So, which analyst firm do we believe?
The one that says optical long-haul will rebound next year? Or the one that says optical will continue to decline for the next two years?
Two weeks ago, I wrote in a column how CIR, a market researcher in New Jersey, predicted recovery for optical long-haul in 2003 based on conversations it had with about 20 service providers. The US$860 million U.S. market for optical long-haul in 2002 is growing to US$1.1 billion next year, and then to US$3 billion in 2006, according to CIR.
Then last week, Dell’Oro Group Inc. issued a release indicating that optical long-haul will be one of the components dragging down the worldwide optical transport industry for the next two years. Dell’Oro predicts 2004 worldwide sales of US$8.1 billion in metro DWDM, long-haul DWDM, SONET/SDH add/drop multiplexers, SONET/SDH Multiservice, and optical switching, down 65 per cent from the US$23 billion peak in 2000.
Modest growth in optical transports won’t occur until 2005, according to Dell’Oro.
Granted, CIR is charting only the U.S. market while Dell’Oro is considering worldwide sales. But which is accurate?
We asked the firms themselves.
“Did you ask Dell’Oro whether they talked to service providers?” asks Mark Lutkowitz, vice-president of optical networking research at CIR. “They’re just an extension of the PR departments of manufacturers. They send out surveys, and if you don’t respond you end up with zero per cent market share, like ADVA (AG) did [one year].”
Lutkowitz maintains that optical is growing this year because bandwidth is still growing. AT&T Corp. in mid-2003 will need to fill up capacity on some of its long-haul routes, he says. Sprint, on the other hand, pushed out its forecast for additional bandwidth into 2004.
Lutkowitz admits market forecasting is an imperfect science, but he also suggests that the optical long-haul market is stigmatized. “Long-haul is a dirty word,” he says. “Overall optimism [of a few years ago] is now excessive pessimism in long-haul.”
Lutkowitz also says his firm, CIR, is not shy about telling it like it is, while others fear offending their clients. “We’re not cheerleaders, we’re not pie-in-the-sky,” he says. “We’re equal opportunity offenders.”
Dell’Oro, meanwhile, says CIR’s data is overly optimistic. Based on its quarterly vendor surveys and, yes, discussions with service providers, Dell’Oro actually projects a more robust market for long-haul optical this year in North America – somewhere north of US$1.2 billion.
But Dell’Oro expects that to drop off by 20 per cent in 2003. In 2006, Dell’Oro projects the worldwide optical long-haul market to be US$3.5 billion, vs. CIR’s forecast of US$3 billion in the U.S. alone.
“If someone is out there telling us that North America is going to rebound in a market that everyone is screaming overcapacity, then I think there’s someone either very optimistic or they believe that there is no overcapacity in the network,” says Jimmy Yu, Dell’Oro’s optical transport analyst. “It’s way too optimistic.”
Yu says there’s only about three or four routes in a carrier’s North American long-haul networks that are “hotspots,” or in need of additional bandwidth capacity – hardly enough to warrant the kind of growth in optical transport and switching gear CIR is projecting.
“Unless they’re saying that more carriers are going to come into this space and build out networks, I can’t see how the U.S. market could grow that big,” Yu says. “Eight hundred million to US$3 billion over the course of four years? That’s a lot.”
Yu does not argue that carriers will consume more bandwidth next year and in coming years, but the price erosion that occurs with optical long-haul equipment will offset any remarkable growth in revenue, he says.
In fact, Yu says, revenue will decline.
“They could buy more units with less price,” he says. “It’s pure economics at that point.”
Duffy is a managing editor with Network World (U.S.). He can be reached at email@example.com.