Overview

Cascade was originally written up on VIC back in 2006 (http://www.valueinvestorsclub.com/value2/Idea/ViewIdea/2479) and warrants another look today.

Cascade Microtech (“CSCD”) is a small cap, semicap test equipment company with increasingly dominant market share in each of its business, and favorable secular tailwinds.  The company is growing both organically and via smart, strategic acquisitions, is poised to see substantial margin expansion, and has $1.28/shr in cash.  In spite of these factors, Cascade is trading at 4.4x 2015 Adj. EBITDA.  It isn’t hard to understand why:  the company has been largely ignored by the market given its small size ($156m MC), poor liquidity ($400k of avg. daily volume) and lack of meaningful sell-side coverage.  The stock has traded off in recent months on the back of two lackluster quarters in which the company hit guidance but failed to surpass expectations.  The stock is also being held back by persistent open-market sales of stock by the company’s aging founder, who has been stepping back from his responsibilities at the company.  Despite his sales, there is substantial insider ownership and a major shareholder sits on the board.  In addition, Becker Drapkin Management, a fund with a record of successful activism in the small cap space, controls more than 7% of the stock, though they do not appear to have made any demands of the company.

About the Business

Cascade operates in two separate but related businesses, each focused on wafer probing – measuring the electrical properties of on-wafer elements.  Engineering systems and probes made up 80% of 2013 revenue, production probes accounted for the other 20%. 

Engineering Systems and Probes (Systems:  $79m of revenue; Probes:  $17m)

In its Engineering Systems business, Cascade sells probe stations and related analytical probes used to test the electrical properties of microprocessor elements during the chip design process.  These systems, which function like a “razor and blade” model – though one where the razors are the primary driver of profitability – are primarily used in R&D labs, and are a critical piece of technology for anyone in the semiconductor ecosystem, including integrated device manufacturers, fabless semiconductor makers, foundries, and government and university research labs.  System ASPs, which average ~$250k, range from less than $100k for low end systems found in university labs to over $1m a unit for the fully-featured tools used by the likes of Intel and Texas Instruments.  A system owner might have a dozen or more probes for their systems, each costing $2,000 to $4,000.

-  Cascade is the dominant player in both the systems and probes side of the engineering probes industry.  As a result of its 2010 acquisition of rival SUSS Microtec, Cascade now controls 60% share of the $125m engineering systems market (more below).  Signatone, the next largest player, has a mere 10% share, and that share is concentrated among less profitable, low-end systems.  Despite this dominance in the systems side of the industry, Cascade only enjoys 50% share among engineering probes, a $35m market.  GGB Industries, a small, privately-held company, controls the other half of the market.  GGB made a living selling probes for SUSS and others’ systems.  Now, with the SUSS acquisition fully integrated, Cascade is capturing more and more share, all the while charging a 10% price premium relative to the competition.  We believe incremental share gains could ultimately drive a 26% increase in engineering probe revenue, and with a greater than ~55% margin flow through, lead to $2.4m in additional EBITDA (15% of LTM consolidated EBITDA).

-  The systems business is at an inflection point.  The acquisition of SUSS structurally improved the quality of Cascade’s systems business.  Prior to the deal, Cascade (35% share) competed aggressively with SUSS (25%) and Signatone (10%), often using price to win deals.  By acquiring SUSS, Cascade enhanced its technology portfolio, inherited a better manufacturing footprint and gained access to an Eastern European supply chain.  The deal brought pricing stability to the industry and allowed Cascade to rationalize 80% redundant R&D, sales teams and production facilities.  Today, Cascade is the overwhelming market leader, with the largest installed base, salesforce, and service organization in the industry.  Though integrating the two businesses took longer than it should have (2 years), the benefit to margins has been obvious.  From 2009 to 2013, Systems gross margins grew from 31.7% to 41.9%, over 1,000bps.  Some of these gains are attributable to the cyclical rebound in system sales, but we estimate half were integration related. 

In October 2013, Cascade paid $27m (45% cash / 55% stock) to acquire ATT Advanced Temperature, a maker of thermal modules for probing systems.  While the deal looks attractive enough on face value – CSCD paid 6.0x EBITDA for a business with 45% EBITDA margins growing at 40% a year – it is even more interesting when strategic considerations are factored in.  Thermals are modules that are attached to engineering probe systems that allow for testing under extreme temperatures.  Historically, 45% of Cascade’s high-end probe systems have shipped with thermals already incorporated.  Prior to the deal, Cascade bought these from one of three vendors (ATT being one of them), packaged them with its systems, and baked the cost of the thermal into its overall bill of materials (BOM).  In fact, the thermal alone could cost as much as 40% of the overall system’s BOM.  By integrating this component, Cascade will expand its TAM (management estimates by $100m), accelerate revenue growth, and increase the gross margin per system it captures.  To illustrate the last point, let’s look at the unit economics for a mid- to fully-featured system.  Let’s assume an ASP of $500k.  The cost of a third-party thermal would be around $115k.  Add to that $175k in other materials, manufacturing costs, shipping costs, etc. and it leaves CSCD with $210k (42%) of gross profit contribution.  Going forward, Cascade will capture the ~60% gross margin formerly earned by ATT.  As a result, per system gross margins (on the same ASP) will now be closer to $280k, a 33% improvement in per-system gross profit.  The full benefit of this deal will phase in over a 3-year period as Cascade’s deals with other thermal makers expire and the company shifts 100% of its business to its now-internal ATT offering.

-  Positive secular trends:  Because Cascade’s engineering systems/probes business is tied to semiconductor design rather than production, customer spending, and by extension system sales, is driven more by the secular trends in the industry than by the cyclical swings characteristic of semicap equipment capex.  In fact, R&D spend by the 15 largest semiconductor companies has grown at a CAGR of 10% over the last eight years, inclusive of the recession, during which spending declined in only one year (2009) and by only 3.5%.  This growth is a reflection of the underlying trends in chip design and production.  Smaller nodes, new materials, and more stringent environmental requirements all necessitate more rigorous and precise wafer testing.  Additionally, the progression to larger wafers bodes well for Cascade over the longer term, though the shift to 450mm wafers is still several years off at best.  If and when the transition to 450mm occurs, CSCD stands to benefit from a wholesale upgrade of the engineering system installed base.

-  New product cycle will contribute to revenue growth and margin expansion:  In late 2012 and early 2013, Cascade introduced three new system platforms.  The SMART 150, the first to launch, is geared at low volume commercial labs, and university and government labs.  Importantly, the SMART 150 consolidates two legacy systems, one Cascade and one SUSS.  By integrating the product lines, Cascade was able to lower their cost and price point, while expanding margins.  This isn’t rocket science, but for a company of Cascade’s size, things like this move the needle.  The second new system is the CM300, a top-of-the-line, fully featured system, capable of automating a significant amount of the wafer probing process.  With an ASP over $1m, this is an expensive machine likely to be found in only high volume research labs at places like Texas Instruments or TSMC.  The last new system, the APS 200 Tesla, is the most interesting of the three.  While the first two systems deepen Cascade’s dominance in the R&D lab, the 200 Tesla positions the Company to enter the adjacent market of on-wafer probing during chip production.  If the product is successfully adopted, it will open Cascade to a $60m TAM, significantly expanding its market opportunity.  Equipment sales cycles tend to be longer in the production environment than in labs, so Cascade has yet to announce any major deals involving the 200 Tesla, but to the extent they land a deal, it could contribute upwards of 20 systems per fab.

Production Probes ($24m of revenue)

Cascade’s production probes, also known as probe cards, are used by semiconductor manufacturers to test wafer reliability prior to the cutting and packaging stages of chip production.  Probe cards are pressed down onto the wafer, connecting the die or integrated circuit under test to the equipment testing it.  Each probe card is custom-designed to match the geometry of the wafer/chip it is meant to test.  During production, probe cards make millions of touch-downs on wafers, wearing down the probe tips and necessitating replacement.  ASPs range from $1,000 to $5,000 per probe.  Cascade’s probe cards are focused almost exclusively on high frequency semiconductors, which includes the radio frequency (RF) chips used in mobile phones.  Key customers in this area include Avago, Skyworks, RF Micro and Triquint. 

-  #1 share in niche market:  Production probes are used to test all types of semiconductors, with logic and memory chips accounting for the vast majority of the $1.2bn market.  These segments are large in size, but highly competitive due to the commonplace nature of the technology in use.  Given the nature of the market, players in the logic and memory markets tend to experience significant swings in revenue and profitability depending on technology cycles and design wins (for a prime example of this phenomenon see FORM).  Cascade’s business, however, is differentiated from traditional p-probe makers in that it is primarily focused on the RF market.  Because of the technical difficulties of testing at high frequencies, only Cascade and a few other small players participate in this slice of the market.  So while Cascade is only a tiny player in the broader market, it is the #1 player in the $60m RF probe market.  Most importantly, because of the more favorable market structure, RF probes have superior economics than traditional p-probes.

-  P-probes have high incremental margins that aren’t readily observable:  For accounting purposes, Cascade lumps its production and engineering probes businesses into a single “probes” reporting segment.  Unfortunately, though both businesses share the probe moniker, they have very different unit economics.  Engineering probes, for instance, require significant human labor, giving them a largely variably cost structure.  P-probes, in contrast, are largely machine-made, and have a larger fixed cost component, leading to high incremental gross margins.  Cascade doesn’t disclose the margin profile of the two businesses, but we can estimate it based on available information.  In the chart below, we compare revenue and gross margins for the probes segment for 2011 and 2013.  During this time, engineering probe revenue was basically flat, allowing us to estimate the incremental gross margin contributed by the change in p-probe revenue.  What we find is that an $11m increase in p-probes resulted in a $7.7m increase in gross profit, suggesting incremental margins of ~70%.

(1)     Italicized figures are estimated based on available information.

-   Growth driven by large secular trends:  From 2010 to 2013, p-probes grew at a 28% CAGR, making it Cascade’s fast growing organic business.  This growth has and will continue to be driven by the surge in demand for the RF content used to power mobile devices.  Every phone and tablet on the planet has some amount of RF content, and all of that content requires wafer level testing.  Global smartphones unit volumes grew 38% in 2013, and that trend should remain strong as smartphones further penetrate developing markets such as China and India.  In addition to growth in unit volumes, the amount of RF content per device is growing.  The progression from 2G to 3G to 4G necessitates inclusion of additional frequency bands, each of which requires more RF content in the form of filters and amplifier.  The movement toward “global” phones – devices capable of operating on networks throughout the world – also means more RF content.  Down the road, wearable computing devices will be another driver, though it is impossible to know today how big that market might be.

We also believe that certain trends within semiconductor design and manufacturing will benefit Cascade over the medium to long-term.  First, as the cost of chip packaging grows, wafer-level testing will become increasingly important to chip makers looking to minimize the amount of cost spent to package defective chips.  Secondly, at some point, we expect to see RF content integrated into SoCs.  If this happens, it will open up an entirely new market for Cascade’s probes.  Viewed differently, this sort of advancement would make Cascade a very attractive acquisition target for someone such as FORM who has a large presence in SoC probes but doesn’t currently have the ability to test at high frequencies.

Improving economic model

Cascade’s business is at an inflection point, and for that reason, the company’s recently reported 2013 results don’t accurately reflect the full earnings power of the business.  For starters, the 2013 figures only bake in 1 quarter’s worth of benefit from the recent ATT acquisition ($10m in annual revenue and ~$5m in EBITDA).  More importantly, though, is the change in margin structure that will take place as the company sees the benefits of its new system launches, share gains in e-probes, and growth in p-probes.  Over the next two years, we expect Adj. EBITDA margins to expand from 17% to 20%, resulting in a near doubling of Adj. EBITDA.  Incidentally, our 20% adj. EBITDA forecast comports with management’s stated “intermediate” term goals.  The last time they shared a long-term model with the Street, they surpassed in less than 2 years.  In the table below, we lay out our assumptions for 2014 and 2015. 

Valuation

At the current price of $9.35, Cascade is trading at 4.4x 2015 Adj. EBITDA and 9.0x cash EPS.  Given most of Cascade’s direct competitors are either privately held, or small pieces of larger public companies, we are not able to observe a good comp set for valuation purposes.  However, given the quality of Cascade’s business, its strong balance sheet, and our expectation of significant Adj. EBITDA growth, we believe a 6x to 7x multiple is more than reasonable.  At these levels, the stock would be worth $13.90 to $15.75, and representing 50% to 70% upside from today’s prices. 

Risks / Concerns

-  Liquidity:  Cascade has a limited float (8.5m shares out of 16.8m FD outstanding) given its large insider ownership and only trades ~$400k of average daily volume, making this one a tough stock to buy. 

 -  Founder selling:  It’s never good to see the founder and 7% shareholder selling, but in this case, we are comfortable that it does not signal challenges with the business.  Eric Strid co-founded and served as CEO from 1984 to 2008.  He then served as Chief Technology Officer until August 2012, when he retired.  He has remained on the Board since.  We view his selling, which has occurred pursuant to a 10B(5) plan, as a natural evolution of his reduced role at the company.  He is not independently wealthy and the bulk of his net worth is tied up in CSCD stock.  As of 4/15/14, Strid still held 1.1m shares.  Even if we assume Strid sells out of the remainder of his holdings, insiders will still control 25% of shares, not counting the 11.2% owned by director J.D. Delafield, giving us confidence that management and the board are well incentivized to do what’s right by shareholders.

 -  Lumpy systems sales can lead to revenue and margin swings from quarter to quarter.

 -  New system sales could be weaker than expected:  Cascade has not seen as quick of a ramp up in its CM300 and 200 Tesla machines as it had expected at launch.  Our research indicates it is a matter of when, not if these products find traction, but further timing delays would lead us to lower our growth expectations.

 -  Consolidation among p-probe customers (RFMD/TQNT) could lead to margin contraction.  We don’t view this as a big risk given the lack of alternatives to Cascade in the market, but it will be important to watch as RFMD and TQNT integrate.

 -  Delays in migration to new semiconductor nodes and processes.