Skynet 5 takes PFI into space
To view the digital edition of this report please click here.
On October 24th 2003, Paradigm Secure Communications Limited (Paradigm) achieved financial close on the Skynet 5 PFI project creating an historic milestone for the UK Ministry of Defence (MoD) and the private sector. While PFI has delivered across a wide range of sectors, the provision of a military satellite based telecommunication service for the MoD is truly a step into new territory. The project, worth an estimated £2.5bn, is the MoD's largest PFI to date. By Andrew Rose and Jonathan Marlow of Canadian Imperial Bank of Commerce (CIBC), one of the three mandated lead arrangers.
Before proceeding into the history of Paradigm's successful bid, it is worth reviewing the requirements of the Skynet 5 Project.
The foundation of the Skynet 5 project is the delivery of a secure military telecommunications network in order to provide services to the UK MoD, with the provision to sell spare capacity to select foreign Governments and NATO. Prior to the successful delivery of Skynet 5, the MoD's requirements will continue to be met by Skynet 4, a constellation of satellites that are approaching the end of their useful lives. Skynet 5 is critical to the national security interest, and all parties are conscious of the importance of delivering the right solution.
The contract that governs the 15-year project is known as the Contract for implementation of Service Delivery (CISD) and this specifies the overall communications service that Paradigm must deliver to the MoD. This service is provided via a communication system which divides naturally into a space segment and a ground segment. The space segment consists of the satellites and their control systems, whilst the ground segment comprises the UK ground communications network including its management segment and the remote terminals.
The upgraded system will be required to deliver sufficient communications capacity to the MoD at a very high level of reliability, and a prescribed level of system diversity on the ground. The system will also have to be able to withstand specified levels of hostile stress and other extreme hazards.
The upgrade of the ground and space segments required by the CISD will occur in phases. Initially, Paradigm is required to direct and perform operation and maintenance of the existing Skynet 4 system to provide assured services, both in terms of a ground segment capacity and satellite communications links. Whilst this is on-going, Paradigm will upgrade the ground segment and deliver an initial set of remote terminals for use at sea and on land.
The next phase of the project requires Paradigm to operate and maintain the enhanced ground segment capability requirement, to complete the delivery of remote terminals, and to continue to provide satellite communications using the existing Skynet 4 satellites.
The design and commencement of manufacturing of the two new Skynet 5 satellites will occur during the initial two phases of the project. Launch, commissioning and testing is scheduled to be complete by March 2007 for the first Skynet 5 satellite, and by March 2008 for the second. Following successful delivery of the second satellite, the upgrade of the entire system will be complete. Paradigm will then be able to deliver the enhanced Skynet 5 services throughout the combined footprint covered by the two new satellites for the remainder of the contract duration until 2018.
Paradigm has subcontracted the design, construction and delivery into orbit of the Skynet 5 satellites to EADS Astrium Limited (Astrium), and the operation of the satellites and operation and maintenance of the ground segment to Paradigm Services Limited (ServiceCo). Paradigm, Astrium and Service Co are all wholly owned subsidiaries of European Aeronautic Defence and Space Company (EADS), the largest defence contractor in Europe and the 80% owner of aircraft manufacturer Airbus.
For CIBC, BNP Paribas, and HSBC, the Skynet 5 adventure began in the Autumn of 1997 when Matra Marconi Space, a joint venture company wholly owned by Lagardere SA of France and GEC PLC of the UK, approached potential lead arrangers for the financing of the project. Selection criteria included a proven project finance capability, industry specialisation in the telecommunications sector and a commitment to the then somewhat unproven Private Finance Initiative (PFI). In the days before debates over OGC Guidance on contract standardisation the Skynet 5 transaction seemed a daunting proposition.
There were many challenges to face.
- Was the MoD really committed to a PFI procurement solution?
- How could an asset with such clear National Security Interest implications be financed in the private sector?
- Would there be sufficient capacity in the banking market for such a transaction and would the telecommunications/defence combination reduce or increase the size of the potential bank market?
- What was the optimal level of risk transfer that would permit the MoD, the sponsors and the banks to craft a mutually acceptable deal?
In late 1997, all parties headed off into the unknown, optimistic that there would be a deal at the end of the rainbow. Over the next six years, the competition for Skynet 5 would be fierce and even the shareholding structure of the ultimately successful bidding entity would go through some profound changes. Paradigm initially had a complex web of shareholders and subcontractors. These would change fundamentally over the initial stages of the bid.
The competition was from Rosetta Global Communications, a consortium led by Lockheed Martin, British Aerospace and British Telecom.
The decision to proceed with a PFI solution resulted from the Strategic Defence Review White Paper issued in 1998. This Paper called for an improvement in MoD Procurement and PFI was ultimately selected as the appropriate procurement methodology for the Skynet 5 programme. It has subsequently been stated that the PFI solution will save the MoD circa £100m, and that real risk transfer has been achieved.
The challenge for the funders was to create a structure that enabled Paradigm to put in a competitive and hopefully winning bid, while ensuring that the finance package would be attractive to a wide range of lenders. By the time the Best and Final Offer (BAFO) bids were submitted in November 2001 the shareholder arrangements had changed, since British Aerospace (now BAE Systems) had acquired Marconi Electronic Systems from General Electric and as a result had become a significant shareholder in Paradigm. Appropriate firewalls were established within BAE, as they were now on both sides of the competition.
A significant level of due diligence was undertaken by the MLAs prior to the BAFO bid and strong letters of support backed Paradigm's bid.
After healthy levels of negotiation, the financing for the bid was formulated with the shareholders standing behind a significant element of the delivery and operational risk, the MoD agreeing to pay a pre-agreed level of take or pay revenue for satellite usage, and a broad risk share of the insurance provisions.
Paradigm was announced as preferred bidder in February 2002, and at this stage although CIBC, BNP Paribas and HSBC had worked with Paradigm since 1997 to structure the financing and support the bids it was seen as prudent to test the financing terms to re-establish that the banks' terms remained competitive. Accordingly, Citigroup, Paradigm's Financial Adviser, evaluated the potential of certain alternatives including a wrapped bond programme. The alternatives were presented to the MoD, who determined that there were many substantial benefits of staying with the bank solution. The economics of the bond were disadvantaged by the negative carry and this was exacerbated by the short (for PFI transactions) term of the concession, which matures in 2018. Additional value was given to the level of due diligence already undertaken by the Banks and the strong level of support achieved at BAFO.
With the senior debt funding method having been determined, the sponsors, MLAs, and the MoD continued to work closely in an effort to conclude negotiations. Before this point was reached, however, there was to be another fundamental change in the shareholding of Paradigm in January of 2003.
The satellite manufacturer Astrium, which set up Paradigm, was a joint venture, owned 75% by EADS and 25% by BAE Systems. In July 2002, the JV partners had announced the intention that EADS would purchase BAE's stake in Astrium. This created challenges in agreeing the appropriate contractual structure, as the preponderance of the manufacturing risk would then rest with EADS, and agreeing the appropriate levels of recourse to each shareholder at the Paradigm level was problematic.
This complex issue was simplified in January 2003 when a financial arrangement relating to Paradigm and Astrium was reached between EADS and BAE Systems, and BAE effectively withdrew from the Skynet 5 transaction, leaving Paradigm as a wholly owned subsidiary of EADS. A commercial agreement on most issues was finally achieved by April 2003.
Throughout the six year bid process financial structuring was critical in achieving the strong support provided by the bank group. In structuring the deal, the MLAs worked closely throughout with their principal advisers, which included Allen & Overy for legals, Marsh for Insurance and Mott MacDonald for technical. Due to the specialist nature of the work, certain subcontractors were required for specialist satellite knowledge.
Significant issues included the following:
- How could the banking market accept satellite design, manufacturing and launch risk?
- What level of offtake risk could be assumed?
- How would insurance be factored into the structure, particularly as the launches would occur years after financial close?
- What was the appropriate level of sponsor support?
These risks and the final position reached on them are analysed further below.
Extensive due diligence was undertaken on the programme for delivery of the assets as well as their ongoing operation and maintenance, and the lenders' advisers produced a comprehensive technical report. In addition to this due diligence, significant recourse to the subcontractors' ultimate shareholder was required to make the deal financeable, with EADS guaranteeing the performance of the obligations of Astrium under the system Prime Contract and ServiceCo under the Service Contract. Significant negotiation took place around the shareholders' financial obligation in the event of contractor default. This needed to be addressed with regard to the termination payments from the MoD.
The concept agreed between all parties was that MoD termination payments would increase only as the MoD received deliverables and value under the terms of the CISD, with the most notable deliverable being the first satellite, Skynet 5A. At this stage, the contractor default termination payment from the MoD would increase substantially in recognition of the successful delivery of Skynet 5A. This payment level rises further shortly after the delivery of Skynet 5B, and eventually rises to a level sufficient to cover Paradigm's full obligations to lenders 15 months thereafter. Commensurate with this escalating level of support from the MoD, recourse to the shareholder under a deficiency guarantee declined over time. From a structuring perspective, making sure that the resulting exposure to the lenders was manageable and financeable was a critical part of the analysis.
A key component of the transaction and the bid was the level of offtake risk the private
sector would take in addition to guaranteed MoD usage levels. Once again, timescales played a major part in the ability of the private sector to take risk.
Under the terms of the CISD, Paradigm was to deliver an assured level of capacity to the MoD, and if this level was delivered in accordance with the terms of the CISD, the MoD would agree a take-or-pay level of usage. This assured level intentionally left incremental capacity that could be sold to other Governments and NATO, collectively known as third party users. In order to make Paradigm's bid competitive, it was an objective to limit the level of take-or-pay revenue. This created significant issues for funders as an analysis of the Government military satellite demand in the period from 2007-2018 might have proved too challenging for many bank credit departments.
The solution to the problem was to create two tranches of debt, a £880m Facility A and a £83m Facility B. Sizing of these facilities had to take into account potential deductions for under performance and delays, and additionally in the case of Facility B MoD (above the take-or-pay level) and third party demand fluctuations. Conceptually, however, Facility A was to finance take-or-pay revenues and Facility B was for third party revenues and MoD usage above the take-or-pay level. Facility B required a level of shareholder support until a business case was established.
While the sizing and terms of these facilities proved hugely complex, the basic structure proved a mutually acceptable mechanism to address revenue risk.
In the CISD and related schedules, the MoD required a high level of insurance from Paradigm due to the nature of the asset, the launch risk, and the problems of having an operational asset in space. The lenders had broadly similar insurance requirements to those of the MoD, however the size and nature of the insurance package created added complexity.
The most significant issue was that the pivotal insured moment (launch) was 3-4 years beyond financial close. Given the size of the insurance market, the volatility of capacity and pricing levels in the market, and the time frames involved, it was impossible to arrange a full insurance package at financial close and many months were spent agreeing the appropriate risk share for the insurance cover.
CIBC was the insurance bank on the deal and led the negotiations on this subject. This required significant legal and insurance expertise as the negotiations covered availability of insurance, capacity in the market, non-vitiation, price of cover, cause of loss, and reinstatement of assets. Given the secure and remote nature of the assets, it was impossible to merely rely on OGC Guidance, and as we all agreed many times, "this is not a hospital we are talking about"! Fortunately all parties were open-minded in understanding who realistically could manage the risks, and the insurance risks were allocated between the insurance market, the lenders, the borrower, the shareholder and the MoD. The concept of uninsurability protection was broadly split between capital U Uninsurability where the MoD provided a level of protection, and little u uninsurability, where the uninsurability was caused by the actions of Paradigm, its shareholder or its principal subcontractors and hence the risks consequently lay in the private sector. CIBC had volunteered to be insurance bank, naively hoping for a light workload! It proved to be one of the most complex and demanding areas of negotiation.
It is often stated in PFI bank financing negotiations that "this is not rocket science", however Skynet 5 proved an exception to this rule!
The Skynet 5 solution is a highly complex one and the banks' technical analysis was conducted by Mott MacDonald, supported by specialist satellite technical experts. Despite the high level of due diligence, the complexity of the solution meant that a meaningful degree of sponsor support was required.
By the time that the final subcontract and shareholder support structure was in place Paradigm, Astrium, the system prime contractor (the builder of the satellites), and ServiceCo were all wholly owned by EADS.
It was hugely important to EADS that Paradigm had a reasonable degree of flexibility in the programme for delivering the satellites and the related communication network, and consequently it was realistic as to the level of shareholder support that would be required.
The system prime contract was to be delivered by EADS subsidiary, Astrium, and the service prime contractor, ServiceCo, would be responsible for delivery of operational service and maintenance. At EADS direction, the structure required Paradigm to take ultimate responsibility for the delivery of the output specification under the CISD.
The common ownership of the entities by EADS caused complexities, however it gave EADS the ability to manage the entire process and provide the appropriate level of shareholder support, particularly in relation to delivery risk, meaning that a bankable solution was achievable.
In addition to this support, the banks structured controls over the process by managing the loan drawdown process against progress under the system upgrade programme. Once again, the negotiated position needed to be one that permitted EADS to manage the process in a manner that enabled it to deliver a complicated solution in a timely manner, and one that gave lenders comfort that their loan was appropriately used. A complex web of potential draw-stops, milestone events, and milestone clusters were crafted so that incentives and risk mitigants were appropriately in place.
Approaching the market
The decision on the finance structure and the resolution of the shareholding structure gave the green light to the MLAs to complete the structuring of the senior finance, revert to their credit committees, and approach the bank syndication market.
Consistent with the Tube Lines and Metronet mega financings, the bank group was identified and approached prior to financial close. Once again, nothing was simple. The approaches were made at a time of hostilities in Iraq and while the SARS epidemic was prevalent, and these factors contributed to a high level of uncertainty over the state and future of the commercial aviation industry. Given the importance of this industry to EADS, banks had to take a view on the long-term implications of these events.
Additionally, the structure was viewed by the market as somewhat of a hybrid. Strong EADS support brought in key relationship banks, but the term of the debt (approximately 14 years) was too long for a corporate deal, but shorter than a conventional PFI deal, and the debt was more expensive than EADS corporate debt, but cheaper (in the construction phase) than a project financing. Certain banks were unfamiliar with PFI or had issue with either the telecommunications or the defence sectors. All of these issues created uncertainty around the demand for the loan in the bank market.
However, despite these complexities the transaction had many merits from the banks' perspective. The importance of the transaction for the MoD, and the key nature of the transaction for EADS, which had openly stated its intention to grow its space business, gave the Project a very strong foundation.
EADS' strong support prior to delivery of the satellites, and strong MoD support once they had received their key assets combined to present a robust structure.
After an initial market sounding, the terms of the deal were agreed by all parties and a bank presentation was held in London on June 18th with approximately 30 banks attending. The presentation was well supported by EADS senior management, which emphasised the importance of the transaction to the group. As a result of the robust structure and strong support, the transaction was very well supported, and a large bank syndicate of underwriters was assembled. After fairly lengthy negotiations on the final documents, and to all parties' relief, financial close occurred on October 24, 2003, almost six years to the day after the MLAs first reviewed the transaction.
PFI for major MoD equipment procurement raises unique challenges and the most important lesson is that the transfer of and/or placement of risk must be very carefully scrutinised and understood by all parties.
Insurance was just one key area, where there was no pre-determined model to follow and the negotiations focused on where risks realistically could be managed. Reinstatement of an impaired asset orbiting thousands of miles from Earth was not contemplated in any Guidance document, and there was a need for an innovative and constructive approach to risk allocation.
A high degree of due diligence early in the process was critical, so that all parties could be confident that the lenders would deliver on their commitments. With all the changes in the banking market, it is a credit to the MLAs' stability that the same three banks started and completed the process over such a long period. We may have all grown a little older, and there were several 'Skynet babies' born during the course of the negotiations, but many of the faces stayed much the same throughout the process, and the MLAs believe that Paradigm chose their financiers well!
The most important lesson in such a complex transaction was the need for constructive co-operation. It will surprise no-one to hear that there were a few fraught midnight sessions at law offices over the years, and the odd heated exchange of views, but it is the writers' views that all parties addressed the challenges with a common goal and a constructive approach to delivering a communications service that will serve
the UK MoD well for many years to come, and that will help EADS achieve its goal of growing its Space Division.
The list of contributors to Skynet 5 is too long to list, but certain individuals at certain firms will not forget the experience for a long time. So for the MoD, the DPA, EADS, Astrium, Paradigm and its long list of subcontractors, the MLAs and bank syndicate, Allen & Overy, Freshfields, Lovells, Burges Salmon, Citigroup, Deloitte and Touche, Mott MacDonald, Marsh, Willis and many others, we really have "boldly gone where no man has been before", at least in the world of PFI!