There are indeed some lions in winter

Optical switching is likely to be one of the more open-ended opportunities in the optical system market over the next couple of years.

By KEVIN SLOCUM AND ROBERT MANDRA

As we prepared to write this month's column, we made certain to go back and read what we had prepared for the prior month. We would have liked to retrieve what we had written and done a rewrite. The reason is we knew as we wrote this article that the Harmonic situation we had referenced in the November issue (page 32) had already taken a turn for the worst and put an exclamation point on our comments before you even read them. Frankly, we probably would have ducked mentioning our buy rating on the shares at the time, because it certainly looks foolish now. We've had a few great stocks over the past few years on our buy list and we have now had a very bad one. Goes with the territory.

Because you hear from us with less regularity than our everyday customer, we try to remain focused on the longer term. What happened to Harm onic's shares speaks volumes about the perils of investments in high- valua tion stocks. The volatility in the group over the course of 2000 only further accentuates the risk. One important financial disappointment in 2000 probably will be followed by at least a few in 2001. After all, we have several newly public companies, many that got to that point without the normal basic training period that is typical for IPO companies. A price will certainly be paid.

All of which goes back to some of our points in last month's column in which we mentioned the need to be more careful with our dollars and our careers. We hope the fall correction has run its course and given way to the rally we expected moving into the end of 2000. The correction was already deeper than we had expected, due to broad-based concerns about carrier spending plans coming under pressure in 2001. We recognize the challenges faced by the carriers, but we also believe that the reality of intelligent optical networking is critical to their getting ahead of the declining price curve per bit of information transported.

Since we are talking about being more careful, let's talk a little about what we expect to be the better areas in the year ahead. We believe that optical switching is likely to be one of the more open-ended opportunities in the optical system market over the next couple of years. That is why we have pressed the Ciena bet most of 2000, and it is a bet we believe we will be pressing well into next year. Look at what has happened to Ciena's valuation throughout the year; it has continued to swell when others have not.

We mentioned in last January's column (page 30) that Sycamore could knock the cover off the ball and get no stock credit because the shares had already priced in unbridled success. That has in effect been what has happened. With good execution on its switch platform, the SN16000, we could see next year possibly being a more rewarding year for Sycamore Networks investors. We don't mean to hedge by mentioning execution, but the company does have a lot on its plate and good execution is critical to success.

Nortel Networks on the other hand is behind in this segment and racing to get an entry to market next year. We believe that the lack of an optical-switch offer ing, which is part of the current carrier decision mix, has probably cost the company al ready from a valuation perspective. An ultra-long-haul platform and accelerating prospects for its metro platform should provide incremental business momentum, but we are not sure they will carry the same valuation sizzle.

In components, we expect the next 12 months to also be a bit more challenging. This year, investors have really begun to be exposed to some newer and innovative component solutions. If you listen to what component-company management says, you start to hear what they really think is important. Look at Raman amplification. Last year, there was a point where only SDL was talking about the promise of this technology. By the Optical Fiber Communications conference last March, all the major optical amplifier companies were talking about their Raman amplifiers.

Today, we are hearing everyone talk about integration of discrete components into planar solutions. Some will mention their enthusiasm for silica on silicon; others will speak of silicon bench technologies. But they all seem to hear from customers the need for integration, and the sooner the better. This could mean good things for Bookham and the silica-based arrayed-waveguide-grating companies that are all beginning to talk about the promise of functionality beyond just multiplexing/demultiplexing. Market analyst RHK reports at least a four-fold increase in the multiplexing/demultiplexing class of devices for 2000. With the added optical channel monitoring and variable optical atten uation features under development, we have to believe that this type of device is going to produce continued phenomenal growth as this integrated functionality moves from design-in to reality.

Elsewhere, we're hearing that the only way to really automate optical com ponent businesses is to deal with fiber-handling. That means two things: Eliminate the fiber through integration or incorporate the functionality into the fiber. Corning has begun to talk about fiber as a component. There is a side of us that looks at that premise and says it is self-serving because they would certainly get a higher stock price if investors valued its fiber business like components. Yet, there is also a side of us that believes it is an honest assessment of optical fiber today. We've all heard of what is going on with fiber Bragg gratings and fused-fiber devices. Add what Corning and Lucent Technologies are doing with designer fibers, and the fiber category is still very much appealing to us despite the ever-present concerns of over-capacity.

Emerging integrated solutions and fiber components will be too small a factor to change business conditions next year, but they will probably have an important impact in 2002. The race to drive costs of production down in some of the mainstream passive solutions of today will be intense, as suppliers like JDS Uniphase work to leverage the investments they made in OCLI and E-TEK Dynamics, while at the same time trying to scale the new technologies. That mix will probably translate into moderation in JDS's valuation as it moves toward 2002. Despite that, we believe the company will have growth superior to many technology companies that will continue to justify their ownership by growth investors. We will therefore probably be a little more sensitive about the prices we suggest paying for new JDS positions.

Optical opportunities in the metro/ access market also feel as though they may be rewarding in the next 12 months. Is every metropolitan DWDM play going to be successful? No, but there does appear to be better momentum in this market, and certainly companies are finding a way to grow by supporting the increased

optical requirement near the edge of the network. Finisar has been just such a play for us since we added it to our focus list back in the April time frame, and since then we have become even more comfortable with the company. We believe there are other opportunities that will emerge in this portion of the network over the course of the next year.

Robert Mandra is a principal in investment banking with Wit SoundView (Stamford, CT). Previously, he was an optical engineer with MIT Lincoln Laboratory for nine years. He can be reached at (203) 462-7361 or [email protected]