Over the holidays I spent some time at my father's home in Vermont. Like most seventy-six year olds, my dad is not a big fan of change. He recently broke down, however, and replaced his twelve-year-old computer. To my surprise, he chose a 27 inch, all-in-one, Dell (DELL) touchscreen desktop. Dad doesn't normally spend a lot on "extras," so paying an additional $400-500 to get a touchscreen monitor was a bit out of character. I asked what compelled him to pay up for this frill; he replied, "I like it." Dad's not all that expressive.
The fact that there is demand for touch functionality in desktop and laptop computers isn't a very controversial notion. It just makes sense that in some applications we would prefer to navigate and manipulate by pointing and using gestures on a screen. Keyboards and mice aren't going away, but touch is a value added user interface for which even crusty old guys are willing to pay.
Touchscreen technologies have evolved considerably over the last thirty years. Each evolution has made the technology either more user-friendly or less expensive (or both), encouraging greater adoption. Apple's (AAPL) introduction of the iPhone, with its projected capacitive (pro-cap) touchscreens, to the mainstream market in 2007 provided a tipping point for pro-cap technology, and has since driven outstanding growth in the touchscreen market. DisplaySearch estimates that the market for touch modules is now $16bn, and it projects that market will double over the next five or six years.
This estimate is likely to prove conservative. The touch market is on the cusp of another inflection which will encourage an almost ubiquitous adoption of touch screens across existing computing and television displays. It will likely cause touch to become a standard interface in a wide variety of existing electronic applications -- and perhaps give rise to some new ones as well.
Incumbent Technology Obstacle - (ITO)
The predominant form of projected capacitive touch that we've come to love in our Apple iPads and iPhones relies upon a conductive material called indium tin oxide (ITO). ITO is well-suited for touch sensing as it can be applied in a thin coating that is largely transparent. This transparency is important when creating touch sensors because it allows as much light as possible from the display to pass through the sensor.
In a pro-cap touch sensor, ITO is arrayed on sheets of glass or film in a grid pattern of rows and columns. The sensor layers fit below the glass surface that you touch but above the LCD display (or inside the display in some cases). When a voltage is applied to the grid, an electric field is created; each connection on the grid can then be measured for its capacitance - how much charge it holds. When a finger or other conductive element comes near the ITO grid, that object's capacitance alters the capacitance at a number of grid connections. This information is analyzed by the micro-controller attached to the grid to determine the precise touch location and apply this information to the corresponding application.
ITO pro-cap touch sensors have obviously worked well for Apple. Since the introduction of the iPhone, pro-cap has become the dominant touchscreen technology in the market (replacing resistive as the market leader). Pro-cap sensors are relatively durable, and they create a good user experience. Yet there are several problems associated with the technology that have limited it from becoming as ubiquitous as it might otherwise.
First, the Indium in ITO is very expensive. It currently sells at about $550/kg, but has sold for as much as $1,000/kg within the last several years. Since Indium comprises about 75 percent of ITO by weight, it makes up a significant portion of the touch sensor cost.
Indium demand continues to grow while supply remains restricted, so Indium prices are likely to remain volatile and expensive for an extended period. Indium is a fairly rare metal, and much of the world's supply comes from China which has imposed export quotas on rare earth metals. Despite expensive prices, demand for the metal has been rather inelastic as it is used in a number of fast growing products such as LCD TVs, OLED displays, touchscreens, and solar panels. Reports that we're going to run out of Indium in the next few years may be a bit hyped, nonetheless the market dynamics have to be a source of discomfort for the supply chain.
It's not just the raw material costs that limit ITO as a touch solution. ITO sensors are manufactured in a costly process that is similar to that used to manufacture semiconductors. This batch process generally involves sputtering (vapor deposition) ITO material in a highly controlled vacuum chamber, and using photo-lithography and etching to pattern circuits in multiple steps. The process requires very expensive, sophisticated equipment, yields can be problematic, and since the process is subtractive (meaning layers of material are deposited and then stripped, leaving only trace amounts), sizable quantities of raw material are wasted. More than 70 percent of the ITO is typically lost in the process as it is deposited on chamber walls or etched from the substrate.
ITO can be difficult to work with - especially at larger display sizes. It is a fairly resistive conductor (electric current doesn't flow easily) and it becomes brittle in the form often used. High resistivity means that the longer circuits required in larger screens translate to weaker sensing and slower responsiveness. There are work-arounds, but they are expensive. And, ITO's brittle nature often necessitates a rigid substrate that can add weight, thickness, and cost to sensors. Again, these issues are exacerbated in larger screen sizes because brittleness issues are more pronounced when longer circuits are run, and deposition is more difficult, causing yields to decline.
The race to replace ITO touch sensors with a less expensive, more workable solution has been run for years, but it has just reached a meaningful inflection point due to the confluence of two factors. First was the introduction of Windows 8 from Microsoft (MSFT) in the fall of 2012. Windows 8 was designed specifically for displays that incorporate a touch interface. The significance of this is that a massive segment of the market was just opened to touch technology. Supplying touch sensors to the larger displays associated with desktop and laptop computers, however, is where ITO falls down and becomes almost prohibitively expensive. Thus ITO's shortcomings are likely slowing consumer adoption of the new operating system.
This is a temporary phenomenon, however. That's because the emergence of the second factor; a couple of inexpensive non-ITO technologies which are finally set to begin commercial production. When Windows 8 touch demand meets lower cost touch technologies, we get the perfect environment for an enormous surge in touch adoption.
Technologies in the Hunt
Let's step back and try to put the pro-cap touch sensor technology market into perspective. There are a wide variety of companies developing technologies trying to produce a better, less expensive touchscreen - in all likelihood they number forty to fifty. Rather than name them all, we can classify them broadly into the following categories based on materials and/or process: new and improved ITO; other transparent conductive oxides; nano materials; conductive polymers; conductive inkjet; optical sensors; and metal mesh.
Each of these categories has its advantages and its drawbacks. Many of the ITO advances are aimed at improving the manufacturing process, or sensor design & placement, to make their manufacture less costly, and the sensor less bulky. You may have heard of some of the technologies currently in vogue; in-cell touch and one-glass-sensor being the two I hear most about. Though these changes will help, the processes remain fairly cumbersome and expensive, and, of course, they won't change the cost of Indium.
Other transparent oxides provide conductors which are much less expensive than indium (zinc oxide or tin oxide), but are often more brittle and/or less conductive and transparent. Nano materials in the form of silver nanowire, silver nanoparticles, carbon nanotubes, and graphene, hold interesting promise. At least one company that uses nano materials has a viable commercial product with multiple wins behind it. Yet some of these technologies are plagued with resistivity issues, and all seem to face cost hurdles. Manufacturing sensors with these materials still involves a subtractive, batch process.
So far, conductive polymers haven't proven all that conductive. The films require a development process that's almost as difficult as the one it aims to replace, making this technology an unlikely winner in the near-term. Laying down molten conductors in an inkjet-like printing process, however, could be a very efficient alternative to ITO. Yet inkjet processes are notoriously slow and have yet to prove themselves viable at a commercial scale. In addition, inkjet printing processes cannot lay conductors in lines fine enough to avoid expensive subtractive processes.
Optical sensors are relatively inexpensive and don't degrade the display image at all, but at this point they require a raised bezel, and cannot sense more than four touch points (which means they cannot become Windows 8 certified). Optical sensors are currently in use, having made strong inroads in the e-reader market and, more recently, in printers.
Finally, metal mesh holds interesting potential for its low cost, flexibility, and relatively simple manufacturing process. The salient obstacle, however, has been low transmissivity - display light blocked by the mesh. Well, at least that was the issue until recently. We'll come back to this later.
Predicting a winner from this jumble of technologies is a fool's game. Engineers who make their living evaluating the various touch sensor solutions don't know who will win, so how can you and I hope to know the answer? The fact of the matter is, there will not be a single winner. Touch sensors are used in a massive array of end markets, each of those markets having its own set of requirements and sensitivities. Some demand very high-end solutions; for others cost is the most important consideration. Screen sizes vary - so do replacement and design cycles. Thus it's likely that a few of these technologies will find their way into the market today in various niches, depending on their strengths and weaknesses.
This is hardly a bold prediction. It's more of an assertion that while the players and the technologies may change, the overall landscape will remain a patchwork quilt of touch technologies. Today's touch market is fragmented; a number of different technologies operate in various markets. Resistive touch screens (which require pressure to sense touch) are still popular in industrial applications, in various terminal systems (such as when you sign for a UPS package), and in many consumer appliances. Within ITO solutions, there are endless variations of the technology that offer tradeoffs between expense, form, and function. In-cell ITO touch sensors are becoming popular in small form factor, high-end consumer devices which benefit from the slimmer profile and lighter feel. Optical solutions have made big inroads in e-readers, and are now beginning to push into other resistive strongholds such as printers and low-end handsets. Three years from now, the touch sensor marketplace will likely still be a hodgepodge of technologies.
Uni-Pixel and UniBoss
Though there is unlikely to be a single dominant player or technology that conquers the pro-cap ITO touch market, one company that seems poised to gain a substantial place in this market is Uni-Pixel, Inc. (UNXL).
Uni-Pixel, an emerging technology firm based in The Woodlands, Texas, has a rich history developing what they call "performance engineered films." In creating specialized films for LCD TVs, electronic circuits, and 3D displays, Uni-Pixel developed certain core competencies around engineering micro-optic designs for high volume, high fidelity, roll-to-roll printing processes on thin films. Out of this experience, the company came to develop UniBoss, an embossed thin film containing a trace copper mesh structure that produces a high performance, low cost touch sensor.
As noted earlier, a major factor prohibiting the adoption of metal mesh technologies has been the lack of transparency in the conductor. Metal mesh conductors block light coming from the display. Were it not for this concern, metal mesh would seem an ideal candidate to replace ITO. The mesh itself is flexible, durable, inexpensive, and up to 2000x more conductive than ITO. More conductivity means lower power requirements, less latency, and the ability to easily scale over large display sizes.
Uni-Pixel has engineered around the transparency problem, and in so doing may have found a "holy grail" in pro-cap touch sensors. Embedded with copper circuitry having a remarkably thin five micron line width, UniBoss's copper mesh is virtually transparent. Uni-Pixel backs this claim with optical tests that demonstrate transmissivity readings in UniBoss film (as high as 92 percent ) that equal those of ITO solutions and that retain extremely low resistivity.
Unseating a part of the massive ITO supply chain requires a competing solution to offer both a compelling cost proposition and functionality that is as good or better than existing ITO solutions. UniBoss appears to check each of these boxes. The film is, without a doubt, much lower cost to produce than an ITO touch sensor. The relative simplicity of the production process makes this clear. That process is comprised of a fairly straight-forward roll-to-roll patterning, followed by an in-line roll-to-roll plating or metallization step (though this step will ultimately be outsourced). That plating step is where, through certain patented (or patent pending) chemistries, copper is deposited in thin traces along the raised design. The finished film is then ready for standard die cutting, bonding to the flex cable that leads to the controller, and lamination to the device or lens. Notably each of these processes can run at normal room temperature and do not require a vacuum chamber. Reed Killion, Uni-Pixel's CEO, is fond of saying that UniBoss reduces the steps required to create a touch panel from forty-one to eight.
While the scalability of a roll-to-roll solution certainly imparts some cost advantages, so does the process of adding patterns on film rather than using a subtractive, photolithographic process. As mentioned earlier, ITO circuitry is created in a complex and costly subtractive process where layers of conductive material are laid down and then etched away. The vast majority of ITO alternatives also use a subtractive process. This is not only expensive but is difficult to engineer and is prone to errors. UniBoss deposits conductive material only where it is required, saving material, and resulting in better yields - all of which translate to lower cost.
Last, the materials costs for UniBoss are significantly less expensive than most other ITO and ITO replacement technologies. Copper is currently 70x less expensive than Indium, and UniBoss's most significant material cost is inexpensive PET film - priced at a nominal twenty cents per square foot (and likely to drop quite a bit as volumes increase).
UniBoss has some noteworthy performance advantages over ITO. Copper's superior conductivity means UniBoss requires much less power and can sense touch without the latency and noise issues which sometimes plague ITO (especially in larger displays). These are significant advantages, especially latency, as anyone who has struggled with slow touch response in larger screen sizes will attest. Copper is also an easy-to-work-with, flexible metal which doesn't require a rigid substrate. This makes UniBoss an ideal sensor for flexible displays as well as traditional rigid screens. It also means that UniBoss can easily scale to large format screen sizes while often eliminating a layer of glass.
How much less costly is UniBoss? A UniBoss sensor for a standard-sized 10.5 square inch smart phone might cost $2.10. This assumes a $25/square ft. UniBoss price that equates to about 18 cents per square inch (I've added 10 percent for material not used after die cutting.) The same ITO sensor is said to cost about $4.05 (MDB Report) or 39 cents per square inch. Thus, UniBoss might save an OEM 40 percent before taking the remaining components and module assembly into account.
That's certainly a noteworthy savings, but the comparison becomes even more appealing at larger screen sizes. An ITO sensor for a 15 inch laptop (diagonal) screen costs about $46 with the cost per square inch rising to 46 cents (a 15 inch screen is 103 square inches). Since UniBoss production costs do not increase with larger display sizes, the UniBoss sensor would be about $19 (18 cents x 103 inches), for a savings of approximately 60 percent. Remember too that the dollar value of this differential will be magnified for the consumer after taking into account the margin structure of the OEM, distributor, and retailer.
How do we put this UniBoss information into perspective? It sounds promising, but how do we know this isn't just corporate propaganda? And if the sensors are that good, how big can Uni-Pixel become?
Some of the people best able to answer these questions are engineers at touch controller manufacturers. The controller manufacturers often coordinate the touch module supply chain - assembling modules, designing electrode layout, etc.. The value they add means the controller companies regularly function as the fulcrum of the touch supply chain. Bringing a new product to market usually means convincing these companies that the new solution offers real advantages over the alternatives. Thus the engineers at these firms occupy an unparalleled vantage point to assess the landscape of touch technology.
I spoke with two engineers who work within very large controller businesses. Despite the fact that they each had business reasons to feel somewhat threatened by UniBoss, they both readily admitted that the UniBoss technology was solid, and that Uni-Pixel would likely win business at larger screen sizes. Analysts with whom I've spoken have also interviewed controller contacts and come away with even stronger positive feedback. One analyst said these engineers left him with the distinct impression that UniBoss was "far and away" the leader to replace ITO, and that, "nothing out there compares with it for commercial potential."
I've also gained perspective through conversations with three companies producing or developing non-ITO touch sensors. Despite the fact that all three have high-profile, well-regarded technologies, and could eventually see themselves in competition with UniBoss, they each acknowledged the strength of Uni-Pixel's technology and the early lead UniBoss has achieved in the marketplace. The managers of these enterprises each expressed the sentiment that the opportunity to displace the dominant incumbent technologies was so attractive that it didn't make a lot of sense to be overly concerned with the various competing non-ITO technologies.
Perhaps the most important validation behind UniBoss, however, comes from the company's several high-profile commercial partnerships. The first is with Texas Instruments (TXN) which announced a deal with Uni-Pixel early in 2012 to integrate TI's controllers with UniBoss's touch sensors. It's important to note that while TI and other controller manufacturers attempt to diversify their exposure and work with a portfolio of new touch technologies, Uni-Pixel's partnership with TI seems to run deeper than most of these arrangements. Uni-Pixel was able to secure the agreement with TI in a non-exclusive arrangement while TI preferred a closer, exclusive deal. The companies work closely together, and TI sees UniBoss as very synergistic to their objective of making inroads in the pro-cap market.
Uni-Pixel's technology is further validated by its agreement with N-Trig, a controller and module supplier that sells the DuoSense pen and multi-touch pro-cap sensor system. Additionally, Uni-Pixel is working with at least three other controller companies as disclosed in its second quarter conference call. While other touch controller manufacturers would likely love the opportunity to engage with Uni-Pixel, it makes more sense for the company to fully develop a few large volume opportunities rather than worry about many smaller ones.
Last, in December Uni-Pixel signed a preferred price and capacity license agreement with a large computer manufacturer - rumored to be Dell . The structure of the agreement will be discussed later, but suffice it to say the deal created a source of demand so large that the company may not be able to fully meet it for years. Owning a great technology is one thing, but it is an entirely different proposition to have the validation that comes with a very large order. In this case, one of the largest computer manufacturers in the world has effectively said, "here's some money to buy equipment; I will take whatever you can produce for the next couple of years."
I also think it's interesting to note that the CEO of Uni-Pixel went out of his way to emphasize that management fought hard to avoid diluting shareholders in the deal with this PC OEM. This is almost as important as anything else in the deal. The large PC manufacturer, who must understand the value proposition and competitive positioning of UniBoss better than just about anyone in the world, was very keen to acquire an equity stake in micro cap Uni-Pixel.
Sizing the Opportunity
Though there will be a number of winners in the race to replace ITO, UniBoss seems particularly well-positioned to win a good-sized share of the purse. How big will that share be? Mike Malouf at Craig Hallum noted in his UNXL initiation report that the market opportunity for UniBoss in ultrabooks alone would be roughly $2.8bn in 2015. He then went on to say that he sees Uni-Pixel in the "pole position" in this market. Of course there may be just as much opportunity for the company in large displays such as tablets, desktops and televisions. Further, UniBoss is likely to also find use in the smartphone/pda segment, and, given its flexibility and scalability, in completely novel applications - think interactive touch media displays, touch walls, etc.. It's also entirely possible that because of UniBoss's low price, the technology may find a place in markets currently served by resistive sensors.
Sensors are projected to be a $17bn market by 2015 (vs. the touch module market of $22bn). If Uni-Pixel were to claim a fairly modest two percent of that market in that time frame they would generate revenue of $340mn. With further market growth and penetration of about five percent, it doesn't seem outlandish to speculate Uni-Pixel might be a billion dollar business before too long.
If the opportunity in front of Uni-Pixel is impressive, the company's business model is almost equally so. The management team has clearly invested time and effort into understanding how to best turn touch sensor revenue into cashflow and earnings for shareholders.
This began with the practice of being parsimonious with the company's equity -- avoiding large, dilutive offerings, dilutive acquisitions, and large options grants. Further, as noted above, management has fought hard to keep from giving up equity to partners and potential partners. The result is that UNXL has only 9.66mn shares outstanding, plus 3.5mn in options and warrants.
Uni-Pixel has priced UniBoss correctly. They are offering savings of 40 percent-plus to the supply chain/OEM while maintaining gross margins in the 50 percent range. Further, I think there's a good deal of flexibility in that gross margin number which will allow the company to either lower prices or take more margin, or both. As stated earlier, the single largest materials component of UniBoss is the thin PET film, and as UniBoss scales, those costs may decline substantially.
Management has also shown financial savvy while at the same time expressing its market strength in fashioning a capital-light manufacturing business vis. a vis. its recent large PC OEM deal. Transitioning from an R&D house to a commercial manufacturing operation would normally be a period of uncertain profitability and large cash draw-downs. Yet in its recent "preferred price and capacity license agreement" with the PC OEM, Uni-Pixel has structured a deal in which it is paid a $15mn royalty-like engineering services fee to buy and set-up the equipment, and begin manufacturing for that OEM. In return, the OEM gets guaranteed capacity and a reduced price on UniBoss product.
Uni-Pixel avoids the big capex hit while building out massive capacity, and receives a highly predictable (milestone based) royalty-like $15mn in revenue. That $15mn equates to over $1 in earnings for UNXL - more than offsetting my projected operating expenses of about $12mn for the year (assuming they recognize it all in 2013). The deal allows analysts to forecast about $1 in earnings for UNXL in 2013 - in what would otherwise be a money-losing year. There aren't a lot of micro-cap companies that could negotiate a deal like this with a behemoth global PC manufacturer.
About That Deal….
The PC OEM deal promises to help Uni-Pixel ramp its manufacturing capacity quickly to meet its partner's volume commitments. The company has set some capacity goals that they hope to achieve over the next couple of years using this partner-funded model. I've laid these out in the table below and shown a corresponding back-of-the-envelope annual earnings estimate for each.
Note that the capacity figures are based on a one square foot (roughly 15.6 inch diagonal) unit. The model assumes that the firm is operating at 80 percent of capacity for the entire year and that gross margins are steady at 50 percent for the entire period. I've assumed earnings are fully taxed at 40 percent (though the company has $50mn in NOLs). Operating expenses are projected to ramp from $13-14 million annually at base and then scale lower with revenue growth, starting at 22 percent. The share count is fully diluted. The model shows negative earnings in the early stages because it doesn't take into account engineering services payments from Uni-Pixel's partner.
Those are some impressive earnings numbers, but are they realistic? At 3.9mn units per month, Uni-Pixel would be generating about $936mn in annual revenue (assuming prices hold at $20/unit). If the market for sensors is approximately $17bn in 2015, UNXL would have captured 5.5 percent in achieving this revenue milestone. That's certainly an optimistic scenario, but it doesn't seem out of the realm of possibility in the third year of production for a manufacturing process that offers the ability to scale quickly and has a partner-funded build-out. The capacity number doesn't actually look all that impressive when juxtaposed against Atmel's (ATML) plans to quickly ramp their XSense product capacity to the equivalent of sixteen million units per month (one billion 3.1 inch screens/year).
The strategic significance of this deal is noteworthy. Because of the prominence of the customer and the size of the commitment it's likely that other large manufacturers have taken note. Technology hardware businesses are notoriously cut-throat and low-margin. Gaining the cost advantage that UniBoss offers could literally make the difference between serving the large-display touch market profitably, or not. In the few weeks since the PC deal was announced I have already seen a distinct uptick in interest at one industry partner.
The Real Question…..Can They Execute?
With this kind of earnings potential, it's obvious that the market is taking a wait-and-see approach, assigning a high discount rate to whatever cashflow we might project. This is understandable because there's still a fair amount of execution risk in the story. From an execution standpoint, however, Uni-Pixel may not have as tough a road ahead as you would think. That's because in 2011 the company hired a highly qualified professional to manage the production transition. Chief Operating Officer Peter Shin comes from Samsung (000830.KS), where as senior vice president of R&D he led "breakthroughs in new display technology, nano imprinting technology, roll-to-roll process architecture, flexible display development…" UNXL's CEO noted that Mr. Shin brings expertise in "LCD panel technology and fabrication, process and architecture development, OEM/ODM product development for display and notebook systems, and major device launches."
Though it's hard to imagine Uni-Pixel finding anyone better suited to manage the commercialization of UniBoss, this doesn't mean it will be easy. The process leaves very little room for error. Achieving high resolution and uniformity at five micron line-widths on the R&D bench is one thing, but something altogether different on a large production line. Having someone like Peter Shin on board, someone who has done similar things at a much larger enterprise, should give investors a measure of confidence that this execution risk is manageable.
It goes without saying that investing in pre-revenue technology companies entails a fair amount of risk. We won't explore all of these risks here, but there are three that I want to highlight.
First, as is the case in most early-stage technology companies, management has a lot on its plate, and doesn't have massive resources. Sometimes they bite off a little more than they can chew, and timelines are delayed. I recommend adding one or two quarters to whatever timeline is given. That said, they're becoming more adept at building conservatism into forecasts, and hopefully they will surprise us positively with their timing in the future.
Second, touch sensor technology is evolving at a rapid pace. While UniBoss appears to be well-positioned today relative to other non-ITO and ITO solutions, this could change tomorrow. This said, the fact that large PC OEMs and controller manufacturers seem willing to invest in partnership with Uni-Pixel is a good sign the company has some runway available. That runway could be crucial if it allows UniBoss to leverage its impressive scale economies to build Uni-Pixel a competitive moat. Regardless, all profit and cashflow projections are speculative.
Third, in late December, Conductive Inkjet Technology Ltd. (CIT), a unit of Carclo plc, issued proceedings in the English High Court against Uni-Pixel. CIT claims that Uni-Pixel
"Made unauthorised use of confidential CIT know-how, relating to the formulation and use of catalytic inks and metallisation technology in touch screen applications, in the development of its own metal mesh process and products (which it calls "UniBoss™") and related patents it has filed (or been issued) on these processes. CIT know-how in that field was disclosed to Uni-Pixel with CIT's consent as part of a project with a development partner, but CIT only permitted Uni-Pixel to use that know-how in relation to a different field of use, unrelated to touch screen sensors. CIT believes that Uni-Pixel has breached the restrictions imposed in respect of its use of that know-how."
Uni-Pixel has issued a response in which it characterized Carclo's claims as "baseless" and "without merit."
Some background will help. CIT partnered with Atmel to help develop Atmel's XSense copper mesh touch sensor. XSense competes with UniBoss and the products are somewhat similar though there are some key differences that separate them. XSense copper circuits are printed on a specialized film using a process that requires photolithography and laser ablation, as well as a clean room environment. This means that the yields and costs for XSense are likely not as competitive as those of UniBoss.
This disparity in the technologies explains the odd feedback I received when doing due diligence on XSense this past summer. Even a contact within Atmel seemed less-than-enthusiastic about competing head-to-head with UniBoss.
Atmel's XSense began shipping this fall in a low volume Chinese handset. This triggered a payment to CIT as commercial production began. Uni-Pixel's UniBoss has yet to ship a single unit in commercial production, yet in its statement, Carclo is asking for injunctive relief against Uni-Pixel to deny it an "unfair headstart" that resulted from applying CIT's special "know-how." Huh? XSense is already shipping yet Carclo alleges that Uni-Pixel has an unfair headstart?
I'll readily admit that my analysis becomes fairly murky at this point, but we're not going to do a lot better given that we're talking about technology litigation. Carclo is upset because Uni-Pixel's technology is better. We know this to be so because young technology companies don't spend the time or effort suing other small companies when they're winning the game in the marketplace. The catalyst may have been a large PC OEM who decided to use UniBoss over XSense.
XSense isn't a big business line for Atmel yet, so there's probably not a big reliance on it for Atmel's financial model. The problem is that if XSense lost this customer to Uni-Pixel, there's a chance Atmel might have lost the controller business as well. That could be important, but I want to emphasize again that this is based on industry chatter, not confirmed fact. If it's true, it may be that Atmel's XSense business has become more of a liability than an asset.
This scenario might have Carclo quickly assembling a proceeding against Uni-Pixel to see if they can rescue their relationship with Atmel. It would help explain the mixed feedback I got about XSense from Atmel, and the odd claims from Carclo that Uni-Pixel had a head start. It might also help to explain why Carclo's claims seem to intimate a misunderstanding of the fundamental process involved in UniBoss's metallization. How can they effectively assert their IP when they don't even understand the technology Uni-Pixel uses to achieve its industry leading metallization? They don't understand the technology Uni-Pixel has created, or they'd be applying it to XSense.
Carclo's intent is to not only regain the favor of Atmel, but also to inject some fear, uncertainty, and doubt around UniBoss. Perhaps they can buy themselves a little time with a strategy like this. I'm guessing that they approach the whole affair as a fishing expedition. They don't really know what lies beneath the surface of the UniBoss technology, as they clearly can't replicate it themselves. Instead they trawl for interesting information by forcing Uni-Pixel to explain UniBoss through its law firm.
Regardless, I always assume a worst case scenario with respect to these situations. I asked a friend who is an IP consultant, has a career's worth of experience covering IP firms as an analyst, and has a law degree, what he thought. He speculated that at worst, a 2-3 percent royalty seemed appropriate. For my model, I assumed a 5 percent royalty on revenue which translates to a 14 percent haircut to 2014 earnings of $2.84 (assuming Uni-Pixel produces at only 40 percent of capacity). If the worst comes to pass, and Uni-Pixel swallows a 5 percent royalty, the firm would produce $2.44 in earnings - about 6.9x the current share price.
Catalysts & Valuation
During the last conference call, Uni-Pixel's CEO noted that the company anticipated signing several partners in deals similar to the one just notched with the large PC OEM. The PC deal covers a number of laptop lines. The other deals might fund capacity to support the supply of tablets, desktops, ultrabooks, smart phones or other consumer or industrial applications. My guess is that a deal similar in size to the one just signed will be announced in the coming months. If the deal is structured like the last one, it could result in the addition of another $0.50-1.00 in earnings this year, and an even larger ramp in capacity toward the end of 2013.
The company has signaled it will announce incremental news about its relationship with Texas Instruments in the coming months. At minimum this would mean a sales and marketing relationship between the two, whereby TXN would sell solutions that use its controllers with UniBoss sensors. For its efforts, TXN would likely get a preferred price on its UniBoss orders. Such a deal is similar enough to the PC OEM "preferred price and capacity license" that we could even see TXN forge a deal along those lines. Texas Instruments is a big player in the resistive market but has had very little presence in pro-cap. They are likely to introduce a new product to serve this market, and a partnership with Uni-Pixel could be the synergistic boost they need to rapidly build a business.
Finally, quarterly reports should be good catalysts for Uni-Pixel. Analyst estimates appear very conservative (25-30 percent of capacity) versus Uni-Pixel's capacity guidance. We will be paying close attention to the quarter-ending production run-rate to get an idea of what the baseline should be for the following quarter. And while the numbers may start slowly, once management gets the hang of it, they should be able to add capacity in a more streamlined, rapid manner.
Few shares outstanding coupled with tremendous operating leverage and growth make Uni-Pixel's earnings numbers potentially combustible. The key to valuing the company is understanding how much of the touch sensor market Uni-Pixel can capture. My base case assumes Uni-Pixel executes against the plan laid out with the large PC OEM, albeit at a slower ramp, and ultimately gains five percent share over the next three to five years. I get to the five percent figure by assuming that the company would sell close to the amount of capacity for which its partner has signed on (producing 3.5 million units per month). That five percent share could mean a valuation of around $180. Below is a table that shows market share levels with corresponding valuations.
UNXL Market Share & Valuation
|Market Share||Capacity Units/Mo||Target Valuation|
I've assumed gross margins are steady at 50 percent and sales are 80 percent of capacity. Operating expenses start in the mid 20's as a percentage of revenue and decline to 19 percent. The assumed tax rate is 40 percent and there are 13.2 million fully diluted shares outstanding. I have used an earnings multiple of 15 to come to the valuation estimate.
Oh, One More Thing….
Diamond Guard is Uni-Pixel's other product. Diamond Guard is a hard-coat plastic that could replace cover glass in all kinds of displays - from phones to televisions. Carestream Tollcoating has a license with Uni-Pixel and has begun manufacturing this product.
Diamond Guard looks like glass and it has glass-like hardness, but it's lighter, it doesn't shatter when dropped, and it's easy for manufacturers to work with. It's this last point that is especially meaningful, because glass is very difficult and expensive to manipulate. Think of all the rounded corners and holes in your iPhone cover glass. Look at that color border that surrounds the iPhone's display. These features give engineer's headaches; Diamond Guard can save OEMs time and money. I'm not doing Diamond Guard justice, but perhaps we'll tackle it another time.
The Author has obtained all information herein from sources he believes to be accurate and reliable. However, such information is presented "as is," without warranty of any kind - whether express or implied. The Author makes no representation, express or implied, as to the accuracy, timeliness, or completeness of any such information or with regard to the results obtained from its use. All expressions of opinion are subject to change without notice, and the Author does not undertake to update or supplement this report or any information contained herein.
Disclosure: I am long UNXL.