Agere Systems's (NYSE: AGR) exit from the optoelectronics business is being lamented by industry insiders and welcomed by Wall Street analysts, but neither group seems to take into account what the particular strengths of Agere have actually been... and how these no longer serve the industry that its predecessors helped spawn. Agere’s (and its predecessors’) strengths in optics have traditionally been in long haul and undersea components. As the industry seems to be shifting focus to metro and access over the next few years, and with Agere needing to show improved financial performance, the company's relative weakness in shorter-reach applications threatens its prospects for success in its main business, integrated circuits.

Agere Optoelectronics grew out of Lucent Technologies Inc.'s (NYSE: LU) optoelectronics division, which in turn grew out of AT&T Corp.'s (NYSE: T) optoelectronics section, each of which drew heavily on Bell Labs. The research results and products sold by these businesses have been among the finest in the optical communications industry and have been critical to the industry’s growth. There is no question that the laments being expressed by industry insiders stem from seeing an extremely prolific institution parting ways with an illustrious past.

At the same time, Wall Street analysts cheer the move. They see carrier capex, especially in optical networking, down for the next several years, with no visibility for a recovery. From their point of view, why should Agere continue to spend money to keep its optics capability in place, when doing so would not only drain cash but also keep the company from doing more on the client side? Agere is rich in optoelectronics technology, but what exactly is that worth when the market is down and the competition can catch up during the hiatus?

Both reactions are legitimate in their own way, but both miss the same point: Agere has traditionally been strong in optical components for long haul and undersea, but not for metro and access, and especially not for datacom. Simply look back less than ten years, before optical components was equated with DWDM, to see this point illustrated.

In the early 1990s, high-end performance was 2.5-Gbit/s Sonet/SDH. At that time, Agere’s optoelectronic business was AT&T Optoelectronics, and it was extremely strong in electro-absorption modulated lasers (EMLs) as well as PIN- and APD-based receivers. The business also had lithium niobate modulators, which at the time were seen as useful for external modulation of AM CATV lasers and for digital applications with distance-bandwidth products beyond those being deployed in commercial programs.

From my perspective as marketing director at Epitaxx (later bought by JDS Uniphase Corp. [Nasdaq: JDSU; Toronto: JDU]), I had the impression that other areas of AT&T Optoelectronics were not as strong and under competitive pressure. These included a variety of laser and detector modules, as well as transmitter and receiver modules, for lower-speed Sonet/SDH applications and fiber-in-the-loop trials. Moreover, AT&T Optoelectronics was competing in datacom, especially for IBM Escom business. When IBM cancelled its purchases one year, the revenue of AT&T Optoelectronics suffered greatly. Whether the cancellation was painful or a plus, however, is unclear because the rumors at the time were that AT&T Optoelectronics was losing money on every unit it shipped. AT&T laid off what was seen, at the time, as a tremendous number of people -- roughly one-third of its 3,000 employees. In addition, the company sold off its multimode datacom business. Moreover, AT&T decided to be far more selective about which singlemode telecom business opportunities it pursued.

AT&T Optoelectronics continued in at least two key ways, that I could then see, to pursue the lower-speed telecom business. On the expectation that convergence of voice, data, and video services to the home would occur, AT&T began development of its Manufacturing Realization Center, a factory designed to produce laser modules by passive alignment and automation. In addition, AT&T began developing singlemode ATM transceivers that did not rely on the company’s traditional high-performance platforms, but instead used package designs and processes similar to those employed by its competitors. Neither of these stratagems, however, was likely to succeed. Volumes for fiber-to-the-home modules proved to be miniscule, making the overhead burden for individual units built in MRC onerous for selling into an extremely cost-sensitive application. The SMF ATM transceivers consumed too much material that was manufactured outside of AT&T, whereas AT&T’s competitors, especially Hewlett-Packard Co. (NYSE: HPQ) (now Agilent Technologies Inc. [NYSE: A]) were more vertically integrated. AT&T’s need to make margin on top of its vendors’ margin left it in a poor competitive position. How AT&T Optoelectronics could continue in the optical communications market, if it stayed at the low end, was a legitimate question to ask.

It was fortunate for AT&T that WDM, and then DWDM, took off when it did. In the mid-1990s, WDM equipment manufacturers could not find many component vendors to provide the high-performance devices they needed. AT&T, and then Lucent, Optoelectronics saw demand from multiple DWDM equipment manufacturers for its EMLs and receivers at 2.5 Gbit/s and for its lithium niobate modulators at 2.5 Gbit/s and 10 Gbit/s. Given limited competition at the time, prices were high and profits robust. Then, as demand grew and supply could not keep pace, Lucent was actually able to move assembly of its EMLs and other DWDM transmission products onto the MRC production line, even reaching volumes that required increasing the capacity of MRC significantly. The automated factory that had been at risk of becoming a white elephant was now making money. At the time, people who had left Lucent Optoelectronics to join startups joked that the company had lucked into being in the right place at the right time, rather than having seen WDM coming and preparing for it in advance. For example, the impressions I drew when I was at Lasertron (bought by Corning Inc. [NYSE: GLW]) and then a market research firm, was that Lucent was not as dominant in WDM-specific products, such as high-power 980nm pump laser modules and arrayed waveguide grating wavelength division multiplexers. The company definitely recognized good fortune when it arrived and cultivated it, but it is unclear whether management actually anticipated the opportunity and planned for it.

With the slowdown of DWDM over the last six quarters, Agere had to decide what to do about its optical components business. Agere had expected to sell optics at higher levels of functionality, combining transponders and integrated circuits in order to sell them together as line cards. This decision was especially smart, because it differentiated Agere from JDSU and almost every other optical component business, all of which lacked the IC prowess of Agere. Unfortunately, the slowdown made this distinction moot and instead forced Agere to evaluate its optoelectronic business relative to other communications segments with stronger prospects. These segments are in metro, access, and datacom, each of which is an area in which Agere's precursors (Lucent and AT&T) had felt greater competitive pressure, or had exited. Moreover, since the mid-1990s, the competition in these areas has grown stronger and the number of competitors has grown. At that time, as mentioned above, Hewlett-Packard was dominant. Agilent today continues to be strong, but now there are other companies -- such as Finisar and the IBM division of JDSU, to name just two -- that have made significant progress in these segments, especially datacom. Agere management spoke aptly when it said that it would be at least as effective to buy transceivers from others as for Agere to make these products itself.

Of course, the decision to exit optoelectronics completely does prompt the question: What Agere would say if high-end optical components rebounded significantly in a few years? I have to assume that Agere has decided that even if the market were to come back, the size still would not be sufficient to justify a business division. Agere must be assuming that its IC business will grow large enough that a high-end optical component business would still be too small, and have too low a market valuation, to justify diverting resources to it.

The Agere decision also raises other questions about the industry at large. For instance, what does this departure mean for JDSU and others? JDSU wins, as it now has one fewer large competitors and an opportunity to increase its market share. For the smaller companies, the departure provides a chance to move up, either by buying some or all of Agere’s business or by the elimination of a competitor bigger than themselves.

How does the Agere announcement affect the sale of Nortel Networks Corp.'s (NYSE/Toronto: NT) optoelectronics business, and of any other optical component business that is being put on the market by a vertically integrated system house? The obvious answer is that the more businesses for sale in a depressed market, the lower the price any one of them can command.

And what about Intel Corp.'s (Nasdaq: INTC) plans? If optical components are not interesting enough for Agere, which is smaller than Intel, why should it remain interesting to Intel, unless the IC giant has a longer horizon or different goals? Intel professes a long-term commitment, but the slowdown is longer and deeper than anyone expected, as well.

In conclusion, Agere's departure from optoelectronics is simultaneously sad and sensible. In addition, it underscores that high-performance optical components are likely to remain a niche business, rather than a market akin to disk drives or integrated circuits, whereas low-cost optical components require having in-house optical components and integrated circuits, as well as high-volume packaging materials and processes. Agere does not want to be in a niche market and cannot see its opto business having a broad enough set of skills to compete against a broader field than existed several years ago. The Agere move lends support to the view that it is very difficult to be in both the high-performance and low-cost optical component business at the same time. Optical component companies are better off seeing where they fit best and being strongest in one, rather than mediocre in both.

— Jay Liebowitz, special to Light Reading
http://www.lightreading.com

Jay Liebowitz is founder and president of Liebowitz Strategies, which helps companies with strategy, business development, and positioning for increasing revenue and achieving corporate financing milestones. Liebowitz -- who can be reached at [email protected] -- was previously at RHK, where he was managing director of advisory services and founder of optical components research, and earlier managed businesses at Lasertron (now Corning) and Epitaxx (now JDS Uniphase).