Why all the West Coast bids were wrong
I have learned a bit more about the mistakes made by the Department for Transport in overseeing the contest to win West Coast Main Line rail franchise.
As I understand it, the entire bidding process was flawed - and that all four of the bidders were given erroneous information by civil servants when preparing their bids.
The nature of the mistake, which I describe below, carries all sorts of implications.
One consequence is that all the bidders were offering far too little protection to taxpayers against the risks of collapse by a franchise holder. In other words, if the mistake had not been picked up, taxpayers would have been excessively exposed to potential losses at some point during the 15-year life of the franchise.
The nature of the department's errors, with all the bidders given flawed data, explains why the Department for Transport has said that it cannot be certain that the outcome of the bidding process would have been different if the process had been robust. In other words, FirstGroup might still have won.
That might make FirstGroup marginally less hostile to the idea that Virgin may be asked to keep running the service for an additional 18 months to two years, until a new tendering process can be completed and the new franchise holder can be in place - because there is no evidence that Virgin as the incumbent would have had an advantage if the last contest had been run properly.
A decision on whether the franchise will stay with Virgin for an interim period, or go into temporary public ownership, will be taken in the next few days.
That said, FirstGroup is considering suing the government for substantial damages resulting from the way it was first awarded and then stripped of the franchise. As it happens, those close to a number of the failed bidders tell me they expect the compensation payments from taxpayers to run to hundreds of millions of pounds, far more than the £40m that has been widely reported.
I am told that ministers and officials are desperately anxious to somehow dissuade FirstGroup from suing, although it is not clear why it would be in the interests of FirstGroup's shareholders to settle without the threat of legal action.
But back to what went wrong. All this is slightly technical and complicated, so bear with me.
When there are contests to award complicated long-term contracts such as this one, civil servants construct their own financial model. This contains certain assumptions about what is likely to happen to important economic variables, such as inflation and passenger numbers, over the contract period.
The model serves two purposes.
It helps civil servants and ministers evaluate bids against their own benchmarks, when those bids are submitted.
But perhaps more importantly, it helps the bidders to submit bids on a basis that allows them to be compared. It creates a level playing field for the bidders.
What happens is that the model produces what is known as a "ready reckoner" which is given to the bidders. This ready reckoner tells each bidder the financial implications of their respective forecasts of how much they think they can increase revenues.
In particular, it tells a bidder how much capital its so-called Train Operating Company (TOC) would have to hold as a protection for taxpayers against the risk that in the course of operating the franchise the TOC went bust or risked going bust.
The point is that any bidder could submit wildly optimistic and unrealistic forecasts for how much its TOC could increase its revenues in order to win the bid, and then get into lethal financial difficulties when actually operating the franchise on that basis.
So to minimise that risk, all the bidders are told that they have to endow the TOC with sufficient capital to protect taxpayers from potential losses.
In the case of FirstGroup, for example, its winning bid endowed its new TOC with £10m of capital from day one, and there was also a £190m subordinated loan from FirstGroup to the TOC and a £45m "performance bond".
All of which may sound like a fair amount of money to absorb the costs of an abrupt termination of the franchise.
But what the Department for Transport discovered in the last few days is that the capital protection being offered was far too little. That was not FirstGroup's fault. It was because of assumptions in the department's own model about what would happen to inflation and passenger numbers - which in turn meant that the ready reckoner given to bidders was wrong.
In fact, the department's mistakes meant that all the bids were offering far too little protection to taxpayers against the possibility of a franchise holder being unable to hold the franchise for the full 15 years.
The implications are serious for the whole process of tendering for rail franchises. It explains why the department has suspended the competitions to run the Great Western, Essex Thameside and Thameslink services, just in case the financial models underpinning those contests are flawed in a similar way.
One very interesting question is how much extra capital First Group and the other bidders would have been forced to hold if the department's model and ready reckoner had been correct.
Here is the big point. It is very expensive for companies to raise capital. It is theoretically possible that FirstGroup - or indeed any of the four bidders - would not have been able to afford the correct amount of capital.
That raises a further question - whether the government will have to revert to awarding eight-year franchises, rather than 15-year franchises, because in theory the capital required to underpin an eight-year contract should be much less.
Here is a tidbit which will add to the government's misery over the West Coast rail debacle.
Virgin tells me that they received the erroneous data from the Department for Transport's flawed ready reckoner and concluded that it was wrong - so they ignored it when submitting their own bid.
The current West Coast encumbent even went so far as to tell the Department for Transport that its numbers were awry.
But apparently the Department ignored Virgin and pressed ahead with the bidding contest.
For what it's worth, Virgin insists that the bid it submitted has a robust financial basis - and it believes it would have submitted the same bid, even if the department had got its own numbers correct.
On the other issue, of whether Virgin will continue to run the franchise for the two years before a new contact can come into force, or whether the West Coast line will go into temporary public ownership, my sense from talking to sources in government and among the bidders is that Virgin will prevail.
The main reason is that there are 150 West Coast contracts that are up for renewal between now and 9 December, when Virgin's current tenure expires. It would be quite a challenge for civil servants, with only a cursory knowledge of the business, to suddenly take over responsibility for those contracts.