For the last two years the Latin American telecoms market has been the financing equivalent of a sure thing. The high profile BCP Refinancing, as well as mirror concession financings such as Vesper, attracted rabid bank and vendor interest. It may, however, have seen its first big-ticket casualty, this one backed by a rock solid sponsor.

The Telefonica SAm-1 cable has been under construction now for some time. It is likely to avoid any concerted easement or landing rights obstructions apart from limited cross-country sections, and is one of the earlier of the new generation of ?carriers' carriers' in the region. The cable is backed by a $600 million bridge loan, assembled in July, provided by Deutsche Bank, Morgan Stanley Dean Witter, Citibank and WestLB. The signing of the six-month loan was followed rapidly by the opening of negotiations on a 7.5-year deal, to be in place by the end of the year.

This will probably not be signed now, after an announcement by the arrangers (the same group as the bridge), that syndication on the $1.3 billion long-term debt was to be deferred into the new year.

The reasons given are familiar ? and understandable. The sharp words from banking regulators about the extent of banks' telecoms exposures were one factor, but equally important have been the string of mediocre results from some long-distance telcos in dissuading banks from committing.

SAm-1's prospects

The deal was sensibly leveraged, and featured the usual element of pre-sales contracts. The debt broke down into a term loan of $1.25 billion and a revolving credit of $50 million. Equity commitments stood at $300 million whilst firm contracts guaranteeing to buy capacity added up to about $500 million, a more than adequate comfort level when compared with recent deals. Telefonica, however, wanting to free up its balance sheet for the 3G licence outlay ahead, will now be hoping that the bridge can be rolled over.

SAm-1 is Telefonica's first sub-sea transaction, and has been operated through the dedicated Emergia subsidiary. Emergia is designed to be a standalone carriers' carrier venture, and is open to competing telecoms players, but has been identified as a growth vehicle for the Spanish state telco. It roughly encircles the continent of South America, crossing land twice in Guatemala and in cutting of the southern tips of Chile and Argentina.

The cable has signed up a number of local carriers, including ventures controlled or part-owned by Telefonica. The mirror concession of Telesp, in Sao Paulo, is one, as are a series of long distance carriers in Chile, Argentina, Peru and Brazil. At present the company is trying to build up a healthy base of mobile subscribers ? now the strategy of choice for developing world telcos looking to build customer bases in countries with a low fixed-line penetration.

But data is set to be the greatest revenue generator this century. While voice traffic can be assumed to grow at around 14% per year, data and internet traffic is growing at between 70% and 200% per year, depending on the analyst (the former types tend to be driving force behind current stock market sentiment). Established operators such as Embratel (now owned by MCI Worldcom) have found it hard building out the necessary fibre-optic backbone quickly enough ? at least one reason why local electricity companies have become important telecoms players.

Telefonica's impatience at the pace gives another clue as to the importance of the SAm-1 cable. It was a shareholder in the ATLANTIS 2 cable, a joint venture of 25 carriers that linked up Portugal and the Brazilian coast. The deal was a consortium financing of the old school that used to dominate cable and satellite building, and Embratel, the leader of the group has arranged for two 40Gps cables between Fortaleza and Rio to be reserved for its own use. These are two of the four landing point chosen by Emergia for its Brazilian coverage.

Latin fibre-optic market tightens

Market studies, including those from Pioneer and Arthur D. Little, have demonstrated that the demand exists for fibre-optic capacity in the region. The new standalone cables, however, tend to find themselves competing for a larger slice of alternative and start-up carrier business than many bankers find attractive. The distinction is important when contracts for use are viewed as an integral, almost quasi-equity, part of a deal's make-up.

Strong PTTs can find a willing group of relationship banks to come in on deals on the back of promised bond deals. The less strong are shifted off balance sheets as fast as lenders can structure the transactions ? as a few vendor finance securitizations, with large Latin American elements, have demonstrated. Sources at Citibank, which recently added to the Alcatel receivables securitization programme, have admitted that the second tranche was a harder sell in the light of market sentiment.

SAm-1 will be selling the majority of its capacity to US carriers rather than local entities. This reflects the fact that much of the new traffic come from US players, belatedly entering the South American markets behind the Europeans. This explains why a simple relationship pull has not worked so well in Telefonica's situation, as well as the simple fact that despite the progress over the last 18 months, the list of banks with appetites for Latin risk is still a short one.

The one area where Latin America may enjoy a real advantage over other regions is that bandwidth prices appear to be holding solid for the time being. Michael Wynne, managing director for telecoms at WestLB Structured Finance, notes that whilst prices for capacity have dropped faster than expected in the North Atlantic, those to the south are not experiencing quite the same falls. ?You tend to assume a 20% decrease in capacity prices per year over time?, he says ?but over the long term revenues to increase with traffic volume?.

The reason for this paradox, he adds, is that cables can be cheaply and incrementally upgraded. Most of the actual dark fibre underwater stays in that condition, but the technology bolted on, whether the ever more complex ways of refracting the light that takes data down cables or the elimination of the need to reconvert this light to electrical pulses, is key. Cables with small initial capacities can receive upgrades at minimal capital cost. Moreover, most of the normal day-to-day expenditure can be passed on to the consumer ? even those companies holding indefeasible rights of use (IRUs) recognise operations and maintenance as pass-through costs.

The competition exists, however, especially from carriers' carriers that also attempt to offer products to large end-users, Global Crossing and 360networks in particular. Global crossing has had a tumultuous year, beginning with the purchase of Racal Telecom and its UK backbone and gorged upon the proceeds of high yield offerings and large corporate finance-type facilities from telecoms banks. Times have changed, exposure worries continue to gnaw, and Global Crossing found itself forced to sell its Globalcenter web-hosting business, a jewel that backed up the fibre-optic giant's claim to provide ?seamless connectivity? to the customer.

Global Crossing has been found returning to the joint venture route (with Hutchison) in Asia whilst its rival FLAG Telecom, buys out its partner GTS on the Atlantic-1 cable. The partnering up may suggest an end to the corporate finance route, but probably not. It recently retired the $800 million loan taken out for the Racal purchase and replaced it with one secured on its profitable Frontier CLEC in the US. More importantly, Global Crossing is now becoming a serious Latin player, and has announced since October alone the linking of Chile, Argentina and Brazil with its global network.

The other contender, 360networks, shows even less inclination to take the structured financing route. Its two most recent deals have been a $700 million standby facility provided by Donaldson Lufkin & Jenrette, and a $1.2 billion senior secured credit facility from DLJ, Chase Securities, Goldman Sachs and TD Securities. The latter has an extension provision for $800 million, subject to various coverage tests, but the CEO of 360, Greg Maffei (formerly of Microsoft) has indicated that the bulk of the funds, sourced to construct an Asian network, will be found from revenues.

360's Latin presence comes from the Atlantica-1 cable, the main asset of the GlobeNet Communications Group, acquired earlier in the year for $600 million and $400 million of debt. The cable itself was funded by an $865 million project loan from TD Securities. It has probably won the race to reach the continent since Atlantica-1, a 22,500-kilometer undersea network linking the United States, Bermuda, Brazil and Venezuela, is scheduled to enter service by the end of the year. A second submarine cable system from Fortaleza, Brazil to Rio de Janeiro with a terrestrial network to Sao Paulo is scheduled for service by February 2001.

The carrier's previous incarnation, WorldWide Fiber, had a tough time syndicating the debt for the Hibernia cable, which linked Newfoundland and Ireland. This found a market that had already taken its fill of the FLAG deal and did not feel that the new venture had quite the pedigree of the refinancing veteran that was FLAG. Head-to-heads rarely make participants happy, unless the two prospects can be shown to be complementary, or technologically different.

Arcos ? the new niche player

Such is the presentation assembled for the Arcos cable, presently syndicated through Barclays and the Bank of Montreal. Arcos is a ring system reaching round the Caribbean, the Gulf Coast and the northernmost part of South America. The cable is more of a niche project, since it links the smaller PTTs of the region and will have a different configuration to cables such as SAm-1. It marks a marriage of the PTTs and an aggressive new player, Global Light Telecom.

Global Light was formed under the aegis of GST Telecom, which filed for bankruptcy in Delaware in May 2000. Global Light was made up of former GST personnel, and GST held a substantial stake in the company. The two had a number of suits outstanding, mostly over the issue of alleged illegal transfers of technology, but this have been settled for some time. Global Light holds a controlling interest in New World Network Holdings, which in turn is majority sponsor of the cable.

It is New World Network Holdings' 87.5% interest in the cable that the bank deal is designed to fund, since the main sponsor has asked for contributions from the other backers of the cable, mostly landing parties for the cable, as equity of around $49 million. Other equity holders include Siemens Project Ventures and Barclays Capital. The last had originally been brought in to raise equity finance and now has a $170 million bridge loan on its books that is looking for a take-out in the high-yield market soon.

A further $165 million is now in the market for syndication, and the arrangers are fortunate that the sum required is dwarfed by SAm-1. The term loan of $150 million and the revolver of $15 million both carry a tenor of seven years and a tasty margin of 325bp over Libor. Commitment fees alone stand at 125bp for a lower than 50% use and 75bp for any level above that. Drawdown of the facility is linked to construction milestones, and the loan converts from a construction loan to a term loan in December 2001. It also has a prepay facility of 75% any excess cashflow, as has become standard on sub-sea deals.

The arrangers are not looking for any large commitments, although Barclays may need to return to the market over the next year or so if the project performs well but the high-yield market does not. This is a realistic scenario given the current game, doing the rounds in the banking sector, of denying large losses on telecoms high-yield bonds to investors. Success on the syndication will wait for allotments to be tied up, so that Siemens and Norddeutsche Seekabelwerken (part of Corning) can start work. They may yet avoid the cable laying ships of Tyco ? currently doing the same work for Emergia.