Space assets have often held a romantic allure for project finance bankers, probably because of the challenges of a structuring a deal around an asset that would require a space shuttle to conduct a site visit. This attitude persists despite some high-profile failures, mostly those associated with telecoms ventures. Skynet 5 is almost as curious - a secure communications satellite system for the UK's Ministry of Defence (MoD). But the strong reception for the £1.3 billion ($2.35 billion) deal reflects a solidly structured PFI deal as much as the spirit of James Bond at the bank meetings.

Skynet's route to close as a PFI deal is a microcosm of the state of the commercial space market. It has been a prospective private project since 1997, when the government conceived of a satellite system that would replace the MoD's current Skynet 4 satellites. These were approaching the end of their natural lives, and the ministry was looking for a more robust and larger service to handle communications with units overseas.

The initial bidding groups were Paradigm Secure Communications, a subsidiary of Astrium which initially was a joint venture between BAE Systems and EADS, the European defence conglomerate, and Rosetta, formed in 1998. Rosetta; advised by Bank of America, also featured BAE Systems, this time alongside Lockheed Martin. The dearth of expertise in the market explains BAE's ability to back both consortia, albeit using Chinese walls.

These precautions were ultimately unnecessary, since BAE decided to quit the business in February 2003, selling its 25% share of Astrium to EADS for £83 million, and it abandoned the edge that being the operator of the Skynet 4 system would have given it. As a result, EADS, which benefited from a combination of the Marconi (also part of Skynet 4), Matra, and DaimlerChrysler space divisions, among others, became the clear favourite.

But Paradigm, advised by Citigroup and funded by HSBC, BNP Paribas and CIBC, was named as preferred bidder at the beginning of 2003, before the departure of BAE. It subcontracts with EADS, BAE, Logica CMG and Cogent for the construction of two satellites and their systems and Serco, EADS, Logica CMG, Stratos, Cable & Wireless, General Dynamics and Qinetiq (the recently privatised defence research labs) for some IT and facilities management services.

The MoD chose the PPP route for the satellites because it believed that it would save the UK government £500 million over the life of the contract. These savings come in part from Astrium's ability to do the job at a lower cost over the whole life, but by also making the equipment cheaper through risk sharing. Paradigm takes on some operational risks, and also has the ability to make capacity open to third parties.

The third party revenue is significant enough to be a separate tranche of the bank deal. This third party revenue comes from contracting with other governments and military organisations for secure communications. The qualifications are that they must be vetted, and that they have a respectable enough credit rating to satisfy lenders. For this reason Nato, or any EU military force, will either need to be rated or have its obligations explicitly underwritten by member governments if it wishes to use the system.

The amount of risk that can be transferred to the MoD is difficult to quantify. The system is a bespoke piece of equipment that will need to be used under conditions of extreme security and most of it is out of the reach of any maintenance crews. Moreover, a satellite launch is a risky proposition - roughly one in six fails.

Since a rash of failures at the end of the 1990s, which afflicted the notorious Iridium deal, premiums for launches have increased. Lloyd's of London, one of the most active policy writers, has gained a good deal of experience in calculating failure rates, but rates are still high.

The PPP contract calls for Paradigm to take over the operation of the existing system, which it did in May 2003, and to source appropriate launch insurance. Should it be unable to find it, the government will backstop the sponsors. The government and Paradigm must also draw up a unique agreement laying out which party is responsible for a wide range and spectrum of system failures.

The three banks have put together an £880 million loan with a tenor of 13 years and four months, and the third-party revenue tranche of £83 million. The second tranche will be priced at 145bp over Libor towards the end of its life, although this is dependent on the quality of the offtaker. Pre-completion, both tranches carry an EADS guarantee, and thus 90bp pricing. The main tranche has upper pricing of 120bp.

The deal attracted a strong response in syndication, a 300% oversubscription, and interest from both PFI banks and EADS relationship banks. It points to untapped bank appetite for PFI assets with this premium - the sponsors of the FSTA aircraft deal should take note.

Paradigm Secure Communications

Status: closed 24 October 2003

Size: £1.3 billion

Description: Largest MoD project to date to operate existing Skynet 4 satellite network, upgrade it and launch the Skynet 5 system.

Concession awarder: UK Ministry of Defence

Concession: 15 years

Sponsor: EADS Astrium

Project debt: £963 million

Lead arrangers: BNP Paribas, CIBC World Markets and HSBC

Financial advisor to MoD: Deloitte & Touche

Financial advisor to sponsor: Citigroup

Technical advisor to lenders: Mott MacDonald

Insurance advisor to the lenders: Marsh

Insurance advisor to the sponsor: Willis

Lawyers to the project company: Freshfields

Lawyers to the sponsor: Lovells

Lawyers to the lenders: Allen & Overy

Lawyers to the concession awarder: Burges Salmon