Uni-Pixel Still Unable To Mass Produce Cost Effectively
Uni-Pixel (NASDAQ:UNXL) is a company that is trying to produce an alternative type of touch-sensor film to the main-stream ITO film. Unfortunately, UNXL has a problem that is similar to that of Gevo (NASDAQ:GEVO), a company I recently reported on. The problem is that while it's not so hard to create a sample product in a lab, it is extremely difficult to create at large, commercial volumes efficiently and cost effectively. Especially the touch screen market, which has large, low-cost vendors in Asia.
Even with the recent operational update, Uni-Pixel's progress is only inches further today than it was two years ago. However, it's in the interest of management to keep investors excited with progress reports and to be as optimistic as possible, because whether Uni-Pixel succeeds or fails, management pays themselves large stock compensation packages.
Uni-Pixel's June 25 Operational Update Exposes a Pile of Problems
On June 25th, Uni-Pixel had an operational update on its production progress. The company declared pilot production status of their InTouch touch sensors, and described the progress as a "breakthrough" in its press release. In reality, what Uni-Pixel has done is still a lab-scale experiment. When analysts asked probing questions on true readiness for commercial mass production, the CEO, Jeff Hawthorne, could only offer vague responses. I did a detailed analysis of the operational update transcript. Below is a summary.
- By pilot production status, the company means that the yield is not only low, but also unpredictable. The yield varies from batch to batch. The company clearly has no clue about the root cause. This puts commercial production at real risk since the production capacity is unpredictable for the fast design cycle consumer electronics (CE) market.
To draw a parallel, it took United Microelectronics (NYSE:UMC) over two years to solve the yield problem of its 28nm process to get to production status. Even to this day, UMC's 28nm production capacity is still limited due to yield issues.
- Uni-Pixel's CEO Jeff Hawthorne's vague answer on gross margin confirms the lack of knowledge in finding the root cause and the lack of know-how to fix the problem. Even more troubling is that the company is unable to comment on the long term gross margin target. A logical interpretation is that Uni-Pixel does not know how much they can improve the currently poor yield.
- Mr. Hawthorne also said that proper equipment maintenance schedule was unknown. This impacts yield. In high volume production, equipment maintenance is very important.
- Mr. Hawthorne said the company is focusing on around the 7" tablet market. This is a rather strange choice for metal mesh sensor vendors. It's well established that the mainstream, proven ITO touch sensors can effectively address this screen size. Hence, the market is extremely competitive and crowded. Atmel (NASDAQ:ATML), a credible metal mesh sensor vendor, only sees the opportunity for metal mesh sensors in larger sized applications, such as all-in-one PCs.
- The company still cannot provide a production shipment date. Best case is end of 2014, which means 2015 product cycle. By then, the market will see further ASP erosion of touch sensors. Asian vendors (incumbents and new entrants with government subsidies) will make progress as well.
Atmel Already Has Its Metal Mesh Sensor On The Market And Is Having A Hard Time Producing And Selling It
Atmel, like UNXL, made ambitious claims about its X-Sense metal mesh touch sensor, but failed miserably in 2013. And Atmel is a very successful touch screen manufacturer with much more resources, revenues, R&D, and experienced technicians than Uni-Pixel has.
Yield is the number of good parts produced out of a batch. Atmel's yield for X-Sense is speculated to be about 40%, not very good. That means out of 100 touch sensors made by Atmel in the factory, 60% are defective.
Atmel's 2013 timeline of X-Sense disappointments:
- From the 1Q13 earnings call: "Due primarily to this delay, we expect XSense revenue to be between $15 million to $25 million for 2013."
- 2Q13 earnings call: "Our expectation right now is $10 million to $15 million for 2013."
- 3Q13 earnings call: "we were not likely to hit that ($10 million to $15 million revenues) due to a customer reducing their forecast."
Wow, Atmel couldn't even hit their reduced estimates! Atmel management doesn't even talk about X-sense much anymore in earnings calls, because it has so far been a huge failure for them.
Uni-Pixel And Atmel Have No Chance Against Asian Touch Sensor Vendors
TPK (3673.TW) has always been the leader in the touch sensor market and still is, but even TPK is hurting from the hyper competition. Its market cap shrank from US$7B in 2013 to $3B today due to shrinking margins. It invented the touch sensor in 2003.
O-film is taking the touch sensor industry by storm. It is the largest touch-module maker in China, and #3 globally. O-film was founded in 2001 in mainland China. It has government support, has access to low cost labors, and is close geographically to major panel makers. It is competing aggressively on price and is supplying to Samsung, and already supplies metal mesh sensors, like for this product. O-film is taking market share from TPK. Both O-film and TPK-Cambrios have commercial adoption of their non-ITO touch sensors.
O-film is described on page 79 of this touch technology report from Intel published on June 1, 2014 for the SID conference among metal mesh vendors. UNXL was not mentioned by Intel at the SID conference among metal mesh sensor vendors. This is because the company is not considered a competitor in the space.
Since the Asian vendors have gotten so aggressive, the touch sensor market has gotten hypercompetitive and the ASPs (average selling price) for touch sensors are dropping dramatically. In 2013, touch module ASP has seen 6x compression. Based on my interview with industry experts, 1Q13 touch module price was $7-$10/diagonal inch. In 4Q14, the price dropped to $2.20. O-film was even quoting $1.60. O-film also has the unfair advantage of receiving government grants, as shown here and here.
Besides these manufacturers' commercial success and scale, they are located much closer to panel makers, which is more important than most people realize. Metal mesh sensors are subject to pattern problems called Moire patterns which are an optical problem, very visible and annoying. ITO is transparent, so doesn't have these problems. However, this problem can be solved by carefully aligning the pattern with the display. This step is best done by panel makers. Unfortunately for Atmel and Uni-Pixel, all of the panel makers are in Asia.
From a supply chain perspective, it is critical to have the touch sensor production site nearby the panel production site to respond to the rapidly changing demand. Uni-Pixel produces sensors in the U.S. This simply won't work from a supply chain management perspective. Consumer electronics (e.g., tablets, notebooks, etc.) have short product cycle. Demand changes quickly. Therefore, panel makers prefer to work with local touch sensor vendors to quickly respond to market change. Uni-Pixel and Kodak only have a facility in the U.S. (on the East Coast, even farther from Asia compared to Atmel's fab in Colorado).
Uni-Pixel's line width is wider than O-film's solution. As shown in this report from Uni-Pixel on July, 2013, UNXL is proud of its 6um line width. On page 22 and 23, it compares it to ITO which has greater than 50um line width. However, as shown in Intel's report on page 80, O-film is at 2um line width. How can UNXL compete with that? This is important since metal mesh is not transparent (ITO is transparent). TPK-Cambrios silver nanowire is even thinner. The specs of Uni-Pixel's metal mesh are on its website. To get a sense of perspective, a human hair has a width of 17 ums (um = a millionth of a meter).
Investors should review the lesson learned from Rubicon (NASDAQ:RBCN), a US-based sapphire vendor. It got crushed by Asian competitors in a brutally competitive market. RBCN was a leader in sapphire crystal production. Its crystals have the best quality. Sapphire is used to make LEDs. After a demand spike in 2009-2010 (LED back-lit TVs), many Asian vendors jumped into the scene. The market has been filled with overcapacity, and the company was forced to sell crystals below cost (negative gross margin) for eight quarters. RBCN stock never recovered, and probably never will. Uni-Pixel's situation is even worse than Rubicon because it hasn't even gotten off the ground yet.
O-film has been around much longer than Uni-Pixel and has a much better understanding of the technical and production issues. Uni-Pixel has only been in the touch-sensor game since 2012. Whereas O-film has been in the game since 2001. It will take Uni-Pixel that many years to get where O-film is today, if ever. Reaching mass production is a different animal than hand-picking a few working samples from a lab for customers to evaluate.
Uni-Pixel, and Atmel to a lesser extent, are playing a catch up game against Asian players who are much bigger, successful, closer to panel makers, have state support, and are willing to take low or no margin.
Disclosure: The author is short UNXL. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.
Editor's Note: This article covers one or more stocks trading at less than $1 per share and/or with less than a $100 million market cap. Please be aware of the risks associated with these stocks.