Cymer Q4 1999 Conference Call transcript, part 1 of 3
Bill Angus gives usual disclaimer. Omitted. Introduces new Director of Corporate Communications and Investor Relations, Terry Slavin (ph): "Terry is an IR professional with more than 16 years experience in investor relations and business communications. While Terry is new to semiconductor industry specifically, he's already begun to familiarize himself with the ins and outs of this dynamic sector, and we expect he will bring himself up to speed quickly." Also issues the following notice: "As an aside, I'd like to mention that, beginning with our reporting for the first quarter of this year, we will issue our earnings releases and hold our conference calls on Tuesdays rather than Mondays."
Bob Akins, Cymer CEO:
Good afternoon, ladies and gentlemen and welcome to our first conference call of the year 200. I won't say the first conference call of the new millennium because we at Cymer do indeed understand the difference and I'll save that welcome for next year. Whether it demarcates a new century or not, this year has certainly brought with it a continuation of the recovery, and overall sense of optimism throughout our industry. The accelerating demand for advanced chips now coming from the communication sector even more strongly than from the traditional driver, PCs, is in turn accelerating the demand for Cymer's most advanced products and services. Telecommunication equipment suppliers tell us that over half of their chips are now being fabricated using sub-quarter micron DUV patterning for critical layers. As demand for functionality expands, and geometries continue to shrink, progressively more layers in such advanced chips will utilize DUV. Many chipmakers, especially foundries, have already announced substantial increases in their capital equipment plans for the year 2000. Along with these trends, we have also seen the continued rise in factory and DUV tool utilization. Remember, DUV utilization has increased from a low of 20-25% in late 1998 to now approaching 100% in many factories, as demand for sub-quarter micron patterning has surged. And as a result, a clear lack of capacity for patterning at 0.18micron and below has developed. This in turn has intensified the industry's search for understanding as to how the critical wavelength transition points - for example KrF to ArF, or ArF to F2 - must be managed to enable the aggressive ongoing geometry shrink. Failure to successfully and smoothly manage these critical manufacturing paradigms cannot, in reality, be tolerated by any chipmaker, and we at Cymer are progressively being pulled more into the middle of such decision-making processes. I will offer some further observations on the subjects of increasing utilization and Cymer's role in determining and managing the wavelength transition roadmap later in today's call. The foundry expansions, chipmaker capacity increases, and transitions to more advanced technologies help explain Cymer's strong results in the fourth quarter of 1999. I assume that you have all seen a copy of our financial results, and have a little time to review them. Our financial performance continued to improve, reflecting the growing transition to our newest, high-performance products as well as growth in revenues. Revenues and bookings again grew strongly and sequentially, and the bottom-line results were better than we anticipated, due in part to higher sales volume, further shifting of product mix to our newest products, and our ongoing managing of costs. I will report on operational activities for the quarter beginning with order rate and new product acceptance. Then I will discuss the progress in the areas of technology and product development, manufacturing, and worldwide customer support activities. Bookings for the quarter were at 97.9 million dollars, and backlog increased to 102.7 million dollars as of December 31, compared to backlog of 81.7 million for Q3, and 59 million dollars for Q2. Revenues continued a four-quarter sequential rise, driven primarily by the shipments of our ELS-6000 model, which represented 45% of new system shipments in Q4, compared to only 20% in Q3. Our order rate for the 6000 increased again during the quarter, as desire for the increased productivity of 2 KHz KrF drove demand, especially for use in advanced high-NA, high-stage-speed scanners. This new model offers compelling higher performance with increased productivity, while providing the lowest cost-of-operation and highest overall equipment effectiveness to date. New orders for the ELS-6000 are again ahead of our own schedule, and we continue to anticipate increased shipments of the 6000s in Q1, and a ramp to higher volumes throughout the year. The number our lasers installed at chipmakers in the fourth quarter totaled 89 as compared to 79 in the third quarter. Our analysis indicates the Cymer lasers comprised 84% of all lasers installed at chipmakers for the fourth quarter, due in part to a higher number of installations by one of our Japanese customers at Japanese chipmakers, which tend to favor our Japan-based competitor's light sources. For the entire year of 1999, 88% of all lasers installed at chipmakers were Cymer's. On a cumulative basis, this translates to a total of 905 Cymer laser light sources at chipmakers at the end of December. These are located as follows: 271 in the U.S., 207 in Japan, 148 in Korea, 135 in Taiwan, 107 in Europe, and 37 in Singapore. We were pleased to announce in November the installation of our 200th system at chipmakers in Japan. This gave us a greater than 80% cumulative market share in Japan, while our overall cumulative market share remains at approximately 94%. As an aside, of the 200 systems in Japan, approximately one-third were built by our manufacturing partner in Japan, Seiko Instruments Incorporated. We estimate that the number of lasers in inventory at our direct customers has stayed flat or decreased slightly quarter to quarter. It remains about where it should be, given our customer's integration times, as well as the 1- and 2-KHz mix of steppers and scanners, and the anticipated industry demand for DUV tools over the near term. As I've discussed with you throughout 1999, we continue to push forward with an advanced development program, spanning four separate wavelengths, and the year 2000 will be Cymer's most prolific year for introducing new products, with plans for three new KrF light source introductions, two new ArFl introductions, and a new F2 process-development prototype, all based our latest 6000-series platform. On the 248 nm front, we see the high certainty patterning down to 0.13 micron for the majority of chipmakers, and promise for some patterning for geometries down to 0.10 micron. Since KrF is expected to constitute the vast majority of DUV tools delivered during the next several years, we have aggressive development programs underway designed to extend our KrF product leadership through new KrF products, with narrower bandwidth and increased stability to power the next generation of high-NA, higher-resolution scanners. The success of ArFl 193 nm light sources will depend on their ability to operate at very high pulse-repetition rates, 4 KHz and above, while achieving an attractive cost-of-operation. Our ELS-5000A ArFl light sources are currently powering most of the process development tools being used worldwide today. We are developing an unprecedented ultra-narrow-bandwidth capability to accommodate lens designs, which could incorporate a minimum of calcium fluoride, as this material has yet to be production proven. Combining high-rep-rate operation with ultra-narrow bandwidth performance will enable our customers to introduce high-NA, high-productivity, high-dosage-accuracy 0.13 and 0.10-micron capable scanners. F2 157 nm light sources are not excimer lasers. They are similar in nature, but at the same time very different in character. We have been conducting F2 scientific investigations since early 1999, and feel that we have gained some fundamental insights as to what this technology is, and is not, capable of. Adequately narrowed and stabilized, F2 can power generations of 157 nm refractive scanners for sub 0.10 production. If narrowing cannot be practically achieved, then catadioptric or partially [unintelligible]-based projection scanners will be required. The quality and pace of our F2 investigations has earned Cymer the industry-determinant scientific leadership position. We announced in November that we received our first F2 order from Ultratech Stepper, for our ELX-6500 F2 light source, to be integrated into a small-field microstepper for next-next-generation process development applications. The light source is scheduled for shipment in the first quarter of this year. We continued our basic research on our 13.5 nm dense-plasma focus light source. We are more enthusiastic than ever before about the potential of our EUV source to power lithography tools in the post-optical era. Cymer's ability to manufacture light sources of ever-increasing complexity, and value-added technology, is proving to be an extremely valuable competitive weapon. And we are of course demonstrating that manufacturing capability both in the U.S. as well as in Japan. Our investment in raw manufacturing capacity, which served us well for the ELS-5000 ramp of 1997, has formed a foundation which now enables us to produce a growing diversity of products. The number of new products and prototypes, spanning three wavelengths, that we plan to ship over the course of the year 2000, will challenge the diversification of our overall capacity, including final test capacity. In spite of this, we have set aggressive targets for improving manufacturing gross margins over the course of the year, and we plan to achieve those through a combination of increased volume and capitalizing on our ongoing process development and improvement efforts. During the quarter, we continued to invest in our regional sales, marketing, and service capabilities. With our installed base of light sources in production at chipmakers numbering over 900, the successful maintenance and uptime of those machines is paramount. To continue to bring benefits of technology improvements to those systems in the field, sales and installations of our ELS-5005 upgrade kits continued from two leading DRAM manufacturers and a number of Taiwanese foundries. We received over 70 upgrade orders that enabled chipmakers to lower cost-of-operation through improved longer-lifetime key modules and extended warranties. Cymer's efforts to improve basic business and manufacturing processes, though still most part in their early phases, are steadily progressing. We are convinced that our attention to processes and the efforts we are making to implement them, though they demand more in time and resource in the short term, will make Cymer a more efficient enterprise, better able to serve the market and continue to enhance shareholder value. Over time, the process investment we make today will pay large dividends. We are actively engaged with the implementation of a new Oracle applications-based enterprise resource planning, or ERP, system to augment our information technology infrastructure necessary to support and integrate critical processes quickly and efficiently. We expect to be online with the new system in San Diego in the 3rd quarter and to deploy it globally in the 4th quarter. I will now turn this report over to Bill for the 4th quarter financial review.
Angus:
Year on year, G&A expense increased 38% ... This increase is primarily attributable to the increased commitment on the part of management to build the process infrastructure necessary for growing the company in the future. This effort will be continued in 2000 with the full implementation of an ERP system. As Bob previously stated, we expect the first phase of this project to be operational during the 3rd quarter of this year.
Looking forward, with the information we have available at this time, we believe our revenues can grow 35 to 45% per year over the next two years. But of course, on smaller quarterly increases than we have been recently experiencing. |
