FLAG Telecom is that rare thing, a frequent user of non-recourse finance in the telecoms industry. And as unease within the sector increases, other telecoms companies previously more comfortable with high-yield and corporate debt may do likewise. But for its North Asia Loop project, FLAG needed to come up with an innovative financing package that took into account its partner's markedly different finance strategy.

FLAG is raising $300 million to complete its share of the work on the Asian extension of its fibre-optic cable network. The project is a 10,000km six-fibre pair loop network that links Hong Kong and Tokyo. There are two legs (designed to create a back up in case one is interrupted), one going via Taipei in Taiwan, and one via Seoul in South Korea. The two staging posts are designed to serve the markets in the region with the most potential ? those where telecoms growth is expected to be the most robust.

FLAG's partner at the time of financing was Level 3 Communications ? best known for its staggering high-yield debt issuance in 1999. Level 3 proposed the North Asian Cable system as a means of extending its US and European network to Asia. It struck a deal with FLAG in January 2001 to create a dual cable loop system that would be stronger and capable of dealing with interruptions. Its side of the loop, linking Hong Kong and Japan via Taiwan, went into service in June 2001.

The financial structure was born out of this two-stage approach to construction, where the two owners oversaw separate construction efforts. FLAG used Alcatel for its FLAG West Asia Cable System, whilst Alcatel and Fujitsu are Level 3's contractors. Security for lenders and the fundamentals of the deal have, if anything, been enhanced by this structure.

In theory, the deal struck between the two owners took the form of a co-build agreement with an equal share of project costs contributed by the two parties. In practice, and to eliminate Level 3 credit risk, the deal took the form of a swap of fibre pairs (so that each owns three sets on the other's leg) and a $300 million presales agreement for capacity on the system.

This is an arrangement that FLAG understands. Presales have been an important component of every FLAG financing, and it has usually been able to treat these upfront cash payments as something approaching equity. Of the $800 million required to build out the cable, only $200 million will come from the sponsor's pockets. And, if past form is anything to go by, a combination of further capacity sales and rigorous cost control will further reduce the need to draw on this facility. Of the $600 million committed to the FLAG Atlantic-1 project (another Deal of the Year in 1999), only $300 million was drawn.

FLAG appointed its usual financing partner, WestLB, to lead the deal. Bank and sponsor put together a package that uses (in the words of WestLB) ?finely negotiated construction funding arrangements and a carefully crafted three-way legal agreement between the banks, Level 3 and the FLAG project company, which supplements the interlocking commercial contracts that underpin the co-build scheme.?

WestLB brought in Royal Bank of Scotland as a top-level underwriter and to manage documentation and took the titles of administrative and security agent itself. The deal was syndicated to ING, KBC and DGZ Dekabank. It is a 6.5-year facility priced at 225bp to 325bp over libor on drawn amounts. As part of the deal, FLAG Telecom Holdings Limited (the ultimate parent) will lend up to $100 million to its subsidiary, which would be reduced if the syndicate group were expanded.

This project presents a number of physical and commercial challenges ? the most obvious being that a telecoms map of the Asia-Pacific region resembles nothing so much as a generous serving of spaghetti. Capacity in the region is plentiful but Taiwan and Korea are often overlooked landing points. FLAG's original cable, running from the Western tip of the United Kingdom to Tokyo, is constrained by capacity that was put in place several years ago. Installation also needs to take into account the volatile geological profile of the region.

At least some of the risks associated with potential overcapacity have been mitigated by the presales contracts ? many of them with FLAG's main shareholders. Moreover, a thorough market study from Ovum suggests that the market can support the potential 2.5-3.8 Terabits of capacity that FLAG will offer. And FLAG is moving away from its initial mission statement of being a carriers' carrier to offering a greater variety of value-added products. These include managed bandwidth products and IP services.

However the story has not all been positive, since FLAG has been forced to abandon plans for a cable between Asia and the US. The project, which would have completed the Fibre-optic Link Around the Globe, was not financable in the current market with so many competing systems in play. Level 3 announced at the end of 2001 that it was selling its stake in the cable system to the Pacific Century CyberWorks/Telstra joint venture Reach. Nevertheless its presence in the region will continue through its capacity contracts, which underpin the economics of the project.

In the current climate for raising telecoms debt the structure will become increasingly appealing. As one source close to the deal explains, ?FLAG has been very conservatively financed and has made sure its projects are funded upfront. As a consequence it has not been left with partially complete networks. It is also aware of the need to conserve its cash reserves.? If joint venture agreements become more common and partners still require separate financing strategies, FLAG North Asia Loop will serve as a valuable template.