"Korean firms don't compete to win, they compete to annihilate the opposition" -- Samsung is keeping Sharp alive with bailout funds, and is a major customer of other competitors including Mediatek, Himax, Sony, and on and on. They make all of the components that they buy, and yet still buy from their competitors. They don't care about annihilation or they wouldn't be resuscitating their competitors.
China can partner its way into producing memory (with SNDK or whoever), but it can't go it alone without endless years of substandard products that are in constant danger of infringement, which would mean they couldn't be exported. Without a partner It would be similar to comparing TSMC and SMIC. SNDK is being acquired because it was the weak member of the herd, and if it partners with China (through WDC) it's for the same reason.
"The more general point that was made from the start of this debate is that the theory that the consolidation that has occurred in memory will lead to excess profits is not correct." -- I don't know about excess, but an improved situation. Look at past lulls in demand and you'll see massive losses and bankruptcies across the memory industry. I don't think we'll see that during this downcycle.
This topic has been debated to death in the past elsewhere. Let's revisit it in 3 or 4 years when China may have had some effect.
When Japan, Taiwan, and Korea entered the market there weren't thousands of patents to contend with, and a decent sized firm was one of dozens in a much smaller market. The technologies and the market were immature, and now the market has a few very strong players in a market in the tens of billions. A couple of decades ago a player the size of Nanya would have been the largest in the world, and now a player like Nanya is irrelevant and has to license technology just to keep going. It's like saying that Japan successfully did well in a little league game despite not having ever played baseball, so China should do well in an MLB game despite never having played baseball. Whole different ball game. China has not lead any of the major techs it's tried to enter, including the foundry business, the display business, the processor business, etc., and memory is harder than those.
You seem to view Samsung as invincible, but they aren't trying to have a monopoly position in the markets they are in, and they often don't have the best technology. They are a major LCD producer, but Sharp has better panels, and Samsung has tried to keep Sharp alive. They are an LCD driver maker, but Samsung buys drivers from Himax. They are a major image sensor maker, but they buy sensors from Sony for their top phones. Speaking of their top phones, they have been losing market share the past couple years, both to upstart Chinese phones and to Apple, both the low end and high end. I could go on about Tizen, app processors, etc., but the point is they are far from an invincible force.
" is it not too early to say who is the winner in these technologies?" -- Yes, it is too early, but the point was that your contention that Samsung is somehow running away with it is...too early.
By the time a new player builds a fab to enter the market, DRAM and planar NAND will no longer exist. If someone new takes over, it will be with a revolutionary technology, and that's too hard to predict.
conversionworrier, you're vastly overestimating China's abilities.
China has wanted a high tech industry for many years and has never managed to foster it. All of the tech areas that China has succeeded at are relatively low tech, low barrier to entry, like displays or consumer electronics. All of the areas that are high tech, high patent dependence, high cost, it's completely failed at. Example: read about the loongson/godson processor that was going to destroy Intel and ended up having zero effect.
As for patents, yes, China cares about them. It has to in order to be a stable, mature player in technology. Why do you think China is currently spending tens of billions on buying tech companies? For the know-how and patents. In back rooms and labs they can ignore patents, but for mass production this would result in huge international trade embargos that would ruin them (and the world in general would suffer).
As for Samsung, OEMs would regard it as a disaster if they were completely beholden to Samsung. Their costs would skyrocket and they are aware of this. Look at Intel's dominance and its high margins. OEMs go out of their way to divy up the revenue in order to keep multiple players alive. Look at the apportioning of sockets by Apple. They intentionally give out sockets to all of the major RF players so as to keep some healthy competition. If it was possible for Samsung to take over the memory world they would rule the world, because memory is omnipresent. Samsung also does not have a lead in R&D, though it does have a lead in process technology (compared to other memory players). Micron/Intel seem to have superior products with 3D NAND and 3D XPoint.
John, as turturo said, ChipMOS Bermuda is a holding company. It has no operations. There's a very limited set of things that the money can be spent on without operations. The premise is that the Bermuda holding company will soon have no reason to exist, so it should also have no use for most of the money that it holds.
Sean, I sent you an email to your Yahoo mail a while back, not sure if you got it.
conversionworrier, agreed, it would be nice if the board was better aligned with shareholders. Shareholders (owners of the company) are concerned with a cash return per share on the investment, whether the cash is actual cash or just an accretion of value. Many directors and managers are just concerned with an increasingly large and stable company, where per share return is a very secondary motive. An ideal board should include someone with an investor mindset.
There's an article on Forbes right now praising KLAC for returning $16.50 to shareholders in a special dividend less than a year ago, as well as paying a regular quarterly dividend. In fact, KLAC took on debt to return that much cash to shareholders. The article makes the point that if the company hadn't done this, Lam could have used KLAC's own cash to buy it out and shareholders would have probably ended up with less in the takeover. This is a lesson for ChipMOS.
Here's one of the paragraphs from the article:
"Had the company not released nearly $3 billion in cash and levered up a bit, a private equity or corporate acquirer might have tried to borrow against KLA-Tencor’s balance sheet to buy it, doing little to generate value for public investors. Dell’s take private by Michael Dell and Silver Lake Partners in October 2013 is a prime example."
There are nonstop deals in the semi industry right now. It would be tragic for all shareholders if management does not act soon and prevent the company from being acquired in this fashion. Use all Bermuda cash for a buyback, or failing that distribute a special dividend.
Sean, yes, I definitely agree with that.
The company has our money with the mandate to invest it for us in a profitable way, not sit on it. A simple and profitable way to invest the cash is by buying its own shares at a cheap price...and once again we have a cheap price.
conversionworrier, the question isn't so much XPoint or not XPoint as the packaging involved because the service that IMOS supplies MU for DRAM is packaging (and not testing). Packaging is always changing, and is different for different end products. Whether a chip uses DRAM or XPoint, the packaging will keep evolving over time and IMOS will need to evolve with the changing packaging needs.
Even if there was a breakthrough tomorrow that suddenly increased OLED yields and decreased OLED prices, it would take years for OLED to replace LCD because of the huge amount of LCD production capacity. I don't think it can be argued that your average internal combustion engine vehicle has technology that matches your average hybrid, and yet hybrids are a very small minority. It's a value proposition. Similar situation with LCD and OLED, where price and production capacity favor the older technology. Either way, IMOS should be able to gold bump both equally well.
sean, good question...NTP is certainly in an interesting position for a former electronics manufacturing company.
Thanks John, I appreciate it. There are a number of strong investors that frequent this board and I hope we all get a chance to meet up again sometime, as some of us did in Vegas. I've occasionally had the pleasure of seeing friends from here at other times too. I've been busier with the fund as you mentioned and so have put less time into posting, but still read the board and also SA.
conversionworrier, neither LCD nor DRAM will be going away any time soon so current equipment will get plenty of use. OLED will be a minority technology for a long time yet, as will 3D XPoint. At least some of the LCD equipment (maybe all?) should work for OLED...no idea on XPoint/DRAM at this early stage, but I'm not concerned. OLED has made moderate inroads in mobile and companies like Himax have prepared for that with new drivers, but it's had no effect in large panel displays, and even OLED giant LG Display doesn't expect to do a lot in that area for a few years.
ristrau, I suppose it's no news flash at this point to say that there's a lull in demand in China, and the market has (over)reacted to that. At heart this is still a company with a lot of long term cash earning power and physical assets.
ChipMOS is capable of doing OLED driver bumping. Note that on their web page that talks about gold bumping they mention "LCD and other flat-panel display driver semiconductors."