Seeking Alpha
View as an RSS Feed

Brian Coleman  

View Brian Coleman's Comments BY TICKER:
Latest  |  Highest rated
  • QuickLogic: An Overvalued Stock In Rapid Fundamental Decline [View article]
    There is so much wrong here it is really hard to know where to begin. I'll just focus on a few of the more glaring deficiencies in the piece. For starters, investors are not about to learn that the display bridge revenues are short duration as the company has been candid about this for the past year. The future of the company is smart connectivity (the Polar Pro 3, which you failed to mention even once in the column) and sensor hubs (the S-family: S1, S2, and S3...). As for the "legacy revenue" the company has been categorizing this revenue as "legacy revenue" for 7 years so it is hardly a surprise that it is not a growth category. As far as being concerned that it is "going away." No, it's not.

    The Polar Pro 3 is targeted at the smart connectivity market which is $150 million annually and currently served entirely by Lattice. There is no reason that in the very near term QUIK can't have 20% of that market. In fact, if you listened to the last conference call the company is very much engaged in design opportunities with Samsung on this front. If successful QUIK could easily have one-half that market.

    On the sensor hub business you wrote: "Management has also been late to even deliver a working SH product line - likely a result of their underinvestment in the product versus much larger competitors with better product offerings, such as INVN, LSCC, XLNX, ALTR, ADNC, ATML, +others)."

    First, in October 2013 the company announced the sensor hub family with first revenue expected in 2H'14. That guidance has not changed so the idea that management is late with a working chip is wrong and is a fabrication.

    Second, INVN, XLNX, ALTR, and ADNC do not sell sensor hubs. And Lattice sells what is best viewed as a sensor data buffer chip which is incapable of running the algorithms necessary for context aware sensor fusion. In fact, here's what LSCC CEO said on their 1Q earnings call: "..we don’t play in sensor hubs as much as people think we are kind of a complementor in the sensor hub technology."

    Lumping these companies in with QUIK as competitors is really all a reader needs to do to confidently dismiss the entire piece as garbage. QUIK's competition in the sensor hub is from the legacy (there's that word again!) suppliers of microcontrollers (MCUs) as sensor hubs.

    Depending on which third party source you use the market for sensor hubs is going to grow from a few hundred million units this year - supplied almost entirely by MCU suppliers - to between 1.3 billion (IHS Research) and 2.5 billion (Semico Research) over the next 4 years. This represents billions of dollars of revenue opportunity and makes the focus on display bridge revenue variability kind of silly (to put it kindly).

    The investment case for QUIK depends on its ability to capture a share of this market. That would have required you to analyze the S-family sensor hubs versus MCUs which you didn't even attempt. Instead you assumed QUIK couldn't compete against completely irrelevant companies like XLNX and ALTR who don't sell anything into the mobile handset market.

    An article on SA where the author can't even bother to get the competitors right, ignores an entire product category (the Polar Pro 3) where first revenue is likely going to be included in 3Q guidance, and assumes QUIK will only serve Tier 3 customers - so he is also completely oblivious to the fact that that QUIK is already engaged with Samsung - is not really the basis for smart investing. I would suggest that anyone looking for a view on QUIK based on actual facts, research, and an understanding of the market, technology and competitive landscape start here:
    Jul 28, 2014. 02:07 PM | 15 Likes Like |Link to Comment
  • QuickLogic: On The Cusp Of A Sharp Revenue Ramp [View article]

    Valuation is clearly subjective, no doubt. In my mind, gross margin and growth rate should be two of the largest considerations in valuing a company on a revenue basis. My rule of thumb - simply based on my experience and observations - is to take gross margin, divide by 10 and subtract 1 to get a fair revenue multiple. So a fair revenue multiple on a 55% gross margin is 4.5x, on a 70% gross margin business 6.0x etc.

    Take a look at semi companies and you can see that this is a pretty good rule of thumb, not perfect but a good starting point. Also, while I think investing on a relative valuation is a poor approach, I would point out that INVN routinely trades at an 8x or higher EV/revenue so using that as a comp my 4-5x assumption is not a stretch.

    But like I said, there is no objectively "correct" valuation at any point in time, a good valuation is in the eye of the beholder.
    May 29, 2014. 12:25 PM | Likes Like |Link to Comment
  • Why I Shorted InvenSense Inc. [View article]
    I think the biggest risk to INVN is that the motion processor becomes integrated into the sensor hub and it goes away as a discrete part. Once a feature becomes standard (which motion processing in devices is) - that is, ubiquitous (all devices have motion processing) and commoditized (is one gyro or accelerometer that much different/better from supplier to supplier? No.) - the function can be absorbed as a feature into a higher function chip. In this case it would be absorbed into the sensor hub - an MCU or QUIK's S-family.

    The parallel to think about is the demise of the market for discrete GPS chips and discrete bluetooth chips. SiRF used to be a big supplier of GPS chips until QCOM and the other AP companies integrated GPS right into the processor. SiRF's business collapsed almost overnight and ended up getting sold to CSR for not much more than cash. I can't think of any reason that motion processing doesn't eventually (in the next year or two) get absorbed right into the sensor hub.
    May 21, 2014. 12:17 PM | Likes Like |Link to Comment
  • Compugen: A Wealth Creation Opportunity In The Making [View article]

    If gazing in the rear view mirror has been serving you well as an investment strategy that by all means continue with that approach. For those of us who don't believe past performance is always an accurate guide to future results we utilize our approach of looking forward.

    I will add this, some of very biggest winners we have had on Domino Analytics (and we research very few names actually) are what are orphaned public companies. Companies that came public with a set business plan and saw technology changes, end markets disrupted or some other event that made their strategy no longer viable.

    These companies often spend many years developing new products, technologies, and strategies under new management teams before the investing public takes note of them again. Compugen clearly fits this dynamic. A company that got its start in selling massive processing capabilities to Pharma; in their second phase they then began using their own processing capability to develop their own discovery platforms. The third phase which started under the current CEO Anat Dayag-Cohen now has the company developing drug candidates for its own pipeline.

    Discovering orphaned public companies with massive turnarounds underway is tough but it absolutely produces the biggest returns possible for investors, in my opinion. You're a skeptic, that's fine. No need to debate this now, let's just see how the next few years unfold
    Feb 27, 2014. 02:33 PM | 1 Like Like |Link to Comment
  • Compugen: A Wealth Creation Opportunity In The Making [View article]

    You wrote that "the company has stated on numerous occasions that they do not want to develop drugs past phase II." I guarantee that you will not find a single instance of the company stating that. If I am wrong please just find one of the numerous occasions where this was stated and I am happy to stand corrected.

    You claim to have been following the company for ten years but you seem completely unfamiliar with the company's strategy since Anat Cohen-Dayag took over as CEO three years ago.

    The company changed its focus from one of constantly developing and improving its discovery infrastructure (having a "science project" mentality as Anat says) to one of actually applying its infrastructure to specific areas of focus and discovering drug candidates and building a Pipeline of opportunity in immunology and oncology.

    As part of this change in strategy the company also engaged in partnerships with BioLineRx, DiscoverRx and Merck Serono to develop the non-core discoveries it had at that time.

    So, the glaringly obvious reason Compugen has not advanced any discoveries to Phase II clinical trials is simple: as the article states, all the discoveries in the current Pipeline were made in the past couple years and are therefore still in pre-clinical. So, how could they realistically have had a drug candidate in Phase II at this time?

    Further, the company only announced within the past year that they would look to advance their discoveries into clinical trials on their own. Their original strategy was to engage in all pre-clinical licensing arrangements but the company has rightly concluded that it can create substantially more shareholder value by holding and developing its discoveries further.

    When asked, they were clear that circumstances at the time would dictate how far they advanced them into clinical trials. Never, contrary to your assertions, did they state they would not develop them beyond Phase II. Whether they do or do not depends on how valuable the drug would be in a licensing arrangement versus the resources required and value that could be created moving forward alone. This is the calculus every drug development company considers.

    So to address your concerns: there have been no Phase II drugs because they are all pre-clinical. And the concern that they have "stated on numerous occasions" that they would not develop their discoveries beyond Phase II... no, they didn't. You completely fabricated that claim.
    Feb 26, 2014. 01:35 PM | 3 Likes Like |Link to Comment
  • Why InvenSense Is Worth The Premium [View article]
    Just want to correct my above post where I said that InvenSense products are MCUs, they are not, they are multi-chip package ASICs. That would explain the significant (favorable) power differential that InvenSense has over MCUs. It also explains why InvenSense would not be in flagship devices as they want to support bleeding edge features while the 9 month ASIC design cycles and lack of programmability severely limits their flexibility.

    By contrast QUIK's AL3 S1 which is a low-powered FPGA has programmable fabric will enable OEMs to push out sensor fusion algorithm updates to devices already in the field. Also the multi-chip package ASIC eliminates any flexibility for an OEM to pick best of breed discrete sensors to include in their design where QUIK is sensor agnostic. This is going to be a huge advantage for mid- to high-end phones. We'll know shortly, but based on everything I can gather I'd be surprised if the Galaxy S5 swapped out the MCU-based approach for InvenSense. In the future it would seem to make perfect sense to swap out the MCU for QUIK's approach though.
    Feb 21, 2014. 06:09 PM | 1 Like Like |Link to Comment
  • Why InvenSense Is Worth The Premium [View article]
    I'm not sure what gyro, accelerometer, or magnetometer are used in the Galaxy S4 but I do know it includes a Bosch pressure sensor. So, I'm confused as to how Samsung could swap out the ATML hub and the discrete sensors in the S4 and go with an InvenSense integrated sensor solution for the S5 since InvenSense doesn't seem to have any integrated sensor solutions that includes a pressure sensor. The data from the pressure sensor is necessary along with the data from the gyro, accelerometer, and magnetometer for providing things like pedestrian dead reckoning.

    Is there a link to the Seeking Alpha article that claims the S5 will use InvenSense. Or am I just missing an InvenSense 10-axis product announcement somewhere? It just doesn't seem possible that InvenSense is in the S5 unless it's going to have fewer features than the S4.
    Feb 18, 2014. 05:14 PM | Likes Like |Link to Comment
  • Why InvenSense Is Worth The Premium [View article]

    That same chart (page 19) shows a QUIK solution compared to MCU based solutions. ATML's sensor hub in Samsung's Galaxy S4 is an MCU as is Apple's M7 which is based on an NXP MCU. Invensense also labels all of their products as MCUs so I am a little confused as to how the MCU architecture for INVN could produce a power budget so drastically different from others (NXP, ATML for example).

    One possibility is that the Invensense link you provided vs. QUIK's chart 19 is not an apples-to-apples comparison as the Invensense product announcement is Motionsensors for "wearables" while QUIK's chart 19 is a comparison chart for sensor solutions for mobile device markets. Is it possible that the INVN products you linked are smaller less functional MCUs (and therefore lower power budgets) than MCU based sensor solutions targeted for the mobile market which is what QUIK's chart 19 is showing? After all, I don't think anyone suggests the M7 could go into an iWatch wearable, so the sensor solutions for mobile markets and wearables are different.

    One other dynamic I have not seen discussed is QUIK's catalog strategy. The assumption is the QUIK's programmable nature will appeal to OEMs high-end models while INVN's integrated off the shelf sensors solutions will be designed into mid- to low-end models. However, QUIK is in the process of developing integrated sensor solutions (what it terms catalog parts) with QUIK providing the hub and partners providing the sensors and the algorithms. This appears to me to compete with INVN.

    Is my understanding correct?
    Feb 15, 2014. 04:12 PM | 1 Like Like |Link to Comment
  • Why InvenSense Is Worth The Premium [View article]
    <<...but InvenSense's always-on solution operates at about 1/3 the power of QuickLogic's reference design.>>


    Very insightful article on INVN. Can you point me to your source for the above comment as it relates to INVN vs QUIK power budgets?
    Feb 11, 2014. 03:43 PM | 3 Likes Like |Link to Comment
  • QuickLogic And Sensor Hubs [View article]
    I've been researching QUIK for many years and just want to point out to you for the historical record the move in QUIK's share price on January 6, 2011 was not driven by QUIK's revenue growth or anything else having to do with QUIK. On that day it turned out Qik Video was acquired by Skype and there was a lot of confusion re: whether QUIK was being acquired. It also coincidentally happened during the CES conference so the price action led some folks to believe there was some big news coming out of QUIK at that conference.

    Once the Qik vs QUIK confusion cleared which caused the CES rumors to dissipate the stock tanked and then as you point out the revenue stalled out and went backwards.

    I think your timing on QUIK now is pretty good (says the guy with the demonstrably terrible timing when it comes to QUIK) and the sensor opportunity is a 180 degree game-changer for them.
    Feb 7, 2014. 08:45 PM | 2 Likes Like |Link to Comment
  • High Octane Stock Unilife Still Very Misunderstood [View article]
    From today's press release from Biodel (appended below) it is stated that the deal with BD is a "follow-on" to the deal with Unilife (i.e. "the auto-reconstitution device" which per the June press release is Unilife's EZmix). In fact, Biodel spent an entire paragraph of today's release highlighting their Unilife deal. And further if you read the description of BD's Uniject device (appended below) - which you conveniently eliminated from your cut and paste - you would see that that device is only used in 3rd world developing markets.

    "Biodel will develop the BD Uniject SCF™ Disposable Auto-Disable Injection System as a follow-on product to complement its auto-reconstitution devices, for which an NDA submission is anticipated in 2015. Biodel's auto-reconstitution devices are designed to offer two-year stability for emergency glucagon administration with little to no training. The auto-reconstitution devices automatically reconstitute lyophilized glucagon and feature automatic needle retraction upon full dose delivery, minimizing dosing errors and the risk of needle stick injuries. These features should make the auto-reconstitution devices compelling products for institutions, emergency responders, and many diabetes patients."

    "The BD Uniject SCF(TM) Disposable Auto-Disable Injection System was developed in the early 1980s with support from the U.S. Agency for International Development to address the need for a low-cost vaccine delivery platform in developing countries....The BD Uniject SCF(TM) Disposable Auto-Disable Injection System has been used in numerous vaccination campaigns in Africa, Asia, and Indonesia."
    Dec 17, 2013. 01:45 PM | Likes Like |Link to Comment
  • The Obstacles To UniPixel's Production And Shipments [View article]
    Wow Richard Roe you're famous they actually discussed this article on the conference call. If you didn't listen, allow me to summarize: (paraphrasing here, just a bit): "You're full of shit and a clown."
    Nov 7, 2013. 05:18 PM | 10 Likes Like |Link to Comment
  • The Obstacles To UniPixel's Production And Shipments [View article]
    "Mr. Killion shipped the first units on April 26, as pointed out in the article. Those units were promptly rejected and discarded within a month, as pointed out in the article."

    The language is so funny. They were "promptly rejected and discarded." Are you sure they didn't also spit on them first and then light them on fire they were so disgusted. I also heard the had a picture of Reed Killion that they started beating with their shoes and the engineer that handled the first touch sensors bathed his hands in acid as to remove any essence of Uniboss from his person. And that was all done "promptly."
    Nov 7, 2013. 04:15 PM | 8 Likes Like |Link to Comment
  • Uni-Pixel Investors Getting Fooled Again [View article]
    Alpha Exposure-

    You come off as a bit of a simpleton by pointing out that the word "revenue" is missing from the latest press release and accusing other of having poor reading comprehension skills while being completely unaware that the term used in the press release - "commercial production" - means the exact same thing as "production revenue" which was used in prior press releases.

    It's pretty clear that you're either dishonest or a complete dolt but I'll allow that those may not be mutually exclusive attributes in your case.
    Sep 29, 2013. 02:08 PM | 7 Likes Like |Link to Comment
  • Uni-Pixel - Progressing Into The Redzone [View article]
    Graycell- Very nicely done. It is interesting to see the typical short slugs (and I mean no disrespect to them) at least throwing in the towel on their old BS ("the market is small" "the technology doesn't work" "the partners aren't real") and it now seems there is agreement among longs and shorts that the risk is in the ability to manufacture to scale.

    But identifying a risk and identifying an actual short opportunity are not the same thing. For a risk to be an actionable short idea one needs to see that the risk is not accounted for in the valuation and that it will actually manifest. Given the prospects for several dollars per share of EPS in the fairly short-term its hard to suggest the current valuation is not heavily discounting these risk. And, more importantly, given how many parasitic short slugs (and again, I mean absolutely no disrespect) there are around the UNXL story, if there were any fire whatsoever to be concerned with, the smoke would have been smelled and brought to our attention by now.
    Sep 16, 2013. 03:18 PM | 6 Likes Like |Link to Comment