Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : JDS Uniphase (JDSU) -- Ignore unavailable to you. Want to Upgrade?

To: Kent Rattey who wrote (1856)	11/8/1999 7:36:00 PM
From: Kent Rattey	Read Replies (1) \| Respond to of 24042

Off MF: JDS UNIPHASE, INC. ("JDSU") I. JDSU Snapshot II. Fiber Optics and the Broadband Revolution A. The Increasing Need for Bandwidth B. Use of Fiber Optics for Broadband C. Dense Wavelength Division Multiplexing (DWDM) - The Key to Unlimited(?) Bandwidth Over Fiber III. Products A. Lasers B. Modulators C. Fiber Gratings D. Receivers E. Wavelocker F. Other Assorted Items G. Optical Modules H. The Holy Grail - Optical Switches IV. Markets, Customers & Competitors A. Overview of Markets B. Discussion of the Major Fiber Optics Markets C. Customers D. Competitors V. Barriers to Entry (or, Why JDSU Will Win) VI. Projections for Increasing Demand VII. Growth Through Acquisitions VIII. Management IX. Financial Performance I. JDSU SNAPSHOT * JSD Uniphase Corporation ("JDSU") is a leading optoelectronics and photonics company that designs, develops, manufactures and markets components and subsystems used in fiber optic telecommunications. * Although its products are also used in cable television, biotechnology, graphics, printing and other industries, the primary focus of JDSU - and this report - is the fiber optic communications industry. * JDSU is traded on the NADSAQ market under the symbol "JDSU" * JDSU is a member of the NASDAQ 100. II. FIBER OPTICS AND THE BROADBAND REVOLUTION A. The Increasing Need for Bandwidth JDSU's business is built upon the exponentially increasing need for "bandwidth" and "broadband" communications. There are lots of technical definitions of these terms, but suffice it to say that "bandwidth" is the amount of information that can be sent in a particular period of time (e.g., megabits per second). You can increase bandwidth by either increasing the capacity of the wires or increasing the speed over the wires. To analogize to plumbing, you can either make your pipes bigger or you can push the water through the pipe faster. "Broadband" is generically used to refer to high speed, high capacity communications, often encompassing many types of information (data, voice, video, etc.). There is a broadband revolution under way. This is due to a number of factors, including the rise of the internet, the ever increasing speed of microprocessors, the continued penetration of computers in the home and in all aspects of business, and the increasing convergence of the computer, phone, and television. Reports state that traffic on the internet is doubling every 100 days. Overall network traffic currently consists of roughly 50% data and 50% voice, but is projected to be 90% data and 10% voice in the relatively near future. (This despite the fact that voice traffic is still increasing 10% each year.) The bottleneck for most people working on networks today is no longer processor speed, but rather networking speed. In other words, your computer isn't slowing you down, your slow network connection is. This is expected to continue, because the demand for bandwidth is believed to be elastic. As computers communicate faster and faster, we will want to send more and more information, such as full screen video, uncompressed music in real time, etc. An investment in JDSU requires a belief that the need for bandwidth will keep increasing for the foreseeable future - in my mind, a very sensible belief. B. Use of Fiber Optics for Broadband Although a variety of technologies have been considered for broadband use, the one with the most proven effectiveness is fiber optics. Most information devices we use today convey information through electricity, which simply put is a flow of electrons. Examples include your television, your phone, and your computer. Information in these "electronic" devices travels over copper wires, and is processed by silicon chips. In contrast, fiber optics conveys information by sending pulses of light over strands of glass. (See Section ___ below for a more detailed discussion of the technical details.) Because light consists of a flow of photons, the devices that handle the light are often called "photonic" devices. Because the information sent via fiber optics travels at the speed of light, it is generally acknowledged that no other method of transmission can be faster. Fiber optics is currently far more expensive than traditional electronics. As a result, fiber optics are used primarily as the "backbone" of communications networks, with the "edges" of the network still transmitting information via electronics. You can analogize to the highway system - you travel long distances using the interstates, but ultimately have to exit and take some local roads to get to your destination. As demand for bandwidth continues to increase, fiber is spreading steadily from the backbone out toward the edges of the network; in other words, more and more roads are being converted to highways. The aim of "optical networking" is to have as much of the transmission handled via optical (as opposed to electronic) means, with the ultimate goal being fiber running right into everyone's home (sometimes called "the last mile"). Until recently, a major flaw with the installation of cross-country fiber optic networks was the problem of amplification. Because light loses intensity as it travels through the glass fiber, it was necessary insert additional equipment every so many miles that would read the (now weakened) light signal, convert it to an electrical signal, and then use that electrical signal to send out a fresh fully-powered light signal. This was not only expensive, but also far too slow. It made no sense to send signals at the speed of light if they were going to have to slow to "electron" speed every 100 miles - it was the metaphorical equivalent of putting stop signs on the interstates. This problem was solved, however, with the creation of the Erbium Doped Fiber Amplifier (EDFA). In english, an EDFA is a fancy name for an invention that amplifies the light signal in a fiber without having to convert it to an electrical signal. It is composed of a special chemically altered piece of fiber and an attached laser which adds the light energy needed to "pump" the signal further down the fiber. (The laser used in the EDFA is called a "pump laser.") The EDFA was an extremely key development in fiber optics, and is one of JDSU's key products. Ever since the advent of the EDFA, telecommunications companies have used fiber optics as their network backbones. This, however, did not solve their bandwidth problems for long. With the explosive increase in demand for bandwidth, they have been continually running out of room on their networks. This presents a major problem, because the obvious solution - just lay more fiber in the ground - is extraordinarily expensive (approximately $75,000-80,000 per mile!). The FAQ for JDSU eloquently describes the dilemma: "Increased data traffic put telecommunications companies in a bind.... In the early fiber optic networks, each strand was only capable of carrying one channel of information. With the emergence of high volume data traffic such as the Internet, the telecommunications companies found that their systems were too small. Another way to think about the situation is this. The telecoms built multi-billion dollar highways with no speed limits that everyone wanted to use. The problem was the highway only had two lanes." C. Dense Wavelength Division Multiplexing (DWDM) - The Key to Unlimited(?) Bandwidth Over Fiber The limitation on bandwidth in earlier fiber optics systems was in part due to the fact that these systems only used one color of light (using the highway metaphor again, these were one lane roads). Engineers have discovered, however, that they can multiply the data capacity of the fibers already in place many times by sending more colors of light through the fiber at the same time. This technique, called Wavelength Division Multiplexing (WDM), turns the "one-lane road" fibers into multi-lane super-highways. WDM operates by separating the color spectrum (picture a rainbow or prism) into a number of "channels", with a different color for each channel, and with different information going into each channel. All the channels (colors) are then crammed together to form one light which is sent through the fiber. When it gets to the other end, the opposite happens -- the light is split into all the different colors, and the information from each gets sent on its way. This works because light is additive - e.g., you can add a red light to a blue light without losing any part of either light; instead you get a purple light, which can be split later back into blue and red lights. Quite simply, the more colors of light the WDM system can handle, the more bandwidth it will give you. In theory, because the light spectrum can be divided into an infinite number of colors, you could get infinite bandwidth using WDM. In practice, it requires more and more precise equipment to distinguish between finer and finer gradations of color, so we currently are nowhere near the infinity mark. The best system on the market now handles 40 colors; companies have announced 80 color systems, but those apparently are not yet available. Systems that handle more than 8 colors are called Dense Wave Division Multiplexing (DWDM), and systems that use more than 40 colors are referred to as ultradense or hyperdense Technical details aside, DWDM is important because it allows telcos to drastically increase the bandwidth on their existing fibers without digging up the ground. All that is required is that DWDM systems are added to the "ends" of the fiber. Consider how much it would cost to lay only one new fiber across the United States (3000 miles x $75K per mile = $225 MILLION), and you can see why demand for DWDM systems has absolutely exploded. DWDM is the core technology for increasing bandwidth in fiber optic systems, and is the foundation for JDSU's growth. III. PRODUCTS JDSU does NOT sell complete DWDM systems. Rather, it sells the components -the basic building blocks - that are used to make DWDM systems. These components are bought by companies such as Lucent, Ciena, Nortel and Alcatel, who use them to assemble DWDM systems, which are sold to telcos such as AT&T, Sprint, Qwest, Level 3, and MCI Worldcom. (See the discussion below re: JDSU's Customers.). Prior to their recent merger, Uniphase was the market leader in so-called "active" components - e.g., lasers, modulators, and amplifiers. JDS Fitel was the market leader in "passive" components (i.e., no electrical input needed), such as couplers and splitters. Together, their product offerings constitute a catalog of virtually all of the components necessary for DWDM systems. These are described below: A. Lasers Lasers are used in at least two ways in fiber optics systems. "Source lasers" create the light that is sent along the fiber. A 40-channel DWDM system will need 40 source lasers, one for each color. "Pump lasers" are used to make the EDFAs work. The most commonly used lasers are semiconductor lasers, which are tiny devices fabricated in much the same way that computer chips are made. They are made in "batch" form, in a reactor that very precisely flows gases over a wafer of semiconductor material (usually silicon or gallium arsinide). The wafers are then processed in various ways to hopefully produce a large number of lasers. The yields from this process can vary widely, however, and I've been told that creating these lasers is half art, half science. Some of my former clients have compared it to alchemy. Lasers come in many different colors (wavelengths), and scientists have found that certain wavelength lasers work best for fiber optic communications. (The reason why is currently beyond me, but it apparently has something to do with the composition of the fiber itself and the need to avoid loss and crosstalk.) For source lasers, 1310nm and 1550nm lasers are used. For pump lasers, 980nm and 1480nm lasers are used. (nm = nanometers, a measure of wavelength). JDSU offers all four of these lasers as products. Source lasers will be roughly 15% of JDSU's 1999 revenue (estimated pre-merger), with Ciena, Lucent and Siemens as major customers. Pump lasers will be roughly 12% of 1999 revenue (pre-merger), with Lasertron and Nortel as major customers. B. Modulators Modulators are devices that encode data onto a light stream for transmission. In english, they flick the light on and off to represent the zeros and ones being sent as data. This can be done two ways: (1) direct - literally switching the laser on and off; or (2) external - keeping the laser on but having a piece of material in front of it which switches from opaque to transparent. Direct modulation is simple and inexpensive since no other components are required for modulation other than the light source. However, for high-speed communications, external modulation is preferred because it is more stable and can operate at higher laser power, allowing signals to travel longer distances before amplification. External modulators can also "turn on and off" faster, thus allowing for the transmission of higher data rates. External modulators will be about 21% of JDSU's 1999 revenue (pre-merger), with key DWDM customers including Alcatel and Ciena (and cable customers including Ortel and Siemens). C. Fiber Gratings Fiber gratings are essentially filters, reflecting one particular wavelength and letting all other wavelengths pass through untouched. (They work much like a piece of colored glass.) Gratings are an essential building block for the "wavelength terminal multiplexer" - the device that sits at the "end" of the fiber and splits the light back into all of its individual color components. They are also used in "add/drop multiplexers," which allow the telcos to add or remove a single color (data signal) from the "middle" of the fiber while allowing the other colors to continue onward. They are currently two popular gratings, the fiber Bragg grating and the planar waveguide grating. Their inner workings go way beyond the scope of this report. JDSU presently makes fiber Bragg gratings only. Planar waveguide gratings are used (and made?) primarily by Lucent. Fiber gratings comprised 1% of Uniphase's estimated 1999 earnings, but will be a far more substantial percentage of revenue post-merger, because JDS Fitel dominated that market. D. Receivers A receiver converts an optical signal into an electrical signal. In short, it is the "eye" at the end of the fiber which detects whether the light is on or off at any given moment. It consists of a photodetector, an amplifier, and circuitry. A DWDM system will need one receiver for every color it handles. Receivers constituted less than 1% of Uniphase's pre-merger revenues. E. Wavelocker A wavelocker is a device which keeps a light signal finely tuned on a particular wavelength. It is used in conjunction with the external modulator to produce the many precisely calibrated signals generated in DWDM systems. JDSU just introduced its wavelocker product in 1999. F. Other Assorted Items "Couplers" are basic building blocks which are used to combine and split signals in an optical network. "Isolators" are used in front of lasers and amplifiers to prevent reflections from entering them. Both are considered commodity components. "Wavelength converters" convert data from one incoming wavelength to another outgoing wavelength. All optical wavelength converters are not commercially available yet. G. Optical Modules JDSU is also beginning to offer fiber optics "modules" - again, not complete DWDM systems, but larger building blocks that integrate a number of smaller components. One of its initial offerings in this area is an EDFA, which is apparently seeing good acceptance in the marketplace. According to the recent conference call, other modules are still in the design phase. These include configurable multiplexers (the devices that take the different streams of light and combine them into a single beam of light) and dynamic add/drops (which will be able to "pluck" different colors out of the light stream on the fly). H. The Holy Grail - Optical Switches Although not a JDSU product, the optical switch is notable because it is the holy grail of optical networking. To date no one has definitively succeeded in creating one (although a company called Corvis claims success, and many others are working on it). Switches are commonplace in computer networking; they take an incoming electronic signal and send it along one of multiple paths, depending on the content of that signal. An optical switch would do the same thing, except that it would be switching light, without ever converting it to an electronic signal.

To: Kent Rattey who wrote (1856)	11/8/1999 7:37:00 PM
From: Kent Rattey	Respond to of 24042

Part 2 IV. MARKETS, CUSTOMERS & COMPETITORS A. Overview of Markets JDSU has competed in three major market areas, but only one of these - telecommunications components - is the primary focus of JDSU. Although all three are briefly addressed below, the vast majority of this report addresses only telecommunications: (1) Laser Subsystems - This was Uniphase's original business and, until 1994, its primary business. JDSU sells a variety of lasers that are used in many different industries, including biotechnology, semiconductor, consumer electronics and industrial process control. This industry is considered mature, and is expected to grow at only 15% for the next few years. In the most recent quarter, this market represented only 15% of Uniphase's revenues. (2) Semiconductor Equipment - until recently, JDSU sold a laser imaging system for examining defects in semiconductors, known as Ultrapointe systems. JDSU recently sold this subsidiary, and thus this will no longer be a source of revenue (or losses) going forward. (3) Telecommunications Components - this is THE area of growth in JDSU, and is the primary focus of this report. This market generally includes not only telecom equipment companies but cable companies as well. JDSU, however, has confusingly placed cable in a separate "transmission and testing" segment. Over the past quarter, telecom represented 59% of revenues, and cable represented 26% of revenues. B. Discussion of the Major Fiber Optics Markets Both the analysts and JDSU's CEO (in the conference call) generally break the fiber optics market into four segments, each of which has its own unique requirements, growth patterns, and customers. This categorization is essentially based on where the fiber is deployed: (1) Terrestrial (Long Haul) - this is probably what you think of when you picture DWDM. Fibers are being strung long distances, from point A to point B, and DWDM and EDFAs are being deployed to increase bandwidth. Demand in this market segment is extremely strong. CEO noted in the conference call that they were seeing a continued progression towards higher channel DWDM systems, which has resulted in greater demand for JDSU's more advanced products. (2) Submarine - conceptually, this is no different than the terrestrial market. Fiber is being strung along the ocean floor to provide intercontinental fiber networks, and DWDM is being used to increase bandwidth on those networks. In practice, however, the requirements for submarine use are far more demanding than those of terrestrial, primarily because of reliability. Because it typically is not easy to search thousands of miles of open sea and then descend to the depths of the ocean floor to replace a small component, the customers demand that the components last at least 100 years without failure. Because of this, submarine products can cost up to twice as much as their terrestrial counterparts. As with terrestrial, the CEO noted a progression towards higher channel systems in this market. In the last quarter, submarine sales comprised 12% of pre-merger revenue, with that percentage expected to rise to 20% in the future. (3) Cable - I didn't understand this at first, but it turns out that the cable companies are spending a lot of money replacing their cable with fiber. This is because they intend to deliver a lot more services through the cable (such as internet access, phone services, video on demand), and these are going to require bandwidth in both directions. (Ordinary cable doesn't really offer bandwidth for information leaving the home, as opposed to entering the home). So they are requiring many of the components that JDSU sells. In the conference call, CEO stated that cable had "very strong growth" over the past quarter, and was a "very strong player" in the company's future growth. As mentioned above, cable accounted for roughly 26% of pre-merger revenues. (4) Metro - This area is still in its infancy, but the expectation is that metropolitan areas will be wired with fiber in the near future, and that DWDM will be a prominent part of these regional networks. Many of the questions on the conference call dealt with speculation as to when this area would "take off." CEO predicted that it "would hit" in 2001, although all indications are that this was a conservative estimation. CEO noted that at the recent SuperComm trade show, there were more than 30 start-ups displaying new devices for the metro market, most of which contained JDSU products. Note that EDFAs should not play a large role in the metro market, because amplification is only needed when signals are travelling more than 100 miles or so. Because the metro market has not reached critical mass, it is unclear exactly how big the opportunity will be in this area. Everyone is in agreement, however, that it will be big. According to the conference call, estimates range from a low of 70% of the long-haul market to a high of three times the long-haul market. C. Customers JDSU's customers are generally equipment manufacturers such as Lucent and Nortel. They vary, however, depending on the specific market. Moreover, because of the breadth of JDSU's product range (and the varying integration of its customers), a number of customers in one area are competitors in another area. Customers for telecommunication and cable television components include: Ciena, Northern Telecom, Lasertron, Marconi, Lucent Technologies, Alcatel, Ericsson, Harmonic Lightwaves, KDD (Fujitsu), Ortel, Pirelli, Siemens, Synchronous. Customers for cable systems include: General Instruments, Scientific Atlanta, Philips Broadband, Silicon Valley Communications. Customers for commercial lasers include: Applied Bio (Perkin Elmer), Coulter, KLA-Tencor, Kodak, Xerox. The above customer list was drafted for Uniphase only. I have not seen any filings for JDS Fitel, but news reports and analyst reports indicate that the two companies had very similar customer lists. JDSU's largest customers are Lucent and Nortel. On a combined post-merger basis, sales to Lucent constituted 17% of revenues in the past quarter, and 13% for the year. Sales to Nortel constituted 10% of revenues for the year. No other customers accounted for more than 10% of combined revenues for either the quarter or the year. D. Competitors JDSU's competitors vary depending on the specific product category. The competition in each area is relatively small. The most significant competitors are probably SDL Inc. and E-Tek Corporation. JDSU also faces competition from vertically integrated equipment manufacturers, such as Lucent and Nortel, that have the ability to manufacture components internally. Competitors in the laser market include: Alcatel, E-tek, Fujitsu, Lucent Technologies, Nortel, Ortel, SDL Inc., Siemens, Sumitomo. Competitors in the modulator market include: Pirelli, Lucent Technologies, SDL, Kyocera, Sumitomo. Competitors in the Bragg grating market include: Lucent Technologies, E-Tek, Corning Competitors in Erbium Doped Fiber Amplifiers include: Pirelli, Corning, Lucent, Ortel, Nortel. Competitors in the CATV market include: AEL, Harmonic Lightwaves, Silicon Valley Communications, Ortel, Synchronous Communications, CFX Communications, Radiant Communications. Perhaps the most intimidating of these competitors are the vertically integrated equipment manufacturers, since these companies are JDSU's customers as well. In theory, there is nothing keeping Lucent, for example, from completely cutting off its business and doing everything in-house. In practice, however, all indications are that the reverse is happening. Analysts have been repeatedly noting a trend towards increasing outsourcing by vertically integrated "competitors" thus favoring suppliers who can offer one-stop shopping and reliable components. Evidence of this trend was given in the recent conference call, where CEO noted that previous estimates for 1999 were for a DWDM market of $3 billion, of which 70% would be from internal sources and 30% would be from merchants such as JDSU. In fact, however, not only is the market twice as big (now estimated at $5.5 billion), but it is split 50/50 between internal and merchant sources. Moreover, analysts are now predicting a market in 2003 of $21.3 billion(!), of which only 30% will be internal and 70% will be captured by merchants such as JDSU. JDSU is seeking to actively push this trend, having publicly stated that its recent secondary public offering (which raised $___ million) was in part to allow it to provide an "exit source for vertically integrated companies that want to outsource manufacturing." In other words, look for JDSU to buy out the fiber optics components divisions of a number of these companies. None of JDSU's "merchant" competitors has its breadth of product lines, and thus none can offer systems manufacturers the "one-stop shopping" that JDSU offers. There is a trend towards consolidation in the industry, but no company appears to be focused enough or fast enough to even keep up with JDSU, let alone gain any ground. A good example of this is SDL, Inc., a company that manufactures a wide variety of lasers. SDL appears to be quite competitive in the 980nm pump laser market, and recently acquired a company named Integrated Optical Components, Ltd., which specializes in external modulators. At the same time, however, SDL continues to focus much of its activity on different lasers for non-telecom applications, such as printing, medicine, etc. By spreading itself thin, and by not growing quickly to expand its product line, SDL is consigning itself to the role of niche player in the industry. V. BARRIERS TO ENTRY (OR, WHY JDSU WILL WIN) A. Patents JDSU owns or exclusively licenses at least 470 patents relating to optoelectronics and photonics. (Uniphase was reported to hold or license 421 patents, and an online search revealed at least 49 patents held by JDS Fitel.) I have not gone through any of these patents, but some reports indicate that a number of these are ground-breaking, particularly in the area of fiber Bragg gratings. B. Top-notch Staff There are a limited number of scientists who can perform the research, development and manufacturing necessary to compete in this field. An enormous number of these scientists work at JDSU - the reports indicate that Uniphase had 116 optical scientists before the JDS Fitel merger. JDSU's continued success relies on keeping those employees happy, but given the company's cutting-edge work, its dominance in the industry, and its stock performance, that probably is not a problem right now. I don't know how many top-notch optics labs exist out there, but a potential competitor would have to get its scientists from somewhere. This would be difficult now, but it will get even harder as JDSU implements its announced plans to expand through further acquisitions. JDSU has not only shown a willingness to move fast, but also, like many of the internet companies today,

To: Kent Rattey who wrote (1856)	11/8/1999 7:41:00 PM
From: Kent Rattey	Respond to of 24042

Part 3 it has the power of a high-flying stock to use as currency for acquisitions. C. Technological Barriers Virtually every technology or internet company claims the difficulty of duplicating its technology as a barrier to entry. Much of the time, this is complete BS, but in the case of JDSU, it is truly an enormous barrier to entry. Even ignoring JDSU's patents, the logistical and technological challenges of making these components is overwhelming. Active components such as lasers are created in reactors by precisely flowing layer after microscopic layer of gas over a semiconductor wafer, precisely chopping the wafer into thousands of individual pieces, processing each piece to form a laser, and then packaging and testing each one individually. Although it requires great scientific knowledge, it also requires an almost artistic "touch" which comes only from experience - my former clients at Xerox PARC compared it to alchemy. Even making these components would be a huge challenge for any new competitor. Manufacturing the components with enough reliability and volume to challenge JDSU would be a far greater feat. Many of these components are required to perform for 100 years without failure, which is an incredibly difficult engineering feat. (Try and think of any electronic component you know of that could run continually for 100 years without failing. It's hard.) Producing them in bulk is even harder. As an example, one analyst report indicates that JDSU is the only qualified supplier of 980 nm pump lasers and submarine qualified fiber Bragg gratings. (During the recent conference call, CEO did not remember for sure whether JDSU was still the only one.) D. Size JDSU is more than twice the size of any of its competitors. This presents a number of strategic advantages. First, it has the ability to make the enormous capital expenditures necessary to keep up with demand in the industry. JDSU will be spending millions of dollars to expand its capacity across its entire product line. Competitors such as SDL are selling all they can make, but do not have the financial capability to ramp operations to the extent that JDSU does. Second, JDSU's sizes gives it the ability to offer an incredibly broad product line that no one can match in the industry. This allows its customers the convenience of one-stop shopping. It also allows it to accommodate its customers by, for example, keeping large inventories of products at the customer's site on a consignment basis; a smaller operation would have neither the manfacturing capacity nor the cash flow to allow this. Furthermore, as JDSU's product line becomes more and more complete, it is able to vertically integrate upwards, offering more and more complete modules to its customers. For example, JDSU is beginning to sell completed EDFA modules, rather than just pump lasers, to its customers. JDSU is also designing other modules, such as configurable multiplexers, switching modules, wavelength lockers; and dynamic add-drops, generally working with the customer to custom-design a solution. No competitor appears big enough to do this on any kind of scale. VI. PROJECTIONS FOR INCREASING DEMAND A. DWDM Industry Demand [UNFINISHED – NOTES ONLY] Demand began with long-haul, point-to-point. Short term growth there, with long term growth in metropolitan market and LAN/WAN applications. Hambrecht & Quist July 1998 report estimates DWDM market at roughly $1B; estimates compounded annual growth rate of more than 150% through 2002, to $7.7B. "Systems providers have said they could have shipped more systems if components had been more readily available. Therefore, regardless of market share, we believe that every company in the optical component market should thrive this year...." B.T. Alex. Brown B. Demand for JDSU Products [UNIFINISHED – NOTES ONLY] Uniphase has two main lines of business: opto-electronic telecommunications components and laser subsystems. The telecommunication components are the most significant contributor to Uniphase's revenue and revenue growth. It is estimated that in FY99 telecommunication and cable television equipment (excluding JDS Fitel's contribution) will be 79% of revenue. There are estimates that this portion of the business can grow by 88% in FY99 and 67% in FY2000. Laser components are used in various applications including biotechnology, semiconductor manufacturing, consumer electronics and industrial processes. It is estimated that in FY99 laser subsystems equipment (excluding JDS Fitel's contribution) will be 21% of revenue. Uniphase is a leading supplier of the sophisticated components that are critical to fiber optic communications equipment. (Market share up to 50%, Kevin Kalkhoven 3/17/99) The company's highly reliable and proven product line is the industry's most comprehensive, and Uniphase enjoys leading market shares for may of its core products Needham estimates revenue growth at 50-70% per year for foreseeable future Pump lasers - tripled capacity to 150,000 per year, and planned to double again by end of 1999. Now shooting for double by Sept 1999, and has ordered another reactor to double again in 2000. Currently production limited - i.e., can sell as many as it can make. Modulators and wavelockers - expected to grow 15-25% percent sequentially by quarter. Source lasers - yield problems have limited supply. Currently yield limited, and can sell whatever they can get off their wafers. Yoy telecom & cable revenue up roughly 100% Strong growth in cable, 980 pump lasers, semilasers passives Growth in JDS Fitel was across the board VII. GROWTH THROUGH ACQUISITIONS Perhaps the biggest challenge facing JDSU (and its competitors) is keeping up with the incredibly growth in demand for its products. Thus far, this has been accomplished by not only internal investment but also by strategic acquisitions. Indeed, although JDSU has started some new operations on its own (notably its cable operation in Pennsylvania), most of its size and breadth of product line is attributable to a string of acquisitions over the past five years. The following table shows the key acquisitions in JDSU's history: Year, Company Name, Location, Products 1995, United Technologies Photonics, Connecticut, External modulators 1996, GCA Fibreoptics Ltd., UK, Laser packaging 1996, Fiber-Optic Alignment Solutions, Chicago 1997, IBM Laser Enterprise, Switzerland, Pump lasers 1997, Indx, Australia, Gratings 1998, Philips Optoelectronics B.V., Netherlands, Source lasers 1999, JDS Fitel, Ottawa, Passive components. Every indication is that JDSU's pattern of growth through acquisitions will continue. JDSU just completed a secondary stock offering which brought in $___ million in cash, and CEO indicated that the money will be used for the following three purposes: (1) technology acquisitions; (2) marketplace consolidation; and (3) exit source for vertically integrated companies that want to outsource manufacturing. Note that all three of these amount to the same thing - namely, a shopping spree. At the same time, CEO has indicated that JDSU will spend millions of dollars to invest in expanding its capacity through new factory space, equipment, etc. Note, however, that all of this new equipment will need new people to operate it, and that such people are difficult to find. As a result, expect that acquisitions will be a very important part of future growth, if only to acquire the talent necessary to grow. My personal assessment is that the difficulty of sustaining growth - not due to lack of demand but rather difficulty of expanding capacity - is the biggest issue facing JDSU, and is the biggest cause for concern in considering an investment. As discussed below, the price of JDSU's stock incorporates very high growth expectations, and operational setbacks could result in serious punishment by the market. VIII. MANAGEMENT I don't have much ability to personally assess the competence of JDSU's CEO Kevin Kalkhoven, but I can say that (1) the company's past performance under his leadership has been outstanding; and (2) analysts that follow the company give him high marks. He appears to be pursuing the right strategy of focusing on the high-growth area of telecommunications, by both concentrating investments in that area and selling off operations outside of that focus (such as the Ultrapointe line of inspection products). Institutional investment in JDSU is extraordinarily high - more than ___ percent - reflecting confidence in management and in the future of the company. "We consider Uniphase management to be a highly capable team that is fulfilling the strategic goals it set for itself years earlier. Through a series of successful acquistions, Uniphase has become the world's leading merchant of optical component supplier to the telecom and CATV equipment markets. We believe management's entrepreneurial spirit is what has enabled the company to succeed in a highly dynamic marketplace. Management and directors own approximately 4.3% of the company." (F.A.C/Equitites, January 26, 1999, Srikauth & Soloff ) IX. FINANCIAL PERFORMANCE [UNFINISHED] It is extremely difficult to analyze the specific numbers for JDSU, because (1) its many acquisitions have generated huge amounts of good will which must be amortized over time, resulting in theoretic "losses" under GAAP; and (2) its merger with JDS Fitel has nearly doubled its revenues and product line, making its most recent earnings numbers difficult to meaningfully compare with past quarters. Most analysts are evaluating JDSU using its "pro forma" numbers, which exclude the huge write-offs of good will from the acquisitions Having said all that, the financial numbers for JDSU (the company) are great. The pro forma numbers for the combined company for the most recently completed quarter (Q4, ending June 31, 1999) are as follows: Revenues $ 191,561 Cost of sales 89,824 Gross profit 101,737 R&D 16,844 General & administrative 23,306 Total operating expenses 40,150 Income from operations 61,587 Interest and other income 2,226 Income before taxes 63,813 Income tax expense 22,132 Net income 41,681 Net income per share $ 0.48 These numbers reflect an extremely healthy company. Highlights include: - Gross margins of 53% - Operating margins of 32% Bad parts include: - Tax rate of 34.7% (that's what happens when you make money) Balance sheet: $234M in cash Combined company generated $146M in cash flow from operations for year, v. $105M in capital spending UNPH Accounts receivable went up from 62 days to 66 days; JDS AR went down from 56 to 48 days Inventories are up - relationships with customers require consignment, etc. Guidance: When asked about the current quarter, guidance was given as "we usually meet or beat the estimates." (Kevin Kalkhoven, 3/17/99) As before, 15% growth sequentially; Challenge is to expand capacity Expect margins to be along recent results - 50+ gross, 29-30% operating margins Tax rate 35% Guidance=50-52% gross margins, but did better than that this past Q. For fiscal Y2000, expect growth to be 60% or greater yoy!! Management believes it can exceed its guidance. Uniphase has no long term debt. The financial numbers for JDSU (the stock) are not as mind-blowing, solely because of its current high valuation. Simply put, like most other successful high-tech companies, investors have begun to discover JDSU and reward it with a very high price. Looking backwards at the $1.47 per share earnings reported for the year ending June 30, 1999, and its current stock price of 180 (not accounting for the recent split), JDSU trades at a P/E ratio of 122(!). Using the company's (conservative) guidance of 15% sequential growth, I get annual earnings for FY2000 of $2.76 per share, resulting in a going forward P/E ratio of 65 - more reasonable, but still quite high.