because you have to decide how much residual value risk you are prepared to take. Remember, you have borrowed money to buy the train, which you are paying back with interest. And with an asset like a train with a nominal 30 year life it is going to take a long time to get your money back.

As a rule of thumb, a new train does not start earning more than the ROSCO is paying out on the capital until around 22-23 years into its service life. Another rule of thumb is that with a 30 year life the annual capital rental is around 6.5% of the cost of the train, 7.5% for a 25 year life and 8.5% for 20 years.

So in today's competitive market for new train leasing, the capital rental a ROSCO offers depends on how long it thinks the train fleet will remain in service after the initial lease expires. This subsequent potential earning power is called residual value and your estimate determines the lease rental.

Now, if you are leasing new trains to a train operator with a seven year franchise, the residual value risk is high and the capital rental will reflect this. Of course, the next franchisee might keep the trains – or they might not. Porterbrook will have caught a massive cold on the Class 458 for example.

For Angel, the residual risk on the SWT Desiro fleet is equally real. To get round this DfT can provide you with what is called Section 54 protection, which guarantees that in the event of a change of franchisee, the incoming operator will continue to lease the trains. This, of course, reduces residual value risk and with it the capital rentals and the cost of the railway to the taxpayer and fare payer.

MOLA

What has this got to do with the MOLA fleets. Well, the banks borrowed the money on the assumption of future rentals, adjusted for residual value risk. Because of this indifference pricing is rolling on into the 21 st Century. But not if DfT has anything to do with it.

DfT reminds me that the documentation covering the original sale of the ROSCOs assumed that indifference pricing would apply only until the MOLA fleet leases were renewed in 2004. Then a competitive market would develop. There was a fatal flaw in this theory.

For a market to develop you need a surplus of supply over demand. This is created by new trains putting older trains with some residual life onto the second hand market.

Unfortunately, the second hand trains were predominantly Mk 1 slam door stock which, as soon as it came off lease went for scrap. Add in the lack of a traction and rolling stock policy from DfT and today, the only market exists is round the margins.

Even worse, the Strategic Rail Authority was a bit slow in waking up to the fact that the initial leases were coming up for renewal in April 2004. So, far from negotiating new terms or trying to create an ersatz market, there was barely enough time to renew the existing leases. That the trains were running on 1 April 1994 was largely down to a series of heroic all-nighters by the ROSCO commercial staff and their lawyers.

Despite its own sins of omission DfT had the brass neck to state in the Future of Rail White Paper, published in July 2004, that the competitive market for rolling stock leases ‘had not materialised as expected'. Against a fusillade of closing stable doors, the White Paper argued that there was a case for investigating ways of improving the operation of the market.

Then, in the Spring of 1995 DfT Rail called in all three ROSCOs for ‘discussions' on the MOLA capital rentals. According to informed sources the tone of the meeting was ‘bad tempered'.

After that the policy switched to divide and conquer, with bilateral meetings with individual ROSCOs. An early gambit was that if the ROSCOs did not agree to reduce rentals on their MOLA fleets by a collective £100m a year they would be referred to ORR. At the same time officials asked for data on rentals and rates of return.

Since tsuch data could be used as part of the threatened referral, reactions were mixed. One ROSCO said it would willingly provide all the data to the competition authorities if the referral went ahead. Another is understood to have complied – and a fat lot collaboration has done them.

As reported in this column, (archive on line in Alycidon Rail at www.alycidon.com ), DfT Rail tried to force the issue through a series of deadlines for the ROSCOs to agree to cough up. During June this descended into farce with deadlines passing and new ones being imposed on a weekly basis. Eventually DfT Rail, and its accounting officers, were in so deep that there was no alternative but to pass the buck to ORR.

Central to this issue is that suspiciously round number £100 million. It has appeared in the various press leaks over the last year or so and informed sources at the ROSCOs confirm that it is the number DfT Rail uses.

But the only way it could be a real figure would be if DfT Rail had first analysed individual capital rentals within the MOLA fleets the assumptions behind them in the original indifference price, then created a simulated market for eaqch class of train and come up with new rentals. DfT concedes that market forces could lead to some MOLA rentals being increased.

Unfair

Now I am sure that DfT Rail's advisors Ernst & Young are clever – but not that clever. This is confirmed by informed sources who claim that DfT Rail's final message to the ROSCOs was ‘give back £100 million a year and you can keep the MOLA capital rentals the same'.

And, in addition to the hapless Mr Twigg, I have been bouncing examples of rentals at DfT officials. But they decline to discuss specific cases and repeat that a ‘fundamental' lack of transparency about the way rentals are set means that DfT is not satisfied that the MOLA fleet capital rentals are ‘fair and competitive'.

One theory for the source of the iconic £100milion, which I find alluring, is that someone in the Department saw that the ROSCOs' total income from rentals was around £1 billion and said ‘I can get 10% out of that'. Now that a stab in the dark hasn't worked, requesting a market investigation at least gets the issue off DfT's desk. ‘Well we tried, but you know what ORR is like'.

Numbers

But what sort of numbers are we taking about? And what represents value for money?

Let's compare two diesel multiple units, the Class 158 from the MOLA fleet and the Class 170 Turbostar. Both offer air conditioned, 23 m long interurban vehicles.

For one TOC's fleet of Class 158 DMUs the capital element of the initial indifference pricing lease rental was £5,700 per vehicle per month. When the ROSCO was sold the expectation was that at the end of the first 10 year lease period, and assuming the lease was renewed for seven years, the capital element would fall to £5,000 per vehicle per month. After that, effectively up to the end of the nominal service life, £4500 per vehicle per month was predicted.

Bombardier's Class 170 Turbostar is the post privatisation equivalent of the Class 158. For one fleet of Turbostars the capital rental, based on the actual cost of finance in a competitive leasing market, is £7200 per vehicle per month.

Now these figures seem pretty transparent to me, with the older vehicle, having a realistically lower rental. But when I offered this comparison DfT Rail they wondered why I had chosen the two examples in that suspicious tone of voice which suggested that the evil ROSCOs had put me up to it.

Similarly, what should the capital rental for a Class 43 IC125 power car be. The current figure of £3000 per month compares with around £11,000 for a notional £2 million replacement with a high residual value. Add in the current expenditure of £300,000-£500,000 a throw on reengineering and that seems good value.

More numbers

Finally, we need to put the £100 million a year reduction in MOLA capital rentals into context. This is what it looks like on the back of an envelope.

Collectively the ROSCOs charge, roundly, £900 million a year. Since privatisation they have bought trains worth £4.8 billion. Assuming an average capital rental of 7.5% gives £360 million a year.

That leaves £540 million which also includes around £250million of non capital rentals - engineering costs and so on. Which means the capital rentals for the MOLA fleet are around £290 million a year – call it £300 million to allow for experimental error.

So the £100million represents cutting a third off current MOLA fleet rentals. ‘We might as well chuck the trains in the sea', said one informed ROSCO source, wearily

Investigation

DfT's case is that there is a lack of effective competition within MOLA market and lack of transparency as to how leasing charges are applied. These provide the grounds for a market investigation into leasing costs.

While DfT Rail ‘recognises' the ROSCOs' ‘important role' in investing over £4 billion in new trains through competitive tender, the Department adds that it also has a duty to ensure that best value is achieved for fare payers and government money. And from its analysis of ‘the available information' DfT Rail is not satisfied the prices charged for MOLA fleets are ‘fair and competitive'.

It also points to falls in interest rates since the MOLA rentals were set and that some risk, such as Mandatory Modifications have come back to Government.

That sounds convincing, until I quote examples and everyone runs a mile – even when off the record. And it seems I am not alone. According to one informed ROSCO ‘You can't have a logical conversation with the Transport Department about level of risk or rates of return, all they say is “give us some money”'

Clearly the ROSCOs expect a more sophisticated analysis from the ORR. Haydn Abbott, Managing Director, Angel Trains, the only ROSCO to put its head above the parapet, welcomed the inquiry ‘and the much needed clarity it will give for our investment decisions going forward'.

Of course, it is impossible to keep former Rail Regulator Tom Winsor out of a punch-up with the DfT. He is advising Angel and described the DfT Rail move as ‘an unjustified political assault on the most successful part of the privatised railway'. Pointing out that the contracts had been renewed three months before the White Paper was published, he added ‘it says something notable about the sanctity of contract with government if. Three months after supervising the signing of a contract, it says it may use its executive power to undermine or unstitch it.'

What next

ORR's initial market study will take three months. The simplest outcome is that it gives the MOLA fleet rentals a clean bill of health.

If ORR finds it all a bit complicated, (and I bet you can't wait for me to explain the Herfindahl-Hirschman

Index) with more detailed analysis needed on which to base a decision, it could decide to undertake a full market review. This would take a further six to twelve months. #

If the Study or the Review find that DfT% Rail has a case then ORR will make a reference to the Competition Commission for full scrutiny under the Competition Commission's market investigation powers.

There is yet another option, ORR could decide that while there is not enough evidence for e referral some tweaking is needed and as a competition authority could deal with this directly.

Sting

So, lots of material for future columns. But by tradition Informed Sources tries to sign off with a sting in the tail.

While DfT Rail's current target is the lack of competition and transparency within the MOLA fleets, looking ahead the Department is already hinting that a similar situation could occur when the leases on the post privatisation fleets come up for renewal.

Read what I have written above about returns and residual value and think what the possible imposition of lower rentals after 10 or 15 years would do to the money market. Perhaps the InterCity Express (formerly HST2) will have be acquired on a finance lease – but don't tell the Chancellor.

Continues.........