I've written about Unipixel (UNXL) before here, but wanted to give an update due to new developments, new information, and the several Seeking Alpha articles that followed mine in the past three weeks. But first, a brief summary on the company.
Unipixel is a micro-cap technology company focused on performance engineered films. It has two main products right now: UniBoss and Diamond Guard. Diamond Guard is a cover film that is extremely thin/light, highly scratch resistant, and virtually unbreakable. UniBoss is a highly customizable patterned grid used in touch sensors as an alternative to current ITO models. For more information on these products, see my previous article (linked above) or Green River Asset's article here. Both will provide excellent background and I highly suggest reading them both before continuing, as I skip over much of the detail included.
I will focus on UniBoss, since it is the more exciting of the two products, although Diamond Guard could also potentially produce significant revenue. As in my previous article, I want to focus on two aspects: The opportunity and the risk.
The touch screen market is huge and growing fast. DisplaySearch estimates that it will grow from $16 billion in 2012 to $32 billion in 2018. Note that this is for the whole touch screen, not just the touch sensor. Based on the following article from DisplayBank, the touch screen panel market size is approximately 65% of this total amount, or about $10B in 2012 and $20B in 2018. However, UniBoss is half the cost of current ITO technology, so really the total market is $5B now and growing to $10B in 2018. This is the expected market that UniBoss is going after.
Keep in mind that touch screens are not a fad. They are here to stay and growing fast. Microsoft is fully devoted with its release of Windows 8. Also, many other companies announced their full support for touch screens at CES: Intel, Lenovo, and of course Dell.
Let's focus on these "large-screen" devices and look at the market on a per unit basis. Morgan Stanley estimates that in 2013, there will be near 600 million large-screen devices (desktops, notebooks and tablets) shipped. If we assume that just half of the notebooks/desktops are touch enabled (less than the 60%-100% claimed by Lenovo, Dell and others), then we can expect approximately 216 + 371*.5 = 400 million large-screen touch-enabled devices shipped in 2013 and 352 + 370*.5 = 550 million in 2015. At $20 per square foot, this translates to a little under an $8B market in 2012 and $11B in 2015 - a little above the estimates presented earlier.
The differences come from the fact that DisplayBank and DisplaySearch are probably not accounting for the large transition in notebooks/PCs to touch screens as this is "new" information from CES this year. It was previously assumed that touch screens would not become too substantial in these devices as the cost was too high: Adding a touch screen to a laptop currently costs about $200-$400 to the end-user ($40-$90 for the ITO touch screen multiplied a few times for various margins). As an example, Sony just revamped its VIZIO T13 line to include touch screens. The base model increased from $770 to approximately $1,100. UniBoss will only increase the total cost by about $50-$150 to the end-user (an estimate given similar ratios).
I have already discussed why UniBoss is superior to ITO and other ITO replacements in the previous articles at the top, and believe that it is therefore not unreasonable to assume a conservative 5% market share. Using the $8B market in 2012 (based on units), this amounts to $400 million in revenue, and does not include any potential revenue from smartphones or other touch devices.
What does $400M in revenue mean in EPS? Well, Unipixel states that its gross margins will be "well north of 50%." We could use this number, however what if margins are much higher? Unipixel has also stated that the most expensive part of the conductive film is the film itself, costing $0.10 per square foot. Throw in labor costs and yield and machine depreciation, and it is clear that, at a sale price of $20 per square foot, the revenue will in fact be "well north" of 50% and could easily be in the 80% to 90% range.
Now, all of this is assuming a 5% market share, or $400M in revenue. But how likely is this? Well, at $20/square foot, Unipixel would have to sell 20M units, or about 1.66M units per month for an entire year. According to Unipixel, it expects to be ramped up to over 3M units by the end of 2013, with its first major customer -- rumored to be Dell (DELL) -- accounting for near 1.3M units per month by the end of 2013, and potentially more in 2014. Unipixel doesn't even have to sign any more deals this year and it will still produce upward of $260 Million in revenue, or about $6.5 in EPS using 50% margins. Give that a 20x P/E and we're looking at a $130 share price, just from the current customer, which is already signed. But how committed is this customer? Well, it will be writing four checks to Unipixel throughout 2013 for a total of $15M just to help it build out capacity. That's right, the customer is putting up millions of its own dollars just to help Unipixel build the capacity to provide it with touch screens.
To further support this data, one can look at the Craig-Hallum research note that was published yesterday (Jan. 22) when it increased its Price Target from $22 to $24. These numbers seem low compared with what I am postulating, however, it readily admits this:
As for 2014, which we are using in our price target calculation, we are only assuming that they produce at 200K units/month for the entire year. We obviously think these numbers are quite low, especially considering that they are ramping capacity to over 6x our estimates by the end of 2013. When looking at a scenario analysis, which we published in our December 10th note, it is clear that the company has the ability to generate more than $9.00/share in earnings next year with just the existing partnership.
Below is a chart to calculate potential EPS given various margin assumptions on the left and unit sales per month at the top. I have assumed the following: OpEx of $12M per year for each 1M units produced per month. Taxes of 35%. Fully diluted share count of 14M (it is actually around 13.5M).
|Units per month in Millions|
I personally believe that a P/E of 25 or 30 is fair given the growth, but even at a conservative P/E of 15 the price targets get very large, very quickly. Keep in mind that Unipixel recently raised its capacity ramp to over 3M units per month by the end of 2013. At just 50% margins, this will allow for over $15 EPS.
And yet, the price is currently at $15 per share. So clearly there must be significant risks, right?
Dilution: As with any small company, one of the risks is dilution. However, that seems extremely unlikely with Unipixel. The company has $15M in cash, zero debt, it will be receiving an additional $15M throughout 2013 from the PC Manufacturer, and it will soon be cash-flow positive. Management has repeatedly stated that it will not be needing to do any more secondaries, thanks in part to the Preferred Pricing Agreements (detailed in Green River Asset's article above).
Manufacturing: Unipixel does not have mass-production up and running yet, and until it does, this will be a risk. It currently owns one printing line (able to produce 1M units per month) and has already ordered two more. It also owns one plating line, which it is currently building / testing. Each plating line can output 100k units per month -- it has already ordered two more for April and parts for eight more for the second half of the year. Based on the orders, it seems it is very confident that mass-production will work and it will be able to fill this capacity (probably because of the current customer's needs and potential future customers). Up until now, it has been producing UniBoss at a 3rd-party facility using the exact same equipment. The only difference is that it didn't utilize the roll-to-roll technology, but it is very confident that this will not cause any issues. Apparently, so is the PC manufacturer.
Other: There have been two SA articles written recently, which were very negative toward Unipixel. They can be read here and here. Both articles have been brilliantly refuted by Green River Asset here and here. But I will touch on them briefly. Both articles focus on the following points: The market isn't really that big, Unipixel's tech is unproven, and there are strong competitors. I feel as though I have painted a very clear picture as to why the market is actually very large. I have done this through estimated touch sensor revenue projections as well as looking at actual units of devices being sold. I even took the liberty of using many of the same articles that the second negative article (by Benjamin Wiley) linked to.
As for Unipixel's tech being unproven, this is impossible for me to discuss. I have no way of proving to myself or anybody else that the technology is viable. But on the other hand, these authors have no way of proving that it is NOT viable. In the end, we can do as much due diligence as we want and speak to as many engineers as we would like, but the only way to know for sure is to view demonstrations. Here is a great demonstration that not only shows the great ability of UniBoss, but also its transparency. Also, keep in mind that the large PC Manufacturer would not have signed up and asked for limited exclusivity and put forward $15M for production if it was not impressed. Also, numerous institutions including FMR, Bank of America, and Wellington, all own over 5% of the shares outstanding and clearly have faith in Unipixel's technology and ability. Most importantly, however, the company believes in itself. Employee compensation is heavily weighted in stock options and just last week on January 15, the company issued stock grants to itself at a price of $16.35 vesting in August of this year. It clearly believes that it we will be much higher by then.
Finally, other competition. The main competition is Atmel's xSense (also copper metal-mesh), however it is not nearly as inexpensive as UniBoss. From talks at CES, it is clear that Atmel (ATML) is advertising its product as an improvement to ITO, and not as a competitor to UniBoss. I have even heard chatter that Atmel is in talks with Unipixel about developing a new controller for the company. Other competitors would be Cambrios and other silver metal-mesh companies. While these are certainly competitive, they are also not nearly as inexpensive as UniBoss and they do not provide any significant advantages (again, information is from talks with people at CES).
I'm not trying to suggest that Unipixel will get 100% of the market share. Even if it miraculously had capacity, this would never happen. However, I am trying to suggest that it is not at all unreasonable to expect a 5%-10% market share, and maybe significantly higher. In all, it is clear to me that the potential for Unipixel this year is huge and the risks are, in actuality, quite small. I expect more customers to sign up in the coming months, especially once they get manufacturing up and running and prove that it is feasible.
Unipixel will be having its earnings announcement for Q4 around March 1 (still unconfirmed), and I'm sure that by then it will be able to discuss much more of this in detail. Until then, I expect potential catalysts to be any press releases of new partnerships / customers or PRs relating to the first plating line being functional.