Here is the text, but I suggest you read it on SA so you can view the links as well:

I think you have done interesting work as a researcher, Ben, but I'm not going to put a lot of faith in your abilities as an investment analyst for several reasons.

In your recent article, you question UniBoss's potential market size and economic viability. You claim to have "gathered and analyzed data from independent, publicly available market reports in order to build a clear, quantitative picture of the potential market for patterned metal grids such as UniBoss." While you did gather facts from market reports, it is your interpretation and analysis of these facts which lack credibility. It seems that your financial involvement with two UniPixel competitors (Cambrios and NanoForge), may have clouded your judgment.

First of all, I'm afraid your numbers don't make a lot of sense. You start off by telling us that were UniPixel to record sales of $240mn in 2013, they would record earnings of $120mn. No, Ben, that would be the company's gross profit. You would still have to take out all of the operating and non-operating costs and you would have to tax the numbers (if you don't give the company credit for $50mn in NOLs). But your math gets screwier still. $120mn in theoretical profits would not equal $2 in earnings, it would equal just over $9/share in earnings assuming 13.2mn fully diluted shares outstanding. Your numbers weren't even in the ball-park.

Note that the theoretical scenario you laid out for 2013 is highly optimistic vis. a vis. Street expectations (as represented by sell-side research). The model I am looking at anticipates a little more than $28mn in revenues for 2013 which equates to a bit over $1 in earnings. Understand, also, that $15mn of that $28mn is actually engineering services revenue, so product revenue is only projected at $13mn in 2013. Despite this, UniPixel is projected to generate over $1 in earnings.

You proceed to draw a distinction between UniBoss's mesh grid and the sensor which includes lamination and tab bonding. Fair enough, but you don't have the actual numbers to understand this distinction. This creates a huge flaw in your analysis.

UniBoss isn't a finished touch sensor, but it's a lot closer than a sheet of ITO film. The Uniboss film already has the bezel circuitry printed, it just needs to be die-cut, tab bonded, and laminated, and it is a sensor (it may actually be printed in a manner to obviate the need for a flex cable, but we don't need to go there now). This is a very low cost, low value-add process that can be done in a clean, ambient production environment - not in a heated and cooled vacuum chamber. The completed UniBoss sensor costs approximately $25 which assumes $3 for lamination and $2 for bonding.

Your patterned ITO film, on the other hand, still needs quite a bit of work to become a sensor. First of all you need two separate sheets of patterned ITO, so you're up to $18 ft2 right there. The bezel circuitry needs to be silk screened which costs $5. Lamination of the two ITO sheets adds $3 and yield issues add another $3. Then you need to laminate to the cover and to the display for approximately $6 and add at least another $3 for yield. Now your sensor is around $40.

Note that the patent that you reference in your note actually undermines your entire thesis. I'd encourage all readers to peruse the abstract to see exactly what I mean. This process is complex. It involves several distinct substrates, and three separate deposition and etch processes using ITO. UniBoss, on the other hand, uses a single substrate, printed on both sides in a roll-to-roll process. No vacuum chamber, no sputtering, no photolithography, no etching, and no batch processes.

In short, comparing a $9/ ft2 ITO film with a $20/ ft2 UniBoss film is a massive error. They just aren't apples-to-apples. As recently as three weeks ago, I spoke with one of the largest touch controller manufacturers in the world; he told me that an ITO sensor for a display in the 12" to 14" range would run $45 to $50. Understand that his motivation is to produce touch modules as cheaply as possible, so I can't imagine he would artificially inflate the cost of ITO sensors.

You have made two errors in trying to ascertain the market size for UniBoss. First, somewhat inexplicably you cite data segmented for small screen sizes, then you indicate that UniBoss's market is limited to smartbooks. Neither of these assumptions is accurate.

Your market size numbers are quite misleading. You get them from a paper which only considered mobile devices under 10.1 inches. This essentially ignores the impact of Windows 8 - the key catalyst for UniPixel. Windows 8 opens up large format screens such as desktop PCs, laptops, and ultrabooks/smartbooks, as well as potentially televisions. Your $8bn market is not small, but UniPixel is after some bigger fish with larger displays.

You then confine UniBoss's opportunity to the smartbook market, and go on to underestimate the size of that market. According to the controller manufacturers with whom I've spoken, UniBoss is highly competitive from twelve inches and up. It is competitive in smaller displays, but it doesn't provide as large a cost savings in smaller form factors. Thus UniBoss is able to address most of the display market (not just smartbooks), but it will be a "no-brainer" sale in the notebook, ultrabook, desktop, monitor, and television markets.

Notebook shipments in 2013 are expected to register at just over 200mn. Mike Malouf's initiation report on UNXL noted that Intel (INTC) expects that ultrabooks (I'm assuming smartbooks and ultrabooks are the same) will comprise 40% of the notebook market. This would indicate around 80mn smartbooks shipped in 2013. If those screens approximate 1 sq ft., then the opportunity in this part of the notebook segment would approximate $1.6bn for UniBoss. Of course they're not going to sell $1.6bn worth of UniBoss into this market this year, but the point is that your market size estimate of 24mn sq ft. is woefully off.

UniBoss is focused on the 350mn unit global PC market which includes desktops and laptops. They will also be very competitive in tablets which are expected to number around 125mn units in 2012. As has been noted in many presentations and comments, Display Search gives a much more comprehensive and accurate view of the touch sensor market. The research firm estimates that this year it was just under $15bn and projects it to be a $17bn market by 2015.

Luckily for all of us, we don't have to rely on you, me, Nanomarkets, Cambrios or NanoForge to tell us how large UniBoss's opportunity is. Dell is telling us with hard, cold cash and volume commitments. Dell will take whatever UniBoss can be produced, and Dell sells over 20mn laptops a year. In the scenario you outline at the beginning of your note, where Unipixel only supplied them with enough UniBoss for 30% of their production (6mn units), the company would produce approximately $240mn in revenue and $3.40 in earnings.

How much sense would it make for Dell to give UniPixel $15mn for capacity to produce a film that sold at a 100%+ price premium to ITO (assuming your claim that ITO is ready to go at $9 ft2), and then put this relatively expensive film into several laptop lines? Not much sense at all, right? Perhaps what you're trying to imply (like many shorts on SA) is that there is no deal with Dell? If so, I'm going to have to put you into this bucket, Ben.

This deal is real, and there's a lot of evidence to support it. Reed Killion noted on a conference call following the deal's announcement that the PC OEM had done six months worth of due diligence on UniBoss, including extensive work integrating this OEM's supply chain. All of this work has led to some industry chatter which is why we know that the PC OEM is Dell. For instance, we have spoken with the controller manufacturer who is working with Dell to supply chips for the touch sensors. This manufacturer is, of course, working closely with UniPixel.

We also know that UniPixel is to be paid $15mn in engineering services fees in the deal. The first payment of $5mn is scheduled to be made in the first quarter according to analyst models. It certainly wouldn't make a lot of sense for UNXL to lay out such a front-loaded milestone schedule if there weren't a real deal.

Nor would it make sense for people like Bernie Marren, Carl Yankowski, and Bruce Berkoff to simply go along with the massive fraud that many people imply. Look where these guys came from - President & CEO Western Micro, Founding President Semiconductor Industry Association (SIA), CEO of Palm Inc., President and COO Sony Electronics, CMO Applied Materials, Chairman LCD TV Association. Are we to believe they were all asleep when the press release went out? Do you think they would risk all that they have accomplished over their distinguished careers to make a few bucks on UniPixel?

You've done interesting work in the space, Ben, yet your analysis of UniPixel reeks of sour grapes.