It could be difficult to get an end of Friday rally today in HIMX with the markets heading down.......and as traders tend to sell off shares so they don't want to pay margin interest over two non-trading weekend days. But, HIMX is swimming upstream and showing good resislence so you could be right and it does continue to rally in the face of these headwinds.
I do see some different MM's trying to control the ask on HIMX. Generally, it is NSDQ with the larger volume asks......but today I see lots of batx, edgx and arcx up there with volume on the asks (my AmeriTrade level II quotes always shows the larger bid/ask MM on top which is generally NSDQ but I am seeing others today).
I've also been watching the Chart thinking the shorts were working to force the 50 DMA under the 200 DMA to help influence some traders to sell and free up shares they can use to cover.......and am no longer able to do this on Yahoo where only the 1 day and 5 day charts seem to work any longer (so have to go to chartsDOTcom). Between that and the new format and message boards.......is Yahoo trying to divert eyeballs from Yahoo Finance as they sell it?
Today is a good reversal in price trend...and on decent volume thus far. Next week's Jefferies Conference will help also.
But, 20 million is a lot of shorts signifying financial power and influence (among MM's, etc.) and they are not going to lay down and take it. I see very few short squeezes and doubt we see a big one here.......unless it was only one entity and the other big players smelled an opportunity worth taking.
I saw nearly 6,000 shares trade after hours last night $0.06 above the closing price also indicating the shorts want to cover where they can.
The recent Hololens tear down has to be relevant here also.......in one direction of the other. I was wondering yesterday but today might be the confirmation HIMX is in it (as we all knew but the doubters needed proof). Now for Wu's claim that all were underestimating their bill of goods in these devices.
As for the HPU, I do wonder if HIMX had some involvement there and perhaps gets a license or royalty on these as I would think they had to be involved to some degree as they work with the LCoS displays and HIMX has lots of driver and timing controller expertise which could have been of value in developing the HPU.
Finally on Sony PSVR system......this is still VR so I wonder what HIMX's bill of goods is here. I checked a few weeks ago and coldn't see where you could pre-order the entire 'starter' system including the console, headset, sensors and controllers. Perhaps I need to look again.......but I am not much of a gamer but have been using my Gear VR headset for various experiences and would be interested in view sports, concerts, travel, etc via VR.
From the Verge article:
Microsoft has finally revealed exactly what's inside its HoloLens headset. While The Verge got exclusive access to a deconstructed HoloLens developer edition back in April, Microsoft has been keeping the details of its special Holographic Processing Unit (HPU) very secret. Microsoft revealed most of the HoloLens specifications earlier this year, and the special HPU is designed to do most of the processing so the CPU and GPU are able to just launch apps and display the holograms. Microsoft custom designed the HPU and it takes all of the data from the cameras and sensors and processes it in real-time so you can use gestures accurately.
At the Hot Chips conference in California this week, Microsoft devices engineer Nick Baker provided a presentation on exactly what's inside the HPU and how powerful it is. The Register reports that Microsoft's special custom-designed HPU is a TSMC-fabricated 28nm coprocessor that has 24 Tensilica DSP cores. It has around 65 million logic gates, 8MB of SRAM, and an additional layer of 1GB of low-power DDR3 RAM. That RAM is separate to the 1GB that's available for the Intel Atom Cherry Trail processor, and the HPU itself can handle around a trillion calculations per second.
Microsoft's HPU is low-powered so it draws less than 10W from the power supply in order to handle gesture and environment sensing. The Register reports it also includes PCIe and standard serial interfaces, and Microsoft has added 10 custom instructions to speed up special instructions made by HoloLens for its augmented reality algorithms. If you're interested in seeing all the components inside Microsoft's HoloLens then check out our full tear down right here.
This HPU would seem to be from HIMX.....and due to their display driver and timing controller technologies.
Hololens if probably more powerful than the VR units out....and those must be tethered to a high powered PC/laptop.
Microsoft strips down Hololens, reveals 24-core custom processor inside
August 24th, 2016 at 6:30 pm - AuthorMatthew Wilson
We have known about Microsoft’s holographic, augmented reality visor, the Hololens, for some time now but we did not know exactly what was powering it under the hood, until now. This week at the HotChips conference, Microsoft stripped down the Hololens for the first time, revealing the custom-built 24-core, 28nm processor running the show.
We already knew that the Hololens was a standalone unit and unlike many of the VR headsets we have today, it did not require any additional computing power. However, this is the first time Microsoft has gone into detail on the hardware.
The processor is being dubbed as a ‘Holographic Processing Unit’ and it uses a custom 24-core design based on the 28 nanometer process. It has 65 million logic gates, 8MB of SRAM cache and a layer of 1GB of DDR3 RAM all in a tiny package. According to Cnet, the chip measures in at just 12mm by 12mm.
This unit handles all of tech inside of the Hololens, including not just the holographic elements being displayed on screen, but also things like gesture recognition, environment sensing and more.
Unfortunately, the Hololens is still a developer-only device for the time being. You could technically buy one if you wanted, but with a price tag of $3000, you should probably wait for the technology to flesh out if you want any real use out of it.
KitGuru Says: I do enjoy following all of Microsoft’s Hololens updates and that custom processor is certainly impressive given all of the work it has to do. Perhaps we will see devices like the Hololens become more common next year when Microsoft releases its Windows 10 Holographic update.
This is huge and part of what Wu was referring to some time back when he or someone coined 'HPU'.
I found articles on this on various sites now over the past two days and this could really be huge......as look at AMD and NVDA stock price gains on the back of graphic processor unit sales.
Microsoft HoloLens is the world’s first fully self-contained, holographic computer, enabling users to interact with high-definition virtual holograms in the real world. The secret sauce of Microsoft HoloLens is its Holographic Processing Unit, a custom chip which allows the device to process and integrate the data streaming from its various sensors and deliver it at a rate its Atom chip can process.
Now at the Hot Chips conference Microsoft has finally revealed the secrets of that chip.
The HPU is a custom-designed TSMC-fabricated 28nm digital signal processor (DSP) with 24 Tensilica DSP cores that can process up to a trillion instructions per second with its 8 MB SRAM and 1 GB of DDR3 RAM.
The chip only takes up a 12 x 12 mm BGA package and performs 200 times faster and at lower power (only 10w) than a software-based solution.
The HPU uses the Tensilica’s instruction extensions to add 10 custom instructions to the DSPs to accelerate specific operations needed by the HoloLens and the data it delivers to the Intel Atom x86 Cherry Trail system-on-chip that powers the Windows 10 operating system is highly processed, meaning there is much less work for the Atom processor to do.
The HoloLens is currently on sale to both researchers and for commercial use for $3000. The Microsoft HoloLens Commercial Suite includes the Development Edition hardware as well as enterprise features for added security and device management. This suite acts as a solution for companies to confidently pilot and deploy HoloLens inside their organization
Think about where the money is coming from which is buying USA bonds.
Then, consider the stress testing of banks to ensure they are properly capitalized.......and they key is that the properly capitalized banks invest their cash primarily in USA bonds which then creates the demand for those USA Gov't bonds.
Agree. Every 1% on our $19 trillion in debt is $190 billion per year in interest payments. As the USA is already running deficits, this additional $190 billion per year PER 1% increase in rates would increase our deficits significantly.
Note that this is also why the Fed, under the Obama Admin, has gone to shorter durations (more 2, 5 and 10 year notes and fewer 30 year bonds than typical).....as short term rates are much lower than longer term rates. Whether that is smart of not will be played out in the next few years to decade.
What is really troubling is that neither Hillary or Donald are really talking about the debt.....but are talking about stimulus and infrastructure which means more deficit spending (although putting a 5% to 7.5% tax on repatriation of overseas funds would help a bit).
Finally, lower interest rates stimulate consumption as there is little benefit from savings. And, at this point in time, more consumption is needed to grow and create jobs.
You need to re-think your comment concerning thinking GNW is only worth $3.92 because a few shareholders are willing to sell their shares for that. $3.92 is the lowest price some shareholders were willing to sell their shares for. So, while you can buy some shares for $3.92.......you can' buy them all for that price.
So, what GNW shares are worth is not as simple as you think. If someone wanted to buy the whole company, I suspect they would not get all the shares at $3.92.
Now, there are lots of reasons to sell (need money, prefer something else, etc.)......but few real reasons why people buy and that is generally to make money. So, the person selling at $3.92 definitely preferred that price to the shares for various reasons.....but the person buying GNW for $3.92 definitely thinks the shares are worth much more than $3.92 (personally, I only buy stocks that I think are undervalued by 25% or more in the future).
Great point.......and why we were discussing the seemingly negative tone of the questions by the analysts at the Conference Call. So, the lack of public upgrades at this point could indicate most of the large investment banks could have been short and are still looking to cover (as in take the share price back down first and then covering).
If you look, the analyst estimates tab has been undated with FY2017 being updated and increased from $0.48 to $0.54/share which is good.....but probably on the lower side of what is should be. Even FY2016 has been updated from $0.39 to $0.42........although 3Q 2016 at $0.10/share is still not at HIMX's own guidance of $0.10 to $0.12/share. 3Q 2016 revenue was increased to 215.6 million and within HIMX Mgmt of 5% to 10% over 2Q 2016 actual of $201.1 million.
So, good point........and why are we not seeing PR's with price target updates? The silence is telling.
We will have to wait and see when the first half of August short is released late next week. But, I don't doubt that the shorts were trying to scare longs away from HIMX as much as they were trying to scare longs into selling.
But, as the price went down from $8.99 to $8.13 from 1-Aug to 10-Aug, I suspect there were not covering very aggressively if they covered at all. But, we should soon know.
Besides what they did in early August, the key is what are the going to do now. They can't be happy with the CC being so good and bullish going forward......including the start of the new AR products Fab Plant along with Pokemon proving interest in AR. All they can do is try to keep a link on HIMX and then start shorting again from a higher level to raise their breakeven point.
Looking at the volume (lower to average) and price action (from $8.99 at end of July and $8.13 going into earnings), I doubt the shorts covered going into earnings. This is puzzling to me as I would have covered.
Now, I haven't look closely at the options as perhaps they hedged/covered using options.
But, I do see the shorts as being in a weaker position now as we have a good quarter report, an easy comparison coming with good guidance , confirmation of the Fab Plant plans (with more debottlenecking of existing plant) and reasonable valuations with Mgmt aware of the shorts (may see more PR's and a pre-announcement in early Oct IF NEEDED).
What is not to like........and what would be the short thesis?
Having written this, I can see the shorts perhaps letting this run to exhaust the buying interest.......and then being able to short more from a higher level raising their breakeven/profit price level.
GNW seems to be turning from beingl priced for failure. My feelings are:
+ Paying 2018 debt early would be great as the debt trades below par so there is a savings to call these........and cash held by GNW is earnings very little.
+ To do this, GNW might need to raise cash by selling assets. They should do that on their lower earnings assets......but they should only hold cash reserves for what they really need. Pre-paying 2018 debt helps reduce the need for cash reserves...which earn very little return.
+ Debt rating is not really needed as they are not taking on new debt......and shouldn't. A lower debt rating only makes taking out debt cheaper.
+ Credit rating is needed for selling policies.......which GNW doesn't seem to be focused on these days as per their own news.
So, I fully support using cash and assets earning less than future debt repayment interest rate and potential discount to market.
I included the institution shares in the float when I added in the number of shorts. So, I do think my float calculation is mostly correct: Total shares less insiders shares plus shorts.
But, my key point is that the float could shrink dramatically as the shorts cover.......and the current 20 million shares is a historically large number............all of which are now down nearly $1.80 in the past 3 trading sessions a they went into earnings around $8.10.
Intersting article on Microsolf widening their Hololens offering beyond developers to corporate enterprises at:
Keep in mind that the float includes the shares shorted as they are 'owned' by two parties:
+ The party that purchased the shares (and whose broker lent them out for cash) and
+ The party that bought the borrowed and resold shares from the shorter
So, the total shares available for buying include the total shares of 172 million less those held by the insiders.......plus the 20 million of shorts. Once can argue that the institutional holdings are also mostly not available for trading.
But, the real key is that the float is higher by the 20 million shorted shares which now count double. This float will get much smaller after 10 to 15 million shorts cover.........so the first shorts to cover would do so with a higher float than the later shorts to cover as the float will be smaller by the number of shorts which cover.
Thanks Jar and Cav,
Ffor your insights and refreshing my memory on these things. I am fairly new to GNW having bought $2 Sept calls prior to earnings as before I was just watching from a distance.
Looks like I need to go back and re-read the 2Q 2016 CC and I am now recalling some of what Cav writes. I was going off the analyst estimates which are more solid than before.......but that is not the full outlook and thus back to the transcript I go.
Fun and game in after hours trading today in HIMX.
Time (ET) Price. Share Volume
19:47 $ 9.69 High 100
19:08 $ 9.69 100
19:03 $ 9.40 Low 100
18:58 $ 9.69 200
18:41 $ 9.69 300
18:41 $ 9.69 500
16:57 $ 9.66 380
16:26 $ 9.58 21,891
16:02 $ 9.4193 134,792
16:00 $ 9.58 100
Selling the Ausi unit to pay debt is also about timing. Yeah, it is nice to please Moody's but they need to carefully consider selling any asset earning more return than their cash........and that could be nearly everything but LTC. So, they should milk the Ausi cash flow as long as they can......unless Ausi is a depreciating asset which it shouldn't be.
So, screw Moody's to some extent. A lower Moody's rating doesn't matter if you are not selling new bonds......which GNW doesn't and shouldn't. So, it now only affects the secondary bond market as GNW has to pay the 'coupon' rate. So, why not have your bonds rated lower and periodically try to take some out with cash at a discount as cash is not returning hardly anything. They just need enough cash flow for recurring debt/claims/benefit payments and a decent cash reserve. They should know their business and needs better than Moody's.
GNW should just balance the return on the Ausi sale with the return on retiring bonds as adding cash to cash investments gets them little return on that cash. So, don't overly liquidate cash producing investments unless and until you need to.
The lack of questions about debt and debt payments might have been to keep that quieter while accumulation takes place. These discussions could well be taking place outside the CC. Better 2Q numbers (including guidance) did not take out the $4+ high after a decent (but not as good) 1Q2016 report. We did get a decent share price increase...from a lower base.
I also don't see the pressure on the ask as before (it is still there......just not as constant) and the shorts covered during both halves of July.
There are fewer good investments with limited downside in today's near record markets........so GNW and other financials are probably being looked over hard by several investment concerns,