Cascade Microtech (CSCD)

A cash-rich, dominant, market-share leader, Cascade Microtech is a semiconductor wafer-probing equipment manufacturer, with a razor/blade (80/20) model. In addition to this stable core business, CSCD has a smaller Pyramid Probe card division growing at 70% a year. The larger division, the Engineering probe business, has a diversified customer base and end-markets, premium product and pricing, a razor/blade (80/20) model, a majority market share (more than 50%); revenue stream of $64M (with 20% operating margins) and is growing at least 7-8% per year – we think this should be conservatively worth $120M to a strategic buyer.

 

Cascade has an EV of $101M or $8.49 a share – market cap $156M, net cash of $55M. At an EV of $101M you are now being paid handsomely to assume the call-options in the smaller but fast-growing Pyramid probe business and the nascent life-sciences business. The Pyramid probe business could become a $75m business (25% market share of addressable market) with 20% OMs in 3-4 years valuing it at another $120M. We think the business could be worth at least $300M or $25 a share in 3-4 years while current downside in these shares is extremely limited.

 

Growth in the smaller, Production Probe division has been held back due to capacity constraints during 2006 and that might continue to be the case despite a 90% y-o-y increase in unit growth. While CSCD has gained share, currently long lead times are causing some customers to go with conventional probe cards. We think the company will leverage from adding capacity in 2007, resulting in higher market share and stronger margins. The company has so far, during 2006, showed solid performance but the major opportunities for the company will be in the year or two ahead. We believe CSCD could achieve $1.39 in EPS by 2008 giving it a compelling sub-7x multiple on FY08 earnings, net of cash. Comparable like Formfactor trades at a significantly higher EV/E multiple of 27x ’06 and 21x ’07. We think is management (with 40% ownership) is excellent and their technology is superior and leading-edge and the stock gets lumped with other, far more cyclical semi-equipment operators. CSCD’s business, we believe, is less cyclical and more dependant on relatively stable R&D budgets than on capacity expansion.

 

 

Price                     $13.10                 2007E

S/O                        11.9M                   EV/E                     10.1x

Market Cap         $156M                 EV/FCF                10.8x

Debt                      0

Cash                      $55M

EV                          $101M

What does Cascade do?

 

·         Cascade makes advanced semiconductor wafer-probing equipment that measures and test electrical characteristics of Integrated Chips (ICs). Cascade makes testing machines and probes (consumables) for both the engineering/design testing and production testing phases.

·         Two segments - Engineering Probe Stations and Production (Pyramid) Probe Cards; Engineering Probes Stations (systems/consumables split 80/20) makes testing equipment and analytical probes for engineering tests and had $64M in revenues with 20% operating margins while Pyramid Probes designs and sells probes for productions testing in logic chips and had $10M in revenues in 2005 (growing at ~70% annually to $17M in ’06).

·         Engineering Probe stations address an approximately $150M market growing at 7-8% average per year while Pyramid probes address a bigger $300M market growing at 15+% per year in which CSCD is rapidly gaining market and mind share.

·         Stable Core Business with a Call Option on Promising New Businesses – CSCD has a more than 50% market share in probe stations (70% share in 300mm wafer technology) making the company the preferred and indisputable leader in an overall market of $150M growing at 7-8% a year. The market has high barriers to entry, has just one other small player and CSCD charges premium prices for its premium technology. Their smaller (revenues - $10M in ’05; $17.5M in ’06) Pyramid Probes business (entirely consumables) is growing at 70% per year and addresses a much larger $300-$500M market with leading technology as well.

·         Good Growth and Improving Margins - Capacity-constrained in the last 12 months, CSCD is hiring new headcount and expanding capacity to address a pressing need for its Pyramid Probes. Pyramid Probes market is fragmented with many small players and CSCD has the potential to win 25-50% share in this business over time. Increased spending on capacity expansion for the smaller division should be mitigated by higher unit volumes by late 2007, thereby leading to much improved overall operating margins. Overall OMs should go from 5% to closer to 14% by late 2007 and 18% in 2008.

·         Diversified Customer Base – CSCD counts Agilent, Broadcom, Freescale, Fujitsu, Hynix, Intel, Samsung and many others are among its Top 20 customers. CSCD also sells to various end-markets within the semiconductor space making it less vulnerable to intense cyclical pressures.

·         Supportive Industry Trends and Muted Cyclicality – Continuing growth in semiconductor sales coupled with intense pressure on chip companies to incorporate increased functionality in its chips while reducing chip size, makes chip design, fabrication and testing very difficult. Cascade’s business cycles are less pronounced than other semiconductor equipment companies since Cascade’s sales are more dependant upon R&D capital spending rather than capital spending for capacity expansion.

·         Solid Balance Sheet with $55M in net cash (or $4.60 per share) – money raised from IPO in late 2004 is still awaiting judicious use. Management would prefer to do an acquisition rather than buyback shares to not impact the low float any more.

·         Sector Doldrums – Cascade gets lumped in with the rest of the semi-equip manufacturers despite its appealing and superior characteristics. Street is concerned that capital equipment-spending cycle might be on the downswing – Cascade’s business is more dependant on R&D cap spending rather than on spending for capacity expansion. We also think business is cyclical but not as much as widely believed and that cyclicality is already priced into the stock.

·         Capacity constraints – Pyramid probes will fall short of management’s internal goal of doubling Pyramid revenues primarily due to capacity constraints this year. Currently long lead times in meeting demand have caused some concerns on the Street.

·         Microcap, liquidity/float – Insiders control 40% of the company and shares don’t trade much (less than 50K a day) making it unattractive for a sizable holding for institutions. 

 

 

A wafer probe card is an essential piece of hardware used in wafer testing by semiconductor manufacturing companies. Wafer testing is the process of electrically testing individual chips on a wafer to determine if that chip is functionally good or bad. This is one of the most important processes in semiconductor manufacturing.

 

A probe card acts as an interface between a tester and the dies on the wafer. Depending upon the complexity of the probe card, there are numerous probes or pins on the card that contact the pads on each die on the wafer. Earlier generation of probe cards had pins/needles that were manually assembled. However, as devices become more complex and smaller, needle based cards are being replaced by micro-electro-mechanical systems (MEMS) based probe cards. These advanced probe cards can test more devices in one single pass (touchdown). Probe cards are customized for each new chip type and physically wear out during usage in production testing (they are consumables). Depending upon the test environment, production probe cards may last for several hundred thousand to roughly 1 to 2 million touchdowns.

 

As chip makers continue to decrease the cost of fabricating chip elements by about 25-30% per year, production test departments must similarly reduce chip test scores. A common strategy is to increase the number of chips tested simultaneously, or parallel test. At the production wafer probe stage, memory chips are typically tested 64 to 200 times at one go. Comparatively, logic and mixed-signal chips are typically tested 2 to 8 at a time due to higher number of connections per chip, increased test complexities, and in many cases, lower unit volumes. Cascade Microtech is focusing its probe card applications exclusively on logic and mixed-signal chip testing.

 

What really drives the market for production probe cards is the production unit volumes of the integrated circuits being tested. The more electronic devices people use, the more chips are needed. And this is leading to sizable opportunities for a company like Cascade. Testing growth drivers can be summarized in following bullet points:

 

·         Unit Growth; As bit growth increases, like it is for DRAM right now, there will be greater need for probe cards

·         Migration to Lower Nodes; Semiconductor device makers are constantly going to lower feature sizes. New probe card designs are needed as feature sizes shrink from 110 nm to 90 nm to 65 nm and so on

·         Wafer Level Test; By testing at the wafer level, chipmakers can identify bad die and eliminate their packaging and final testing, thus saving time and money

·         Multi-Chip Packaging (MCP); Miniaturization (smaller cell phones and other consumer devices) has led to increasing use of multi-chip packaging. One single bad die can kill the entire package, making it extremely important that each and every die that goes into a multi-chip package is fully functional. This can efficiently be achieved by wafer level testing

·         New architectures; This will necessitate the need for new probe card designs

 

 

The total probe card market is worth around $1B and growing at 20-30% per year. Half of the market consists of memory chips and half of logic chips. Market research firms suggest that the total available market for advanced probe cards will increase from $470M in 2005 to $1.3B in 2009.

 

Cascade has initially focused its efforts on the advanced logic portion of the overall probe card market and this market is today worth around $150M, and growing at a 20% clip to $300M+ in 2010, according to VLSI Research. The company commands a competitive advantage in RF (radio frequency) cards, including superior electrical performance, longer probe lifetime and multi-chip parallel testing capabilities. Their current RF probe card applications include chips and components for cell phones, wireless LAN, Bluetooth and wideband communications applications. Cascade has also expanded the test capability of their probe cards allowing them to address emerging test requirements for mainstream logic devices, especially parallel testing of fine-pitch wire-bonded logic chips. Their current applications include testing of logic and mixed-signal chips such as automotive controllers, cell phone processors or digital signal processors. Cascade is expecting to generate revenues of around $17.5M in 2006 from its probe cards business. Thanks to its competitive advantages, a market share of 25% in its addressable markets is not unreasonable in five years, making the growth potential really interesting.

Cascade Microtech was founded in 1983 in Portland, Oregon, by the current CEO and VP of Engineering. Management has built a well-regarded franchise in IC testing that sells to most semiconductor companies around the world. They are known for leading-edge technology, innovation and high quality. All management members have spent many years at the company and in the industry. They are very conservative, smart and capable people and total insider ownership measures up to 40% of the shares. Financial compensation is conservative.

 

Management has owned around 40% of the company since before the IPO about 5 quarters ago. They did not sell s single share in the IPO. Farallon Capital and Royce Associates have owned a significant piece since the IPO.