SUNDAY JULY 11, 1999
Jumping Off the Bandwidth Wagon – Long Distance Carriers Regroup
by Seth Schiesel
In a dingy warren of cables and wires on Manhattan’s Upper East Side, phone calls are starting to
look a lot like pork bellies.
There, in his cramped office above the Pulse Night Club, Alex Mashinsky, 33, has created an
electronic trading floor for telephone service that resembles nothing so much as the international
market for traditional commodities.
Communications carriers from around the world, including giants like AT&T and Nippon
Telegraph and Telephone, link to the Web Site for Mr. Mashinsky’s company, Arbinet
Communications, to trade unused long-distance telephone minutes from New York to every
corner of the globe. On Friday, a minute to Israel sold for as little as 8.4 cents, while a minute
to Hong Kong cost as little as 4.6 cents.
Although it is tempting to hold up Arbinet as a creator of a new communications landscape, in fact
it is only a stark symptom of a much larger shift in the electronic world: While consumers and
regulators focus on the communications bottleneck in the so-called last mile of wire to homes and
businesses, long-distance communications capacity — or "bandwidth", one of the hot buzzwords
of the Information Moment — is fast becoming a commodity.
Deregulation, rapid advances in optical technology and generous capital markets have combined
to create an explosion in communications supply — a supply that exceeds even the surging
demands in communications unleashed by the Internet. Accordingly, long-distance prices are
falling sharply. On Friday, AT&T sold its biggest business customers a minute of
telephone time to Israel for 35.4 cents, less than a third of the cost three years ago,
and a minute to Hong Kong for 27.2 cents, down 57 percent. The big gap between the
retail marketplace and spot markets like Arbinet’s cannot last for long.
More significantly, for the companies and investors alike, margins in some parts
of the communications business are following suit, as long-distance bandwidth for
both telephone and data services turns from scarce to plentiful.
The shift is driving many of the strategy reversals and multibillion-dollar deals
that have become almost commonplace in the industry. It explains AT&T’s burst into
the cable television business and the continuing, soap-opera-like bidding war between
Qwest Communications International and Global Crossing for the Frontier Corp. and U S
At the same time, the fierce competition in the long-distance arena only highlights
its relative absence in local markets. That, in turn, reinforces the almost overarching
value of local communications networks — and is prompting some investors to rethink
their portfolios for the digital future.
A New Profit Formula
The changes in the long-distance industry are undermining what is perhaps the fundamental
rule of telecommunications economics — that margins stick to assets.
After AT&T began to face significant competition about two decades ago, it
became clear that long-term profits in the business stemmed largely from
controlling physical assets like fiber optic lines, copper wires and
communications switches. For years, the cost of running a network was a big
part, perhaps a third or more, of running a communications business. A major
component of making money was keeping the continuing network costs low, and
the main way to do so was to own the network.
That is one reason Sprint emerged from a venture of the Southern Pacific
Railroad, which could use its railroad rights-of-way to lay fiber-optic lines.
It also explains MCI’s original name, Microwave Communications Inc.; by using
microwave links it had developed, MCI could provide long-distance services for
less than AT&T. (MCI eventually built its own fiber optic network, to help the
company compete more broadly.)
Most of the hundreds of other new long-distance companies that emerged after the
1984 breakup of AT&T, however, leased communications capacity from others, and
few have had consistent profits and growth.
Drawing from that experience, new companies like Qwest, Level 3 Communications, Global
Crossing, IXC Communications and the communications units of the Williams Companies and
Enron, two energy giants, have been founded — and financed — on the logic that owning
physical assets was essential. Each of the companies is building a new fiber optic
long-distance network, some in the United States and some globally. Combined, the first
three have raised or borrowed more than $10 billion over the last three years, mainly to
build their new networks. At the same time, new optical technologies have been expanding
the capacity of existing networks by factors of 16 and 32, with a potential of 160.
To grasp just how much bandwidth is being created, consider this: According to Telegeography,
Inc., a research firm in Washington, in 1997 the world’s public telephone networks carried 81.8
billion minutes, or 155,632 years, worth of international phone calls. The newest optical
systems can transmit the digital equivalent of all of those calls over a single fiber
in about 11 days.
Hence the new carriers’ predicament: Their investments are alleviating the
very scarcity that made their original business plans so attractive.
The problem is not that a glut of bandwidth has been created. It is clear that
new, bandwidth-hungry services will fill all of the available communications
pipes. However, the big profits in the industry are migrating away from raw
bandwidth and toward the applications that use it.
An early sign of how prices would turn volatile as bandwidth supply played
leapfrog with demand was in the market for T-3 Internet lines, which can carry
45 million bits of digital information a second.
When they first became widely available in 1995, the lines sometimes cost less
than $20,000 a month; buyers included corporate customers and Internet service
providers. But as the use of the Internet exploded, demand outpaced supply, and
prices shot as high as $70,000 a month in early 1996.
Since then, however, prices have fallen consistently, by about 10 percent a year,
as new and existing carriers have built up their networks. T-3 lines now cost about
$50,000 a month, and many in the industry expect the trend to continue or even
"People have been thinking about bandwidth from a Malthusian standpoint:
‘How do I ration this?’" said Daniel Smith, chief executive of Sycamore Networks,
a communications equipment company, whose previous business, Cascade Communications,
another equipment maker, was bought by Ascend Communications in 1997 for $3.7 billion.
"And now we’re saying, ‘How do we think about this from a standpoint of plenty?"’
Many of the nation’s communications carriers have come up with an answer:
They need to evolve up the digital food chain — from mere bandwidth
suppliers to providers of advanced services — to maintain their profit
margins in the future. They are scrambling to acquire direct links to homes
and offices — not for the assets per se, but because in the next century,
margins in the communications business will stick to services and customers.
"If anything, the cost of communications is falling faster than the cost
of computing, so assume unlimited bandwidth,” said Howard Anderson, managing
director of the Yankee Group, a technology consulting firm in Boston. In that
case, he said, “you are either the application service innovator, or you’re dead
Moving to Plan B
The shift has shred business plans across the industry and helped fuel the
paroxysm of takeovers reshaping the communications landscape.
AT&T provides a classic example. Many strategic subcurrents have been at work
behind AT&T’s $90 billion leap into the cable TV business, from a need to
protect its long-distance base to the simple desire to shake the company out of
a torpor. But at the root has been a quest for sustainable profits, and the brain
trust at AT&T’s headquarters in Basking Ridge, N.J., decided that such profits were
not to be found merely in the long-distance business.
Having a long-distance network is "a necessary, but not sufficient, condition
to succeed in the future," said Frank Ianna, president of AT&T’s network unit.
Instead, AT&T concluded that it again needed to connect directly to its
customers. "That’s where there’s the least amount of competition," he
explained, referring to local markets, " and that’s where you have the
ability to differentiate from your competition."
AT&T is betting that owning cable lines directly into customers’ homes will allow
it to deliver interactive television, high-speed Internet services and local phone
service — all products that should support higher margins than its core long-distance
MCI Worldcom, the No. 2 long-distance provider, after AT&T, has almost the
same philosophy, but with a focus on serving business customers. It, too, sees
its future in services and applications, not simply bandwidth. And it also wants
to own the direct link to the customer.
"Our story is very different from offering raw capacity," said John
W. Sidgmore, vice chairman of MCI Worldcom.
In Worldcom’s evolution from communications nobody to industry titan over the
last decade, the company’s deal to acquire MFS Communications in 1996 looks at
least as important strategically as its acquisition of MCI last year. The MFS
deal gave Worldcom not only direct, local fiber optic links to business
customers, but also the Uunet Internet operation run by Sidgmore. Uunet, coupled
with MCI’s corporate data communications operation, is now driving MCI Worldcom’s
growth, by developing the advanced data applications, like intranets, demanded
by big business customers and delivering them worldwide.
Since MCI and Worldcom announced their merger in the fall of 1997, telephone
service revenue has dropped from 68 to 61 percent of the combined company’s
overall sales, while Internet and corporate data services have grown from 22
percent to 30 percent. In the first quarter of 1999, telephone service revenue
increased only 7 percent over the comparable 1998 period, while data services
sales grew 30 percent and Internet revenue 60 percent.
At Qwest, the shift in focus toward the top of the food chain has happened
in fast-forward, and some investors have found it unsettling.
When the company first offered shares to the public in 1997, it was seen on
Wall Street as a bandwidth play. Investors expected Qwest to concentrate on
using the latest technology to build a long-distance network with more, and
cheaper, capacity than those of companies like AT&T or MCI, letting the
upstart take market share away from slower, less-advanced competitors.
Over the last year, though, Qwest has repositioned itself as — surprise! —
a provider of advanced communications services for end users. This year alone,
Qwest has announced partnerships with companies including KPMG, BellSouth,
Siebel Systems, SAP America, Hewlett-Packard, Oracle and Automatic Data
Processing’s Brokerage Services unit to develop high-margin applications
like software for sales forces, aimed at various market segments//cut for space.
And now, Qwest is making an unsolicited run at acquiring Frontier, a
long-distance company that also owns Globalcenter, one of the biggest
Web-hosting concerns, and US West, the regional Bell company based in Denver.
U S West is the slowest-growing Bell, but it still has millions of customers,
whom Qwest needs for the new services it is developing.
Qwest is trying to wrench Frontier and U S West from the grasp of another
erstwhile bandwidth play, Global Crossing. While both Qwest and Global
Crossing have made cases for why their deals make sense, their sudden
transformations have left some investors scratching their heads.
"It’s created confusion," said Eric Strumingher, an analyst at
Paine Webber. "Qwest and Global Crossing have represented themselves as
growth companies, as nimble companies based on modern technology and that
had the most flexibility. Now they’re attempting to combine with a company
that was not representing itself like that, and is in fact not like that.
That is the big source of the indigestion."
To James Q. Crowe, perhaps the most prominent executive still carrying the
flag for the pure bandwidth strategy, this is all tantamount to heresy.
Crowe is chief executive of Level 3, which is building a national fiber optic
communications network based on Internet systems and the latest optical
technology. Crowe says he has no intention of selling advanced services to
consumers. He says that through shrewd investments, he can sustain a technical
dvantage over his competitors. That, in turn, would let Level 3 sustain a cost
advantage, enabling the company to persist as a wholesaler of long-distance
"I think I can drop prices faster than anyone else can keep up with and have
margins that are sustainable," Crowe said. "If you can get a six- or nine-month
technical advantage over your competition, you can have a long, long, long life."
The stock market has hardly abandoned Crowe’s world view — shares of Level 3
are up 56 percent this year — but others in the industry say he is ignoring
the commoditization of bandwidth at considerable peril.
"I respect Jim and I’ve worked with Jim, but I think he’s absolutely, completely
dead wrong," Sidgmore said. "It really is black and white. The fact of the matter
is that if you just stand back and think about it, Jim is not going to invent any
new technology. It’s coming from Lucent or Cisco. And do you think they’re going
to give anyone an exclusive? If you wanted to give someone an exclusive, it would
be the company that could buy the most product; it would be MCI Worldcom or AT&T."
But Level 3 is different from those companies — if not from Qwest or Global
Crossing, in their original guises — because it has no existing customer or
revenue base to protect. That, Crowe says, will keep his company more focused.
"History tells us that when you get bigger and bigger and more and more vertically
integrated, you break businesses apart," Crowe said. "That’s like starting a computer
company in 1980 and saying the model is IBM." Integration is also the exact opposite
of government policy and market activity in, for example, the power business, which
is splitting into distinct segments that generate electricity, transport it over long
distances and then deliver it to users.
Indeed, some communications experts think that consumers will suffer as carriers
seek vertical integration — offering advanced services as well as raw communications
pipelines — to fight off the margin-eroding tides of commodity pricing.
"As someone on the buying end of it, rather than the selling end of it, I’d like
the companies that provide bandwidth to do a really good job of that rather than do
what the local telcos have done, which is bring services into their networks," said
Robert M. Metcalfe, one of the founders of the 3Com Corp., the big data networking
Metcalfe, now an industry consultant, contends that the Bell companies, which
dominate local communications, have done a poor job of making advanced services
available to consumers, even as they try to fend off competition.
For investors, it may not really matter who is right. Even some people who own or
recommend the stocks of companies like Level 3 acknowledge that the pure bandwidth
strategy may not endure for the long term — but add that many such companies will
probably be acquired in the next few years anyway, perhaps as foreign carriers
expand into the United States.
Still, some professional investors are shifting their focus to the remaining points
of scarcity in the communications business — local assets and expertise in advanced
services, especially on the Internet.
That is why Michael J. Mahoney, director of telecommunications investment at
Dresdner RCM Global Investors, likes MCI Worldcom but is also becoming more bullish
about local communications companies.
"We have a decent position on the local side," he said, "and if anything,
we’re steadily growing in our affection for the local side." Mahoney
particularly likes GTE, which is trying to merge with Bell Atlantic, and
SBC, which appears close to completing its deal to buy Ameritech.
Mashinsky of Arbinet, the communications exchange in New York City, watches
the new world of long-distance evolve every day on his computer screen. To
him, the shift is clear. The future is not in electronic pork bellies; it is
in the fiber optic equivalent of processed meats.
"Over the last three years, $402 billion was invested in telecom worldwide,
and a lot of that went into infrastructure," he said. But many of the
entrepreneurs "have not been able to fill up their networks, because there
are too many people building networks. "So a lot of these guys are refocusing,"
Mashinsky added, "and saying that if we want to retain our margins, we have to get
a hold of some customers and add some value."