2.40 During the delay in closing the deal MOD chose to continue with bank finance as the method of financing in the absence of a clear case for change. The differential between the cost of bank and bond finance was volatile and if a decision had been made to change financing routes, MOD considered that by the time of financial close the differential may have moved the other way. MOD always considered that the deal was close to being closed and it did not wish to delay the deal further by changing the financing arrangements. MOD considered that it left the bank or bond finance decision as late as possible without causing further delays to the project. It therefore saw no clear durable advantage from changing the financing arrangements.
2.41 In hindsight, following price movements in the markets, bond financing may have provided a less expensive financing cost at the time of closing the deal in May 2000, but the impact on the value for money of the deal from using bond finance is uncertain.
2.42 Interest rates for bank finance became relatively more expensive compared to bond finance between June 1999 and May 2000. Two other government accommodation contracts were let at a similar time using bond finance: a £170 million bond was used to finance a contract let in May 2000 for the Treasury Building redevelopment and a £407 million bond was used to finance the contract let in June 2000 for the new GCHQ Building. There were nevertheless other bank-financed accommodation deals which were closing at the time such as UCL Hospital and Glasgow schools. We and MOD estimate that, had bond finance been achievable when the MOD Building deal was finalised in May 2000, it might have been between £1 million and £22 million5 cheaper than bank finance. But, because of uncertainties surrounding these estimates and how the markets at the time would have priced a bond for this deal, such an outcome from using bond finance, in respect of the pricing and timing of the deal, cannot be viewed with certainty.
2.43 There are various areas of uncertainty. It is possible that, in order to attract sufficient investors for three bond financed government deals, the timing of completing the MOD Building deal (or one of the other two deals) may have had to be changed and/or that the pricing of one or both of the deals may have moved unfavourably. The size of bond that would have been needed (£550 million), as the largest PFI bond to have come to the market and the complexity of the transaction, may have affected the ability to raise that level of bond finance or would have affected the terms on which it could be raised. MOD also has concerns that it may have had to agree to less favourable commercial terms on issues such as risk transfer to enable the bond to be underwritten.
2.44 As a general rule, departments should leave the choice of financing as late as possible so that they maximise their chances of the method of financing being optimal at the time the contract is signed. In this particular case, MOD, in seeking to close the deal without further delays, considers it left the choice of financing as late as possible. It continued to prefer bank finance for a number of qualitative factors including that it provided greater flexibility to cope with any necessary contract variations, and it had concerns about whether the bond was deliverable and likely to deliver the same risk transfer. For these reasons, and because of volatility in the markets and uncertainties about the terms on which bond finance could have been raised for this large deal, MOD saw no clear durable advantage from changing the financing arrangements. In the context of this deal, MOD considers, therefore, that it had good reason to continue with bank financing in finalising the letting of this contract.
_____________________________________________________________________________________
5 The upper end of this range is based on the saving that would have been achieved had it been possible for MOD to arrange bond finance on the same terms as that achieved in the GCHQ deal in June 2000. Both projects, in different ways, were highly complex large deals although the GCHQ financing was approximately £140 million less and was for new buildings rather than the redevelopment of existing buildings. The terms of the Treasury Building bond in May 2000, if applied to MOD's deal, would also have produced a similar level of savings but the terms were obtained on a considerably smaller redevelopment project. The contract period for the Treasury building was longer which also favoured a bond solution. The lower end of the range would have arisen if the pricing of the deal had had to be amended in a number of ways from the pricing achieved on the GCHQ and Treasury building deals due to the complexity, size, and length of contract.