PYRAMID RESEARCH Africa/Middle East Perspective
At the end of this year, the SAT-3 submarine cable along Africa's western coastline will become operational. Spearheaded by Telkom South Africa, SAT-3 is expected to provide an initial capacity of about 20 Gbps, which can be
expanded to 40 Gbps. Africa One, the proposed rival to SAT-3, intends to circle the continent with more than 25 landing points and provide an initial capacity of about 10 Gbps, extendible to 80 Gbps.
Exhibit 1 details the characteristics of these two major submarine cable projects. As argued in this analysis, the Africa One project faces challenges that are substantial enough to question its feasibility.
THE PYRAMID PERSPECTIVE
Demand for international bandwidth is on the rise
• The implementation of the SAT-3 and Africa One cables is designed to meet the rising needs of the African market for international band-width and reduce the continent's dependence on European traffic hubs. The explosion of the mobile market and the emergence of voice over IP have boosted both outgoing and incoming international traffic.
Recorded international traffic growth has grown to an annual average rate of about 17 percent since 1998, a pace it should maintain over the next two years. The rise in non-recorded traffic (most notably bypass traffic) suggests that actual annual growth rates are around 20 percent.
• For all the traffic growth associated with voice, data transport is arguably the most attractive application for international bandwidth service providers. Africa's international Internet bandwidth has been increasing exponentially, though nearly not fast enough to keep up
with pent-up demand. In South Africa, international Internet band-width has tripled annually over the past three years and should double again this year; Egypt's has quadrupled, and should grow by 1,000 percent this year as more ISPs link up to the FLAG submarine cable system and bandwidth costs are halved.
Africa One faces feasibility hurdles
• The hurdles to Africa One's implementation are plenty. First, Africa One is, logistically speaking, one of the most ambitious projects Africa has ever seen, intending to bring together about 20-30 African landing points. In a region where socio-political complexities have a history of creating logistical and administrative nightmares, bringing
together dozens of African carriers is an arduous task.
• More important, the project will have significant financing problems:
the extended telecoms slump has made it extremely difficult to raise funds for large-scale telecoms projects, particularly when there is little visibility for financial performance. With some of the major African carriers out of the game (financing participation in SAT-3 was difficult enough for them; participating in Africa One cannot even be
considered), Africa One has to raise the bulk of the needed $1.6 billion from private investors. Considering capital markets would be unrealistic, as would be an over-reliance on vendor finance. In the span of one year, Africa One's financing options have narrowed considerably.
Supply and demand fundamentals are concern…
• Supply and demand fundamentals are an additional concern for Africa One. Using a bottom-up, country-by-country forecast model developed for our Fixed Telecommunications White Paper, Africa's Bandwidth Glut, we anticipate that Africa's international bandwidth will rise tenfold over the next five years, from about 650 Mbps at the end of 2000 to more than six Gbps at the end of 2006 (see Exhibit 2).
While this growth is substantial, total demand would still be only 5 perent of the total bandwidth supply for African markets in 2006, and 25-50 percent of total supply over a 15-year period.
• In West Africa, Africa One would compete directly with SAT-3, and West African carriers are unlikely to back a project competing directly with one in which they have invested. In Southern Africa, the project lacks the backing of Telkom South Africa. This company's weight in the international traffic market means its participation is almost a sine qua non for success, at least in the short term.
In North Africa, fixed carriers can live without Africa One, as they are already served by three submarine cables. Pretty much the only region for which the project would be critical is East Africa (mainly Kenya and Tanzania), where there are no alternatives to satellite links.
…But Africa One is needed
• This is not to argue that Africa One is unnecessary. To the contrary, strong competition between the two systems would go a long way toward reducing international bandwidth costs in African markets, notwithstanding the substantial chasm between supply and demand. Without competition, incumbent telcos will control bandwidth costs, and they have a deplorable track record of doing so.
• Rather, our argument is that Africa One may simply be ahead of its time, relying on a competitive market model at a time when most African markets remain dominated by monopolistic behemoths. A deployment around 2004-05 would be better timing; by then, the critical markets (South Africa, Nigeria, Senegal and even North African markets) should be full of players that would be too happy to shake
off their dependence on incumbents for international bandwidth.
Guy Engon Zibi, Pyramid Research
This Perspective provides Pyramid’s view on a significant development in the communications industry. Perspectives are a component of Pyramid’s Advisory Services.
© 2001 The Economist Intelligence Unit Ltd. All rights reserved. Pyramid Research is a division of the Economist Intelligence Unit
© 2001 Mena Report (www.menareport.com)
