Growing Pains Afflict Robust Optical Networking Market
New Fiber Optics Online columnist William Magill, principal with NationsBanc Montgomery Securities LLC, offers a Wall Street perspective on the optical networking industry. Reviewing 1998 activity and looking ahead to 1999 and beyond, he concludes that the optical networking market remains the most exciting investment opportunity in telecommunications.
By: Bill Magill, NationsBanc Montgomery Securities LLC
Contents
•Market pings and pangs
•New model year
•Peeking under the hood
It's early March 1998, and a light snow is falling on the Swiss hamlet of Davos. Inside the Meierhof Hotel, a small party is gathered at the cozy bar, mostly chatting about the day's skiing. A few of the patrons exchange thoughts on a new technology market: photonic networking. The gathering is an odd assemblage of futurists, scientists, investors, and analysts sharing drinks and dinner, with more than a few toasts to Uniphase for arranging this boondoggle: the opening of the company's new laser fab in Zurich.
The ribbon cutting provides a perfect backdrop for this impromptu summit. Despite the varied background among the guests, everyone shares a sense of good fortune. Carrier interest in photonics is exploding, suppliers are responding with fantastic product claims, wavelength division multiplexing (WDM) has become the hottest buzzword in telecom, and companies even remotely associated with the new technology are feeling an updraft in valuations. Possibilities within the industry appear as endless as the blue sky over the Alps.
But the skies over photonic networking markets have clouded. Rest assured that the optics revolution continues in full blaze and demand for products remains strong. But lest we forget economics 101, it takes both demand and supply to make a market. Recently, the supply side of this equation has been stinking up the joint.
top
Market pings and pangs
As the WDM industry matured in 1998, the market experienced its first growing pain. Although the optical networking market has shifted, it remains the most exciting investment opportunity in telecommunications. Investors will no doubt search out optical component suppliers, which hold substantial market power because they dictate technology advancements across the industry. Those advancements will yield incremental improvements in the performance of WDM systems in 1999, but cost cutting will escalate in priority relative to performance. Nevertheless, we see the optical networking market brimming with $3.3 billion worth of activity in 1999 (see Figure 1).
At the time of the soiree in Davos, only two companies—Ciena and Nortel—were shipping WDM systems capable of 40 Gbps or more on a truly commercial basis. Planners at the long-distance companies were preparing hara-kiri swords under the looming threat of network meltdowns, and bidding for the limited WDM supply among early adopters like Sprint and WorldCom was keen. To exacerbate the situation, the mother-of-all-carriers, AT&T, was evaluating the technology, placing even greater pressure on perceived demand. Flourishing in this early market period, Ciena was growing revenues at 130% and pocketing a quarter on each dollar of sales.
The carriers' incumbent systems suppliers would not let this stand (to quote a fearless George Bush). By the end of the year, at least six additional vendors entered the fray with dense WDM (DWDM) products. The surge in supply over-ran demand, and low-ball pricing, became a favored tactic for securing a foothold in the market. Prices tumbled precipitously, cutting Ciena's gross margins in half, slipping sales growth into reverse, and turning profits negative. In dollar terms, the WDM market finished 1998 25% below our forecast set the previous year. What happened? The WDM market began to mature, and experienced its first growing pain.
top
New model year
In 1999, the deployment of WDM systems will remain primarily in long-distance networks, be they on land or undersea. The market's most generous advocate—Sprint—may cut back spending in 1999, after more than two years of aggressive WDM deployments. But accelerated spending at many other carriers will more than fill the void. WorldCom, for example, will return from the 12-month WDM spending moratorium that it launched to digest its MCI acquisition. Carriers competing for both long distance and local services will step up spending on optical systems as they allocate new network capital to the latest, most capable technologies. On top of that, we count 11 new undersea builds—all deploying WDM.
Point-to-point WDM transport will remain the primary optical networking application in 1999. Incremental improvements in performance, including more channels (16 to 40 to 128) and more speed (OC-48 to OC-192), will comprise enhancements to this year's model of WDM—as opposed to sharp changes in utility. While stripped-down WDM econo models for metro networks will hit show rooms in 1999, full-size luxury systems will continue to capture most of the spending. While price pressure has knocked some of the luster out of market growth, we still expect growth to $3.3B this year, and 37% compound annual growth through 2001.
top
Peeking under the hood
The real fun in 1999 surrounds the optical component suppliers. Because they largely determine the optical networking industry's technology progression, these guys wield substantial market power. While the 1999 focus will continue to be on bigger (more power, more channels), better (narrower, more stable wavelengths) and faster (high line rates), cheaper is quickly becoming top priority to remain in good standing with the systems industry.
For a crystal ball view on the systems market in 2000 and beyond, take a look at development activities in component labs now. Dynamic systems are required to make WDM a real networking technology instead of just a big fat pipe. Tunability and addressability are key—in wavelength, dispersion, and attenuation—to which considerable R&D capital is being applied. Component integration and packaging also commands attention, because both lead to more efficient assembly and lower costs. Systems suppliers don't want to align and piece together 1000 discrete components with expensive and time-consuming manual labor. Vendors want to pick and place pre-assembled modules onto fiberglass cards, test them, then ship.
Are we betting that a plug-and-chug supply chain will support dynamic WDM any time soon? No, but it will happen. Development activities in these areas are brisk, drawing significant capital investment. And capital markets are amazingly efficient at getting worms to the squeakiest chicks.
Driving all of this activity is the demand for bandwidth, which shows no signs of abating. To skeptics who envision a saturation of bandwidth—who believe this is all too much, too soon—consider that Internet users are doubling in number every 100 days. SBC is rolling out ADSL to a quarter of its service territory in 1999. AT&T will leverage its considerable capital muscle to accelerate consumer cable modem service over the TCI network. And Lucent has given Winstar a $2 billion charter to bring wireless broadband to the masses.
Soon, I'm off to Uniphase's second annual ski weekend in Davos. Dinner conversation will surely be animated, drunken with anticipation, and the skiing will be world-class. Taking part will be the bandwidth brokers—the laser makers, filter designers, and systems assemblers. Although the optical networking market has shifted since the last trip, these players are in on the most exciting game in telecommunications.
top
About the author
Bill Magill is a principal with NationsBanc Montgomery Securities LLC. He can be reached at wmagill@montgomery.com.