# of shrs o/s: 16.2mm (Cypress owning 46%; mgmt owning another 17% - leaving 37% float); Fiscal yr ends Spet.

CPII is a highly profitable business with 20% average EBITDA margin (22% before corporate overhead) and free cashflow is very strong approximating accounting earnings. On the other hand, there appears to be no big “sexy” organic growth prospects for the overall business. Focus of using free cashflow in the next 12-24 months will be paying down debt ($200+mm net debt o/s now). This, however, doesn’t preclude potential acquisitions as management feels comfortable with debt level and if opportunity arises, they will do a deal.

Long term organic growth is expected to be mid-single digit on the top line. 20% EBITDA margin is sustainable but not much more. This will lead to double digit EPS growth w/ delevering balance sheet or accretive acquisitions.

For potential acquisitions, CPI looks for synergy to arise from leveraging its existing customer base and technology sharing across businesses. It doesn’t believe it’s constrained by B/S even today. Potential areas for acquisitions include radar, which could add other supplementary components on existing radar platforms CPI serves and medical, which could also add other components on existing imaging customers CPI is already supplying. CPI will never become a system vendor outside satcom to compete with its own clients. Typical valuation multiple for radar/medical companies is 7-8x EV/EBITDA (11-12x for growing or young/emerging tech companies) and mgmt will not do a deal that is EPS dilutive.

There appears to be no divestiture under consideration as CPI had already evaluated everything before the IPO (April ’06). Even for satcom, which is cyclical, the management is optimistic about potential secular growth opportunities in the military applications. 

Organization structure

CPI currently runs five factories in Canada, Boston and California after Eimac in San Carlos was shut down in FY06. These five factories are run on their own P&L and deal with specific end markets as follows:

Ontario, Canada (2): Satcom and Communications & Medical Products (CMP)

Beverly, MA: Radar/EW

Palo Alto, CA (Microwave Power Products, or MPP): Radar/EW (also supplying components to other plants)

Woodland, CA (Econoco): 2^nd-hand

CPI also has 40 service centers over the world.

Despite their independence, there is constant communication among five plants and CPI conducts semi-annual review of all plants each yr. They share best practices quite often, too. In addition to division President for each plant, CEO is based in Toronto so he has hands-on operational responsibility of two plants there. COO is based in California, so he watches two plants there. Monthly reports are available for senior management to track overall performance

Division or plant managers are compensated based on P&L and cashflow with roughly 50:50 split as well as overall corporate profitability.

Sales are centralized in California and run by VP of sales, though detail contract negotiation is still done by factory-level executives.

Sales breakdown by plants

MPP: $130mm (the one in Palo Alto) – roughly $25mm internal sales to Satcom+CMP

Satcom: $50-60mm

CMP: $50-60mm

Boston: $50-60mm

Econco:  $40+mm 

Total: $340mm

EBITDA Margin rank

Medical: highest (>25%)

Satcom: by definition the lowest (10-15%; low-teen norm) due to integrator nature; also no proprietary protection

Common threads among products: microwave tech + customer interface

Cost structure

COGS                                                 for non-Satcom                                    for Satcom

            Raw material:                30-35% (direct exposure                      60%

                                                To commodity only 5%; rest

                                                finished machined parts)                       

            Labor                           20-25% (10% touch)                           10%

            Utility/energy                3%

Key competitive advantage: designing capability and "extreme" technical know-how (no patent, though)

Production/Manufacturing (I only persoanlly visited one workshop within Palo Alto plant – this unit produces Helix TWT)

4,500+ products are mostly based on one basic foundation: Vacuum Electron Devices (VEDs).

By definition, vaccum environment is required since it can’t afford to have any air otherwise would comprise the product.

-         CPI uses a bell-like device to pump vacuum. There are 32 “ports” in each device so it allows 32 tubes to be pumped vacuum each time. The whole process takes several days under 600-700 © degree heat-up and upon completion  tubes are sealed. Success is measured by as few # of leaks as possible.

It’s also important to have clean environment and CPI says it has industry’s leading “clean” work room. “cleanness” is measured by # of airborn article and CPI measures such indicator three times a day. One hard part of VE manufacturing is keeping beam travel straight in the tube. In the clean room, workers use laser welders to achieve precision.

Two key mechanisms in the VEDs. First, high energy beam of electrons is created and accelerate electronic beam’s travel through the vacuum tube. It interacts with a low-level microwave input and as a result of interaction, kinetic energy is transferred from electron beam to the microwave signal. Effectively, microwave signal is amplified  and then is extracted from the device with a much higher power level.

Two key basic parts of a VED are electronic gun, which creates and accelerates electron beam and a collector, which captures energy generated (as electrons dissipate, it creates a lot of heat, and needs copper wire to collect).

There are a lot of different forms of creating interaction btw. electron beam and microwave signal. Some examples are:

            Helix traveling wave tubes (TWT): electron beams interact with a helix-shaped coil or spring in the device, enabling very wide bandwidth (at relatively low power (10 to hundreds watts)

            Klystron: electron beams interact with a series of resonant cavities linked by a beam tunnel, generating high energy (hundreds to megawatts) but narrow bandwidth (500 KHz to 30+ GHz) signals

In sum, different product forms include

1) High power/narrow band (frequency) – like klystrons

Usage includes every area

Using oil to cool

2) Less power/wider band 

Usage is largely in communication/medical

3) Lower power/wider band - Helix TWT

Usage largely in electronic warfare

Using air to cool

For Palo Alto plant I visited, 70% is Helix (@ $10-13K per unit) and 30% is Klystron (@ 35-50K each unit).

In the low frequency area, there are Chinese competitions. However, infrastructure cost is huge ($0.5mm for a small rack of machines alone and there are perhaps hundres of such racks I saw in just one plant) for more complicated products. And a lot of machines are actually assembled by CPII itself so they’re proprietary.

All the products are very customer specific with proprietary design and they’re complex and difficult to manufacture, requiring extremely high tolerance. That’s why very few capable competitors are out there

-         Klystron: 1-2 competitors

-         Helix: 1-3 competitors

80% of total parts are hand made or assembled with high precision.

It consumes quite a lot of precious metal (copper, platinum, cupronickel (70% copper/30 nickel), molybdenum, etc.). However, for a $25-50K product, copper (for copper wire collector) only costs $2,500 or 5-10%. Nevertheless, copper collector, if faults, can’t be reused. Therefore, scrap cost could be high if yield is low (similar to semi wafer production) – a key differentiating factor btw. good operator and bad operator.

CPII says its “yield” is 90-95% for matured products; for new products, could be as low as 80%-ish initially. “Method sheet” is an effective way to control quality, which is like a direction map with precise instruction. Testing in each step of production under different environment (from extremely hot to extremely cold, for instance) also helps reduce ultimate product failure.

Internally made components account for 20% of total and the rest 80% is purchased from outside, further reducing direct exposure to commodity risk.

Lead time for most products is 100-150 days; for stacom, much shorter (therefore requiring more inventory capital: $4-5mm more for safety stock in FY06).

CPII has extra capacity, which is evident during my visit and probably necessary due to long lead time. Subject to individual products, utilization is anywhere between 50-80% (at most). No much capex required beyond maintenance, which is $5mm p.a.

The avg. tenure of employees is 16 yrs (avg. age @ 46-48 yrs old) while avg. tenure of management is 25 yrs. Workers are paid $75-85K avg. salary – believeably competitive within the industry. 3-4 new hires are employed each yr for the Helix unit I visited. Most employees are graduates with degrees in physics, mechanical engineering, EE, etc. Though being next to Google and other Silicon Valley companies, CPII believes their employees are a different group of workers compared to people working for those hi-tech companies and low employee turnover seems to demonstrate that.

Competition is based on reputation (reliability and quality) and price. CPII believes it’s “Mercedes” in the industry with great reputation. Price-wise, CPII actually charges premium price based on its high level of reliability, which is important to customers as they can’t afford any failure in using CPII’s parts and components.

Helix TWT unit is mirror image to the whole firm that has achieved great level of diversification and production flexibility. Traditionally Helix TWT had been used only for EW. Now it is applied in almost every market. This offers tremendous operation flexibility. For example, if satcom business is soft, Helix unit could move labor that worked on satcom products to other end products so that unit cost can be still competitive across all the products. This is important because most contracts are fixed-price contracts. And as sole-supplier in many cases, CPII has to disclose its raw material cost upon bidding. A sustainable low-cost position for each individual product is critical to winning business in a competitive bidding. And labor flexibility and end market diversification have enabled CPII to achieve that low cost position.

Business Division

Radar (most recent update: order strong; sales strong): $115-120mm sales

CPII provides VEDs and amplifiers for air-, ground-, and sea-based radar systems for military customers. Key platforms include AEGIS destroyers (AN/SPY-1 radar and Phalanx weapon system), Firefinder artillery locating radar, and the PAC-3 air-defense guided-missile radar. This business is expected to grow about 1-3% annually, driven by the DoD’s emphasis on combating terrorism and by aftermarket demand (including upgrades) in support of the large installed base of more than 200 systems. CPII has participated in every major AEGIS radar upgrade in the program’s history. Key radar components including high power microwave devices, control products and multifunction assemblies. Do things like fire control, ground search, weather and tracking, etc.

75% of sales is recurring (replacement or upgrade) and rest 25% is from new system sales. 2/3 of total sales are sole-sourced. It’s very stable with spares and repairs sold to a large installed base (200 different platforms, i.e. types of ships). These platforms are expected to stay around for decades (2015-2030). Even there could be new Navy vessels to replace old ones, but in reality older ones never just go away, continually giving rise to replacement demand. Every five yrs generates replacement and upgrade needs. Upgrade is favorable since it’s so expensive to have a new system. For example, for Phalanx close-in weapon system, it replace missiles with machine gun, which is an upgrade.

downside: Risk from budget pressure is minimal since all programs are already funded; timing of order receipt could cause delay for 1 or 2 quarters – but again, the diversity of 200+ programs offers stability. Any delay most likely occurs in 4Q (Sept.), which is also the end of fiscal yr for DoD, since in that quarter funds are likely run out before new FY starts.

Upside: also limited (base is too big!) with long term growth in the area of +1-3%(high single digit in '05-'06 is not sustainable)

Key products

Aegis SPY-1D + MK-99 (fire control) + Phalanx close-in weapon: $15mm (all used on on DDG)

-          assuming 2 new ships @$2-4mm btw. '06-'07 + service on 50 installed DDG base

-          post '07, worst case being no new ships but installed base will have been expanded to 56 ships

-          upside could arise with int'l orders as Int’l demand started to open up among allied nations, like South Korea and Taiwan (though CPII"s shr of content might be less than that on Aegis)

Rest (200+ different programs): $0.5mm each on average (largest being $2-3mm, such as "Hawk" missile and TPQ-37 Firefinder (used in Iraq))

Market position: +1 (34% of shr in ’05)

Competitors: L-3, Thales, e2v, Teledyne

Major customers: US Navy, Raytheon (Patriot missile), Ericsson, Alenia

Medical (most recent update: order strong; sales strong): $60+mm sales

CPI’s microwave technologies generate the radiation in cancer therapy equipment and control elements of MRI equipment. CPII is the sole provider of high-power microwave devices for Varian Medical Systems’ cancer therapy machines (Varian is world’s leading maker of radiation oncology equipment). CPI’s medical market revenue has grown at a 20% CAGR over the past few years, and further expansion (8-9% at least) is likely given: (1) CPI’s strong relation with Varian, which continues to report solid order growth rates for cancer therapy equipment; (2) an aging installed base of x-ray equipment; and (3) outsourcing of development and production by large OEMs (e.g., GE, Siemens, Philips). Also, CPII plans to expand its digital imaging offerings to complement amplifier sales.

Sales is roughly 2/3 imaging ($40mm) & 1/3 treatment ($20mm).

CPII has a particularly strong position in the treatment area with 80% shr in the high-energy market thru Varian Medical Systems (VAR). Annual # of units sold/yr is 400-500 with avg. price (forzen@) $21,000. Price is viewed good for both VAR and CPII. So annual sales to VAR for therapy as sole-provider is around $10-12mm. The advantage over rivals (Siemems which use different products) is being cheaper and perhaps better. However, CPII can't break into rivals due to VAR relationship.

CPII also supplies to VAR for imaging (but not as a sole vendor) with sales running at $2-3mm p.a. Major OEMs are increasingly outsourcing, providing growth opportunities for CPI.

Major products

-         X-ray generator for imaging

-         Klystrons + electron guns for VAR radiation oncology linear accelerators

New customer (OEM outsourcing and int’l) and incremental penetration into existing customers with new products (such as digital imaging workstations) are two main ways to grow.   

Medical business tends to have seasonal spike in 2Q(March) when big order ($10-30mm) comes in.

Market position: # 1 (42% shr in ’05)

Competitors: e2v, Sedecal, Thales, L-3 (while VAR competes with Siemens and Phillips)

Major customers: Varian, MKS, Richardson, GMM, GE
Electronic warfare: $27-28mm sales

CPII sells amplifiers for electronic warfare (EW) systems that protect ships, aircraft, and land targets protection against radar-guided weapons by jamming or deceiving the threat. Also, CPII is developing amplifiers that would destroy IEDs (roadside bombs) and provide “Active Denial,” or non-lethal personnel deterrence. This is still a relatively small market for CPII, but as the DoD upgrades existing platforms with advanced EW technologies and expands the use of Active Denial, this market could generate growth for CPII. It's noted that the Active Denial and counter-IED markets are extremely competitive. CPII’s traditional decoy products are sold mainly to the aftermarket.

Below-$30mm revenue includes $2-3mm development dollar.

Decoy (stuff behind plane to cheat enemy) usage is low due to lack of competent enemy (Iraq war didn’t have a chance to use).

Nothing exciting before 2008/2009 when non-lethal weapons like active denial, counter-improvised explosive devices (IED) and electronic attack start to comes along ($5-10mm/yr incremental for each) - these are very expensive weapon, so taking longer to deploy

CPII has strong established, often sole-provider, position in key programs (selected products listed below)

Towed decoy

Sole supplier of mini-traveling wave tubes (“TWTs”) to ALE-50 program

Qualified supplier on IDECM

Shipboard decoy

Sole supplier of TWT on Mk-53 NULKA and European DLH programs

Sole-provider of mini-TWTs to ALQ-184 (electronic warfare jammers) for Airforce

Sole provider of mini-TWTs to SLQ-302 (multi-beam phased array system) for Navy

Market position: Tied # 1 (24% shr in ’05)

Competitors: Teledyne, L-3, e2v, Thales

Major customers: US Navy, Raytheon, BAE, ITT, Lockheed
Communication: $110mm sales

CPI’s products amplify and transmit signals for satellite (all frequencies), terrestrial broadcast (AM, FM, shortwave radio, VHF, UHF), and over-the-horizon (microwave) communications systems. In 1H06, Communications sales increased 24% y/y due to amplifier sales for direct-to-home TV systems. This, however, caused a cyclical downturn in sales to DTH customers for the next year as DTH customers are now focused on deploying these purchased amplifiers and not ordering more. Outside DTH market, robust growth should continue @high-single/low-double digit as the military invests more in Ka- and tri-band based programs.  This business is different than the other divisions in that CPII plays a role as a communications system integrator, combining the VED, power supply, and control systems – in the other divisions, CPII is a component vendor.

Overseas/domestic sales split is 50:50. Top clients include

- Echostar: $10mm

- DTV: $10mm

- China: $5-10mm

- India: $5mm

- EU: $20mm

Total unit sales is around 2,000 with an avg price of $50,000-70,000. New amplifier sales is roughly $60mm (including power source) and replacement sales is $40mm.

military sales is $15-20mm, which is growing @ +40-50% w/ new applications for field communications.

I.                    satellite: 75-80% of segment or Satcom

a.       satellite uplinks with installed base of 19,000 amplifiers

                                                               i.      It’s cyclical in a sense that direct-to-home satcom customers, which represents 15-20% of sales, periodically place large orders as they expand infrastructure but reduce demand as they deploy and digest capacity. Over time, # of channels (local channels via georgraphy expansion) and HDTV (consuming more capacity) drive demand

                                                             ii.      While severity of such decline is unknow, if we assume 50% decline on 15-20% of sales, and the rest of communications growing at 10%, total communications sales could decline by just 1% - not a big deal

b.      New products include

                                                               i.      30 gHZ or KaBand for both commercial (teleco, data, not voice, focuesed) and military

                                                             ii.      military comm. (Triband with three discrete frequency bands)

1.      GD partner

2.      switching from direct sales to reseller 

II.                 Broadcast (15-20%): AM/FM/UHF/VHF ratio and TV

a.       Used to be supplied by Eimac; with declining market, reason for consolidation

b.      Also reason for purchase of Econoco to supply second-handed products

c.       New products include shortwave digital radio. It offers incremental growth but also faces a lot of competition.

III.               Over-the-horizon (5%)

a.       Over several hundred miles

b.      No satellite required => easy use/alternative to satellite-based communicatons

With new products, gross margin could be initially low, offset by higher volume only later.

Market position: # 1 (30% shr in ’05)

Competitors: Thales, Xicom, e2v, MCL, Teledyne

Major customers: Ecostar, DirectTV, Chinese/Indian government
Industrial: $20-25mm sales

The company sells VEDs for material processing, instrumentation, and voltage generation in various industries including pipe/plastic, textiles, and semiconductor fabrication. Industrial sales roughly doubled between FY03 and FY05, but growth is expected to moderate. Large installed base of industrial equipment, textile, semiconductor equipment should provide a predictable base of business from aftermarket replacement demand. 

Major products include amplifiers (fully integrated), high power microwave devices (used in instrumentation for electronmagnetic interference and compatibility testing).

Tied # 1 with 26% mkt shr.

Scientific: $6-7mm sales

CPII manufactures equipment used in “big science” research projects such as reactor fusion programs and particle accelerators for the Department of Energy. Order rates tend to be volatile in this business, and CPII anticipates demand associated with next-generation accelerators and electron lasers.

Products include high level of microwave or RF energy for reactor fusion program and accelerator for high energy particle physics (“big science”). It’s estimated that there are 60+ high energy particle accelerators in planning, design, development, construction phase.

Sales should fall within a range from $5-12mm – it’s lumpy, though.

Scientific market also provides funded R&D that generates innovations which can be applied in other markets.

# 2 with 30% shr.