Barron's cont'd.
Iridium's chief rival, Globalstar, continues to lag some months behind. A
project dreamed up by Loral, Globalstar plans to start service in early 1999.
Globalstar's network will have 48 low-orbit satellites; eight have been
launched. Assuming no problems with its three Zenit satellite launches,
Globalstar will provide Iridium with tough competition. For one thing, the
system will be a lot less expensive, $2.7 billion, versus $4.4 billion for Iridium.
That will allow Globalstar to undercut Iridium's pricing -- figure 65 cents to $1
a minute at retail, versus $3. Globalstar's handsets will be cheaper, too,
running $750, rather than $3,000 for Iridium's.
"We've constructed a satellite system that's very simple, with all the
complexity of the processing done on the ground, not in the air," says Bernard
Schwartz, CEO of both Loral and Globalstar. "Iridium does
satellite-to-satellite cross-linking, which is expensive. They would argue they
gain some quality, but I would doubt anybody could discern that." Globalstar
will also have far more capacity. Schwartz says Globalstar can handle seven
to 10 million subscribers, versus five to six million for Iridium and ICO
combined.
Globalstar, though, has its own particular risks, not the least of which is their
above-mentioned dependence on the none-too-reliable Zenit rocket.
Assuming the Zenits don't cause a problem, Globalstar will have to establish
itself in the face of a marketing blitz from Iridium. Thomas Watts, an analyst at
Merrill Lynch, pointed out in a recent report that Globalstar will be far more
reliant than Iridium on local partners to market the service. Where Iridium will
spend a fortune to establish a brand, Globalstar has chosen to act as an
unbranded wholesaler.
Schwartz isn't worried. "In 2002, our marketing plan calls for just 80,000
subscribers in India," Schwartz says. "That's not a big hurdle. We expect
20,000 in Greece, 200,000 in Brazil, 200,000 in Russia. Our capacity will be
used up by low-hanging fruit." In all, Globalstar forecasts three million
subscribers by 2002, producing more than $2 billion in annual revenues. For
investors, though, that doesn't leave a lot of room for error, with the stock
effectively trading at close to four times the revenues that are expected four
years from now.
Starting in 2000, Iridium and Globalstar will face new competition from
London-based ICO Global Communications, a company originally part of
Inmarsat, the international maritime satellite consortium. Now independent,
ICO faces the obvious quandary of coming to market at least two years
behind its primary rivals. But it should be a formidable adversary, thanks to a
list of investors that includes all of the more than 50 members of Inmarsat,
many of them huge state-owned phone companies.
ICO will use medium-orbit satellites, built and launched under a $2.5 billion
contract with Hughes, which has an equity stake in ICO. While ICO will use a
traditional bent-pipe design, there's a twist. The company has set up its own
fiber-optic network, ICONet, to help keep down the cost of the earth-bound
portion of each call. In a sense, they'll do on the ground what Iridium does in
the air -- carry traffic off the normal public telephone network.
ICO figures its network will be easier to build and maintain. It won't be any
cheaper, though, costing $4.6 billion, on a par with Iridium.
Like Iridium, Globalstar and ICO will target both industrial markets and
business travelers. But Globalstar and ICO also plan to offer basic service to
people now not reached by cellular or land lines, a market Iridium has
decided to avoid because of cost. ICO Executive Vice President Fredrik
Verkroost figures there are three billion people who fall into the underserved
category, as many as 100 million of whom would sign up for telephone
service if they could get it. One way to address part of that market will be
with a type of remote, solar-powered phone booth.
The key problem faced by Ellipso and ECCO, the other primary satellite
phone projects, are their late start -- being the fourth or fifth player in the
market isn't ideal. Each has some unique features, though, which could help
them survive. Ellipso, created by privately held Mobile Communications
Holdings, plans a 17-satellite system, most operating in an unusual elliptical
orbit. Ellipso has heavyweight partners -- Boeing will be the prime satellite
contractor, and Lockheed Martin will make the ground stations.
Thanks to those elliptical orbits, Ellipso's satellites will spend two-thirds of
their time over the Northern Hemisphere, which Mobile Communications
considers more lucrative than the Southern Hemisphere. Ellipso thinks it can
keep costs to 50 cents a minute during peak times and 35 cents off-peak. In
fact, Ellipso say out loud what their rivals won't: that the company intends to
bring prices below those of terrestrial cellular service. Says spokesman Brian
Willard, "Our goal is to bring low-cost telephony to the world." The planned
debut of Ellipso's $1.4 billion project is 2001.
That same year should see the rollout of ECCO Equatorial, the first phase of
a project sponsored by privately held Constellation Communications. ECCO
Equatorial will use 11 satellites to provide service to a quarter of the Earth's
population. By the end of 2003, Constellation will unveil ECCO Global, with
the addition of another 35 satellites. Orbital Sciences will build satellites for
the first phase of the project, which is backed by Bell Atlantic and Raytheon,
among others. The smaller first phase of the project should cost less than $1
billion, the second phase about twice that.
There will be some additional competition from regional projects using one-
or two-satellite high-orbit systems. While suffering the usual geosynchronous
signal latency, those systems will be cheaper to build and operate than the big
low-orbit systems like Iridium and Globalstar. ACeS, a project led by Pasifik
Satelit of Indonesia, Philippine Long Distance Telephone and Jasmine of
Thailand, will target Southeast Asia, India, China and Australia. Etisalat, the
United Arab Emirates phone company, is building Thuraya, a project targeting
the Mideast, Turkey, Iran, India, Central Asia, North Africa and Eastern
Europe.
One thing the satellite phone companies won't do effectively is transmit data --
at a top speed of 2,400 baud, they wouldn't be much good for surfing the
Web. That's where Teledesic and its rivals in broadband satellite service
come in. These companies will provide fat data pipes for Internet access,
virtual private networks, video-conferencing, interactive gaming and myriad
other applications. The market won't lack for competition. The Teal Group's
Caceres counts at least 60 proposed broadband satellite systems, with a
combined price tag of $130 billion.
The most imposing project is Teledesic, which has received funding not only
from Bill Gates and Craig McCaw but also from Boeing and from another
billionaire, the Saudi Arabian investor Prince Alwaleed bin Talal. Teledesic
recently struck a crucial deal with Motorola which instantly eliminated its most
direct competitor. Motorola, which been planning a comparable service called
Celestri, instead agreed to combine forces with Teledesic, providing $1.5
billion in cash and technology for a 26% stake. This suggests that Motorola
executives think the project is already worth almost $6 billion.
Teledesic, born in 1990, will provide extremely rapid data transmission, 64
megabits per second downstream, to be exact, which is about 2,000 times
faster than a 28.8 modem. The uplink is slower, though at two megabits per
second, it's still faster than the T-1 line you may have in your office.
In short, it's the wireless equivalent of fiber optics. While Teledesic could
theoretically be used for some wireless-phone applications, what the system
will do best is provide broadband data to fixed points, using an antenna the
size of a laptop computer. The first satellite launch is planned for 2001, with
full service available in 2003.
Teledesic has the potential to provide the equivalent of "fiber to the curb,"
delivering nearly limitless data-transmission ability to homes and offices. Few
people have real fiber access today. For the most part, such fiber lines are
used for "trunking networks," capable of aggregating large volumes of traffic.
For individual users, says Teledesic CEO Russell Daggatt, the economics of
fiber don't work. "Even in this country," he says, "we're virtually starting from
scratch when it comes to broadband access."
Though the bulk of Teledesic's business will initially come from corporate
customers, to the average bandwidth-starved technology reporter, Teledesic
sounds like Nirvana. "We'll be able to provide anyone, anywhere, the
equivalent of optic fiber," says Daggatt. "No part of the world will be left
without that capability." While telephone companies are installing fiber at a
rapid rate, Daggatt doubts they'll ever catch up to demand. "The terrestrial
infrastructure will never be anything approaching ubiquitous," he says. "To
provide global connectivity, you have to look to satellites."
Technologically, Teledesic is hugely ambitious. Like Iridium, its 288 satellites
will be able to pass signals from one to another, avoiding the terrestrial
network. But Teledesic will have more than four times as many satellites.
Adding Motorola to the team, Teledesic gains the experience that was gained
engineering Iridium's satellite-to-satellite switching, while also getting
Motorola's expertise in making satellites. When required, Motorola cranked
out an Iridium satellite every four days.
Having absorbed Celestri, Teledesic's primary rival in LEO-based broadband
services is SkyBridge, a project sponsored by Alcatel and Loral. SkyBridge
plans a simpler approach. "Basically, SkyBridge is a wireless broadband local
loop system," says CEO Pascale Sourisse. The system will provide "last mile"
data access, linking consumers or businesses to the terrestrial fiber network
via a bent-pipe design. "We'll have gateways in each country in which we
operate," Sourisse explains. "We've tried to simplify the space segment, to
keep the intelligence on the ground. That allows us to re-use broadband
technologies used for other purposes. Switches in our gateways, for instance,
don't need to be designed specifically for SkyBridge."
SkyBridge is expected to be operational at the end of 2001, beating
Teledesic to market. That's despite the fact that SkyBridge recently increased
the size of its system to 80 satellites from 64, lifting the project budget by
$700 million to $4.2 billion. That's still less than half Teledesic's price tag, and
the modification should boost SkyBridge's capacity by 50%.
Surprisingly, Sourisse says SkyBridge plans to offer voice services in some
markets. "SkyBridge will primarily be used for things like high-speed access
to the Internet, delivering training, and Intranet connections," she says. "In
developed countries, that's where the demand will be. But we'll have capacity
available everywhere, including countries which may not need such services.
Instead, we can provide voice service at very attractive prices, 10 cents a
minute or lower."
That would pose a real threat to ICO, Globalstar and regional systems
planning to offer more expensive fixed-site voice services in developing
countries.
Several companies are developing broadband systems based on high-orbit
satellites. While unappealing for interactive applications, they would make
sense for corporate data networks, some Internet traffic and broadcast
applications. Among the players: Astrolink, from Lockheed; Cyberstar, from
Loral, and Spaceway, from Hughes.
Cyberstar will start service later this year, initially using capacity leased from
other Loral satellites. Spaceway, from Hughes, will target not just broadband
data, but also fixed telephony service in underserved markets. Hughes will
spend $3.5 billion on an eight-satellite network that includes celestial routing,
like Iridium and Teledesic. Lockheed's Astrolink will spend $3.2 billion for a
nine-satellite system, with launches starting in 2001.
No question, both voice and data satellite services have the potential to be
huge businesses. But for these ambitious, capital-intensive projects, the risks
are high. Most observers think both Iridium and Globalstar will find long-term
success, but at the moment neither stock looks especially cheap, and both
face weighty near-term risks.
SkyBridge CEO Sourisse notes that customers care about service and
pricing, not impressive technical feats. Access to technology gets you into the
game, but the real challenge will be turning the hardware into a real business.
And, take our word for it, not everybody will pull it off. |
