Same Old Story on Chinese State Procurement

The petition earlier this month by an assortment of business associations from North America, Asia and Europe protesting the not actively circulated circular (Notice No. 618) issued jointly by China's Ministry of Finance, Ministry of Science and Technology and the National Development and Reform Commission (link here) is just howling at the moon. This notice just lays out formally one more way that China tries to steer procurement towards domestic companies. And there is not much foreign companies and countries can do as China is not a signatory to the WTO's Agreement on Government Procurement. But since China's state procurement from foreign companies was quite small to begin with, there may not be much for foreign firms to worry about. China has used procurement as one of many industrial policy tools to try to promote local technology. The recent announcement by MIIT setting inter-network fees for TD-SCDMA (the "indigenous" mobile technology) subscribers at half the rate for subscribers using other technologies just goes to show how anxious China is to think up news ways to promote domestic firms and technology and continue old ones like procurement.

End of Taiwan Memory?

Taiwan's Legislative Yuan (the national legislative assembly) has cut off further support for Taiwan Innovation Memory Corporation (formerly Taiwan Memory Corporation). However, the Ministry of Economic Affairs (MoEA) does not seem to be completely giving up on this project since the cut-off applies to funding for next year. As the prices look better than they did in the recent past, the MoEA may find it very hard to pursue its plan of gobbling up failing memory makers on the cheap, without or without new funds from the legislature.

Richard Chang's Exit: Quick Note

Rumors are flying fast around Richard Chang's departure from SMIC. Some claim that Chang's departure was required by TSMc as part of the deal to end TSMC's lawsuit against SMIC. Others speculate that bringing in David NK Wang means that SMIC will merge with another firm. Would that firm be HuaHong? I'll be back more with more in a few days after reading all the tea leaves.

Is ChiNext China's NASDAQ?

ChiNext just started its trading. After one day, things look quite frothy. So frothy that according to Business Week (link here) some foreign institutional investors interested in it are sitting on the sidelines for now. The real question is will ChiNext actually serve as a NASDAQ-style exchange for entrepreneurial high-tech companies. Given that the first 28 companies were hand-picked by the government to list on the new exchange, the signs are not that hopeful thus far. Indeed, one need only look across the river to Hong Kong's weak GEM exchange to realize that new exchanges alone cannot serve to create another NASDAQ when the other institutions supporting tech entrepreneurship are missing.

Acer passes Dell

Acer has finally passed Dell in terms of global computer sales (link here). The rise of Acer and other Taiwanese brands, such as HTC, proves the wisdom of Acer's decision to focus on brand while hiving off its manufacturing operations, including manufacturing services. The Taiwanese fretted for years about their inability to create brands in sad contrast to big Korean brands like Samsung, but since adopting the same strategy for brand that they did for other parts of the value chain (i.e. segmentation of the value chain and then narrow focus on one small slice of the value chain), they have achieved great success without needing the scale and scope of Korean chaebol.

More news on the Grace-Hua Hong merger

I was in China about a week ago and a well informed source close to parties in both firms suggested that the merger is definitely going ahead. The likely CEO of the merged company will be Grace's Schumacher not Hua Hong's TY Chiu. Indeed, the source speculated that TY Chiu would probably leave Hua Hong for a second time because it is too hard to run this state-owned company due to government interference. Supposedly the planned merger will be basically financial with the two companies still maintaining operational independence.

First Solar to China

First Solar, US-based and the world's largest solar panel maker, has been seen as competitive with Chinese imports because it uses thin film (technically speaking it is cadmium telluride) technology. Admittedly this blog has also subscribed to this optimistic point of view (see my Aug. 26 post). However, now First Solar has announced plans to build a 2 GW facility in Ordos, Inner Mongolia (link here). One could decry the offshoring of arguably technology-intensive manufacturing, but the price of silicon, which is the material with which Chinese firms make their solar panels, is falling dramatically. In fact, the price of silicon is falling so fast that some Chinese solar panel firms, such as Trina, claim that their unit costs per kwh will be cheaper than First Solar's by 2010. Thus, First Solar's offshoring is an understandable reaction to these shifting materials costs, but this case provides no reassurance to those who aspire to compete with China-based manufacturing through investments in technology.

GlobalFoundries and Chartered: Impact on SMIC

The purchase of Chartered by GlobalFoundries should impact the foundry industry for years to come, but is it good or bad for SMIC? Mark Lapedus at EE Times (link here) argues that SMIC is a winner in the wake of this merger because there is one less leading edge foundry firm in the marketplace. I am not convinced the implications are actually so positive for SMIC because the GlobalFoundries-Chartered combination is more formidable than Chartered alone. And let's face it, among the big four foundires, SMIC and Chartered were battling to see who would not end up in fourth place so enhancing Chartereed's competitiveness impacts SMIC more negatively than TSMC or even UMC. With AMD as a steady client, sharing the cost of IBM's licensing fees over a larger scale firm and the deep pockets of Abu Dhabi's Advanced Technology Investment Corporation (ATIC), the merged firm is much more of a threat than Chartered alone despite Singapore's lavish support for the firm over the years. Furthermore, the merger does not seem to consolidate the amount of mainstream CMOS capacity out there since GlobalFoundries did not have much of that prior to the merger.

Ctrip takeover as a window on Taiwan's debate on China

Some have assumed under Ma Yingjiu's leadership further substantial economic integration with Mainland China was a given for Taiwan. At least they did until Ma's ratings were thrashed by the ramshackle government response to Typhoon Morakot. However, even before this natural disaster, the media commentary on Ctrip's (China's Travelocity/Expedia) unannounced effective takeover of a controlling share of ezTravel, its Taiwanese equivalent, demonstrates the deep fissures in Taiwan on the issue of economic integration. The narrowly focused economics media, such as CENS, treated this news without alarm (link here) and the China Times, which occupies the middle ground in Taiwan's political specturm with a slight tinge of KMT blue, even heralded it as creating a Chinese travel services brand (link here) but the very green (i.e. pro-Taiwanese independence) Liberty Times treated it with alarm as an invasion by a Chinese company which was just masquerading as a foreign firm (link here).

China Solar Hype?

Keith Bradsher's piece in the New York Times yesterday (link here) suggests that China solar panel makers are going to take over the US market even as the US federal government subsidizes investment in solar energy. He makes the good point that some of these Chinese solar panel operators are heavily government supported i.e. subsidized, but there are some reasons not to count out US competitors just yet. While the Chinese use crystalline silicon to make their panels, the world's largest solar module maker, Arizona's First Solar, uses thin film technology to remain the world's cheapest panel maker per KWH.
Meanwhile EE Times has suggested that Centron Solar will provide the organization that links disparate Chinese solar panel makers to its own sales operation in the US (link here). However, there are questions about Centron's scale as a company and how many panel makers it actually represents (link here).

Inspur (Langchao) and Qimonda

Inspur (浪潮) has now publicly confirmed (link) that it is buying Qimonda's design center, reports of which I wrote about in my August 11 post. The rationale of realizing synergies and expanding across the value chain offered by Inspur for the purchase still seems rather dubious to me.

Jingxin and Freescale

A report on Netease claims that Jingxin, a fabless design subsidiary of Techfaith, has bought Freescale's mobile chip unit (link). If true, this represents a major coup for Techfaith and Jingxin. The acquisition will bolster the value of the design services Techfaith can offer and Jingxin through its ties to Techfaith will present a strong challenge to likes of Spreadtrum.

CAS and Legend Holding's Corporate Governance

Given that the Chinese Academy of Sciences (CAS) has over 500 corporations under its umbrella, the announced goal to reduce the CAS' shares in each firm to 35% or less makes some sense since CAS can't possibly monitor all 500 firms effectively. Still, it was amusing to hear CAS state that one of the reasons it wants to sell 29% of Legend Holdings (CAS' total stake is 65%) is to improve corporate governance when it also insists that the buyer of the stake must to commit to not pressing for any change in the management or strategy of Legend Holdings for the next five years.

Buying Qimonda's China Assets

According to Taiwan's China Times newspaper(link), the Shandong-based Langchao Group will buy Qimonda's R&D facility in Xi'an and the Huarun Group will buy Qimonda's assembly and testing plant in Suzhou. The Taiwanese newspaper sees this as a strategic move for China to occupy upstream and downstream memory technologies and complete the DRAM production chain in China if one includes Korean Hynix's fab in Wuxi. There are several problems with this analysis. First, who cares about completing the production chain within one country. The critical segment of memory, the fabrication--which itself is hard to separate cleanly from the design in these memory products, is still in Korean hands. Second, it is really unclear what use the R&D facility in Xi'an will be to Langchao. Can this group of engineers in Xi'an create a manufacturing-ready cutting edge DRAM design? Probably not. And even if they could, would Langchao be able to do anything with it?

Perhaps Langchao's move is related to the rumors of the Shandong provincial government's own plans to build an IDM or pureplay foundry in the provincial capital of Jinan (see my very unromantic February 14 post). Combining this R&D facility with an actual fab makes the most sense in terms of utilizing a memory design R&D facility, but entering the memory business at this stage seems foolish from both a public policy and commercial perspective. Unfortunately, that may not stop Shandong and Langchao from investing.

Lenovo's Retreat to the China Market

In my July 11 post, I asked if Lenovo's efforts to consolidate its sway in rural China and the domestic market more generally were creating a position from which it would try to compete again in the global market or simply represented a retreat from the global marketplace. Second quarter results are suggest a tentative answer. On the surface, Lenovo appears to be regaining global market share, but the reality is that Lenovo is gaining in China while losing more the global market in Europe and North America (link). As Bruce Einhorn observed recently at his BW blog, Lenovo appears to be sliding back to the very China-oriented market position it had before purchasing IBM's PC division (link).

One Chairman, Two Companies? More Signs of a Grace-Hua Hong Merger

Hua Hong after four months of reorganization has announced that Mr. Wenbiao Fu will be the new chairman of the firm (link). This appointment suggests that the Grace-Hua Hong merger (see my post from November 1 2008) is finally going to reach fruition as Mr. Fu is already serving as chairman of Grace, a post he assumed on February 12 after stepping down as Chairman of the powerful municipal government agency, the Shanghai Municipal Informatization Committee.

NucTech, Namibia and State Procurement

The corruption scandal in Namibia involving NucTech (link), a firm spun-off from Tsinghua and run by Hu Jintao's son (Hu Haifeng) until he was promoted to head Tsinghua Holdings last May, reminds me of my visit to NucTech a number of years ago when it still had Tsinghua attached to its name. The company seemed to be reliant on state procurement rather than leveraging any internal capabilities just like many other xiaoban qiye (school-run enterprises). The ongoing probe by the EU to look into allegations that NucTech used soft loans from the Chinese government to win customers in the EU (link) suggests that state backing is still critical to its business.

Globalfoundries as an Emerging Competitor?

The announcement that ST will sign on as Globalfoundries' first major outside (i.e. non-AMD) customer (link) could be seen as a sign that the Big Four pureplays (TSMC, UMC, SMIC and Chartered) are headed towards an even more competitive foundry market. The speculation that ST is turning to Globalfoundries because ST was disappointed with TSMC's problems with 40nm technology reinforces the image of Globalfoundries as a serious competitor. I suspect the hype around the ST signing will prove to be just that and Globalfoundries will remain more of a niche player.

Tencent's Search Engine as Competition Policy?

Rumors are floating about that Tencent and Alibaba will release their own search engines. More intriguing still, according to the Marbridge Daily, Tencent's search engine has significant government support because the government wants a domestic competitor to Baidu. My guess is that the government's concern has to do with the lack of competition in this market rather than any dissatisfaction with Baidu's ability to keep the foreigners, well really just Google, at bay.

Lenovo Reinforces Rural Position

Lenovo's new "1+N+N" strategy of having county-level stores, plus township-level retailers and village-based "convenience windows" appears to be a bid to retain its strong hold over channels in rural China. Whether this market will be a base upon which Lenovo can leverage to compete in urban China and abroad or is the beginning of a long slow retreat to the periphery of the PC market to be seen.

WAPI's reappearance

China has once again submitted an application to make WAPI a global standard after several years of quieit on the wireless standards front (link). At least two of China's three major telecom service providers, China Mobile and China Telecom, are requiring WAPI support for all new wireless LAN gear. At least one Chinese report (link) claims that ISO is reconsidering WAPI because WAPI has proven to be far superior in its performance to the IEEE 802.11i standard. I remain sceptical of that claim and the standard (or at least the motives behind it).

China's brands gaining in China

The FT reported this week that Best Buy's Shanghai stores have witnessed a rapid turn away from Japanese and Korean brands and a move towards Chinese ones. Lenovo, Haier and TCL were mentioned specifically. It would be interesting to find out what products are involved in this trend. I suspect foreign brand mobile phones remain strong, but perhaps Chinese consumer are moving towards cheaper Chinese brands in computing.

Morris is Back!

Morris Chang is once again CEO of TSMC. While chairman and CEO of TSMC since its founding, Chang gave up the CEO role to Rick Tsai in 2005. A quick reading of this breaking news would suggest some troubling implications. First and foremost, TSMC may have an Apple problem. Apple without Steve Jobs has proven to be an ineffective company. Is TSMC dependent on Chang in the same manner? Probably not, but does this mean that Tsai is not the answer at CEO for future? If so, then TSMC may face a problematic search for a successor to replace the irreplaceable. Understandably it is hard to replace a semicon industry icon like Chang who is rightfully called the father of the foundry industry, but Tsai did not do too shabby of a job until the worldwide downturn hit. If they aren't happy with Tsai's performance as successor, who will make the board and Morris happy?

Taiwan Memory's waiting game

There is now a lot of speculation that Taiwan Memory is simply going to wait out the Taiwanese DRAM companies and buy up some of their assets on the cheap when they fold. If Taiwan Memory actually follows through this strategy, it could turn out to be a smart business move as well as a face-saving gesture for a company (to use that term loosely for this state-organized entity) so rudely rebuffed by Taiwan's DRAM makers when it offered to save them through M&A. Promos and Powerchip are particularly ripe for picking. Still, TM would have to gather most of Taiwan's DRAM capacity to really make a go of it, and Nanya is likely to continue to hold out with the help of the surprisingly resilient Micron, which has surpassed Elpida for third spot in the global DRAM rankings as of Q1 of this year.

Lenovo's continued slide

Lenovo's most recent annual financial results are not pretty. The firm lost $268 million over its financial year that ended March 31. With Liu Chuanzhi back in charge, one might think the firm will turn around. That's highly unlikely. Lenovo's reliance on the Chinese market and Chinese government procurement has made it perfectly unsuitable to compete against the big boys outside of its home turf (it even racked up losses in the Asia-Pacific region ex. China). The current strategy seems to be a further retreat from the global marketplace. This move may curb losses in the short term, but will do little to push Lenovo to upgrade its capabilities.

Intel vs. EU

In addition to Intel executives, some outsiders are crying foul over the EU's fine for Intel's anti-competitive practices (link). Peter Gumbel in Forbes believes that the EU picks on American firms for alleged anti-competitive practices because many European firms with monopolistic tendencies are owned by various EU member states so the EU is politically unable to go after them. Others, such this unsigned EE times editorial, see the Intel ruling as just another confirmation of Intel's anti-competitive after similar rulings against Intel in recent years in Korea and Japan (link). Looking at the ten biggest antitrust fines imposed on companies by the European Commission, four of the companies are in fact American. This could be construed as evidence of an anti-American bent to the EU's regulations, but three of the four fines were imposed on the recalcitrant Microsoft, which kept dragging its feet to comply with the initial European Commission ruling. Moreover, five of the other ten largest fines have gone against firms from EU member states with all five hailing from the three of the four largest EU economies, France, Germany and the UK (the eighth largest fine went against a South African firm). Given what we know about Wintel's market power, it looks like the EU's moves against Intel were probably as well justified as the EU's moves against market-bending Microsoft were.

UMC buys Hejian from UMC

The news that UMC has purchased the 85 percent of Suzhou-based Hejian that it did not "own" is quite amusing (link). After all, everyone knows that through affiliates UMC has always indirectly owned Hejian and treats it as one of its fabs. Well, everyone other than Chen Shuibian's government, which was shocked (shocked!) to find UMC had invested in China illegally behind its back. Thankfully, the technology investment restrictions by the Taiwanese government have significantly loosened over the years.

One Step Forward, One Step Back

Hynix and Samsung for the first quarter reported rising ASPs in NAND if not DRAM and credit cuts in production in Taiwan for these rising ASPs. Furthermore, they are considering capacity shifts from DRAM to NAND (link). Unfortunately, given the Japanese government's announcement of support for Elpida, the necessary reduction in DRAM capacity may take quite a long time to work its way through the global system as each country offers support for its own firms. Unless it can get Taiwanese government support despite initially rebuffing Taiwan Memory, Nanya and its partner, Micron, look like the odd men out.

Bizarre addition to Silicon 60

EE Times' Silicon 60 version 8.0 (link) contains a rather odd addition, Beijing's BLX. This firm is typical of the state-linked firms that feed at the government trough. Moreover, BLX has never put out competitive products in the open marketplace. Instead, this firm has relied on selling its inferior products to other state-linked firms, which in turn rely on selling their computers to the Chinese state. I hope for version 9.0 EE Times removes BLX and puts in its place any one of a number of worthy design start-ups from China.

Move into RF for ZTE?

Freescale is preparing to sell the jewel in the crown of its cellphone operations, its RF transceiver group, and ZTE is rumored to be one of two serious suitors (link). If ZTE could pull this off, it would be a remarkable upgrade for its IC design division. Unfortunately for ZTE, the other serious contender to buy the group is Samsung and Samsung has strong motivations to buy as Freescale's RF transceiver group would complement Samsung's other assets in IC design and downstream in mobile phones.

Spreadtrum Scores a Win

Spreadtrum looked like it was in bad shape given the competition from MTK in the domestic mobile phone gray market and its lack of headway in other markets, but perhaps things are looking up. Apparently, Samsung has agreed to use Spreadtrum baseband chips (link).

Movement to Services a Threat to China's Telecom Equipment Makers?

A growing trend in telecommunications has been the outsourcing of network operations from the telecom service providers to the telecom equipment markets (see for example this Wall Street Journal article). While Huawei has been cognizant of this trend for several years and has even published reports about it, a real question remains as to whether or not Huawei and ZTE will be able to meet the challenge of the move to services-cum-equipment model that appears to be emerging.

The DRAM Debacle Part II: Whither Taiwan Memory?

Taiwan's government has backed away from one possible blueprint for the government-sponsored memory chip firm, Taiwan Memory, announced last week to consolidate the DRAM sector into Taiwan Memory. Taiwan DRAM firms have been expressing a distinct lack of enthusiasm for the prospect of joining Taiwan Memory (link here), and Minister of Economic Affairs Chii-Ming Yiin announced on Thursday that the tentative plan now is to rent fab capacity from the separate DRAM firms for Taiwan Memory. Are the memory firms in Taiwan holding out for a better deal from the government? Probably so as these firms cannot survive the downturn on their own. What further complicates any merger are the separate deals these local firms have with larger foreign partners, principally Elpida and Micron. Last week the government suggested that Taiwan Memory would choose one or the other of these foreign firms as its technology partner so both firms have incentives to keep their current partnerships rather than losing them in consolidation.

Intel's Fab in Dalian Delayed

There have been reports (link here) that Intel's fab in Dalian will be delayed. Intel denies it of course, but Intel would be wise to deny it since any delay will anger China's government. Intel would also be wise to delay it given the harshness of the current downturn. Given these realities, chances are Intel is delaying it but won't ever publicly confirm it is doing so.

The Plot Thickens in Long-tan Science Park Case

Lee Jie-mu, the former head of Hsinchu Science Park Administration under Chen Shuibian, has admitted that the deal to buy land from the Koo family was done at the behest of President Chen. Apparently, the National Science Council's chief was kept in the dark about the deal although vice premier and head of the Council of Economic Planning and Development, Lin Hsin-yi, knew about the deal and was opposed to it (link here).

Vivek Wadhwa on the Reverse Brain Drain: The Implications for China

Vivek Wadhwa on his Business Week blog recently wrote about his research team's findings concerning the reverse brain drain of skilled immigrants from the US back home (link here). His research on this topic has focused mainly on Indian and Chinese immigrants. While I pointed out this worrying trend three years ago in an op-ed in the San Jose Mercury News (link here), the work by Wadhwa and his colleagues has gathered evidence of the growing dimensions of this reverse brain drain. Undoubtedly, losing these skilled immigrants is a loss for the United States, but the implications are not all positive for China.

What should worry China most about Wadhwa's findings is the relative youth of the returnees to China. The average of these Chinese returnees was 33. We often tend to think of technology workers as relative young and that this youth can bring a fair amount of new insight, energy and entrepreneurial enthusiasm. However, I have heard plenty of complaints from VCs, technological entrepreneurs and other industry insiders in China that many of these young returnees returned too soon without the requisite experience to be entrepreneurs or top managers in new China-based start-ups. Evidence from Wadhwa's study shows that these returnees enjoy a substantial move up in the ranks of management upon their return to China, but what is good for individual returnees may not be good for the development of China's tech sector. Ironically, these insiders generally feel China would be better off if these returnees stayed in the US to accumulate further experience before returning home. In recent years too many inexperienced returnees have taken on levels of managerial authority in China-based firms that they are ill equipped to handle.

What should also worry China are the factors that may turn the tide back in favor of emigrating to the US. It is clear that over the last several years Western VCs were becoming less and less willing to throw good money after bad in China even before the financial crisis dried up credit all around. Thus, even without the financial crisis, the VC environment was becoming much tougher. Although the survey by Wadhwa and his colleagues was done in 2008, it probably did not capture the full effects of this change in funding opportunities possibly because of some selection bias among the respondents. One could think that the financial crisis, having started in the States, would impact America more than it would China, but I firmly believe the entrepreneurial side of the tech sector in China will be hurt badly. Certainly China will probably not slip into negative growth, but with its export-oriented economy suffering from lack of demand from the rest of the world and the Chinese consumer missing in action, there were be many fewer appealing opportunities for China-based ventures. In effect, the China bubble of overeager VCs has popped just as the world has been hit by the financial crisis. It is hard to imagine the type of venture funding of years past will be readily available over the next several years. Even before the financial crisis I had heard many stories of disillusioned returnees heading back to the States after they realized the road blocks (institutional, legal and technological) that still act as barriers to sophisticated tech entrepreneurship in China. The growing crisis may just accelerate the reversal of the reverse brain drain and this "return" to emigration could last at least several years.

Return of TY Chiu

According to HuaHong NEC's website, TY Chiu has returned to Shanghai and the firm to take over as CEO effective February 1. Despite TY's past success in running fabs, I think HuaHong NEC will be hard to turn around. Still, given the state's strong support, HuaHong NEC will not collapse as many private IC firms surely will over the course of the current downturn.

Corruption Scandals and Taiwan's Tech Sector

Allegations keep coming out about various Taiwanese technology business leaders giving bribes to former First Lady, Wu Shu-jen (吳淑珍 or Wu Shu-zhen in pinyin romanization), while her husband, Chen Shuibian, was president. According to Next Magazine (台灣壹週刊), technology leaders, such as Hon Hai's Terry Gou (郭台銘 or Guo Tai-Ming) and TSMC's Morris Chang (張忠謀 or Zhang Zhongmou), were named in a February 3 affidavit by Wu (link here). Chang has denied any wrongdoing and TSMC has publicly donated to both the KMT and the DPP in the past (link). TSMC as the best run and probably most transparent company in Taiwan is the least likely candidate to be involved in such shenanigans. Another large tech firm is getting tainted by similar allegations. Quanta, the large computer contract manufacturer, appears to be under investigation for allegedly bribing the Chen family for land in the Longtan Science Park although the firm has denied it (link).

Shandong and Qimonda?

Infineon's supervisory board chairman hinted in a recent interview that a Chinese province had been interested in investing in Qimonda before it filed for insolvency (link here). I suspect the province was Shandong because the Shandong provincial government has been pushing for the creation of a large IC base in Jinan, the provincial capital.

Hejian in a New Light

News has come out that casts the prosecution of UMC executives over investments in Hejian, UMC's fab in Mainland China (link here), in a new light. Apparently, the Chen Shuibian administration allegedly tried to hit up Bob Tsao for political contributions and threatened to influence the outcome of the ongoing prosecution over the investment in Hejian if Tsao did not cough up the money. Essentially, the government's prosecution case claimed that UMC had violated Taiwan's investment laws by investing in Hejian, which is located Mainland China. What was strange about the prosecution is that it started in 2004 long after UMC's involvement in Hejian was publicly and widely reported, such as in an excellent article in WSJ on April 1 of 2002 (p. A8) by Jason Dean and Terho Uimonen. Given these reports, the Taiwanese government could not plausibly have not heard about UMC's reported investment in Hejian, but the government waited two years before bringing a case. With these new reports about demands for political contributions, the delay may be explained by the fact that Chen's administration never real wanted to prosecute as much as shakedown UMC.

Elpida Taiwan merger

Elpida apparently has been able to agree to a merger with Taiwan's Powerchip, ProMOS Technologies and Rexchip (link here). While the details have yet to emerge, this merger might mean that Elpida has agreed to some Taiwanese direct or indirect state representation on its board as the Taiwanese government had been demanding that. In return Elpida should receive some Taiwanese government assistance. On the optimistic side, this merger may be spell the end of Taiwanese state support for Taiwan's failed DRAM firms, but we won't know until the details of the merger are made public and maybe not even then.

The DRAM Debacle Part I

Years of over-investment, particularly in Asia, has led to the DRAM debacle. The global financial crisis has exacerbated what already was shaping up to be one ugly example of excess capacity. Kudos then to the European Commission for telling Qimonda to buzz off (link here). Of course, it appears that Infineon's wise decision to refuse to throw good money after bad played a key role in preventing a Saxony state-led bail-out for Qimonda.

In contrast to these wise choices, the Japanese government still seems to think that DRAM is a critical technology for the rest of its electronics industry (I guess they are stuck in the 1980s or maybe the 1970s) and is preparing to bail out Elpida (link here). At the very least Elpida is probably viable given its scale and technology so the Japanese government is not throwing money down some hole.

One cannot say the same thing about Taiwan's industry. Taiwan's DRAM makers have never attained the scale to support independent technology development so they have remained dependent on foreign partners for technology and been stuck in the low margin DRAM foundry business. Eight years ago two electrical engineering profs at MIT and I pointed out that Taiwan's DRAM sector was underperforming and would continue to fare worse than Taiwan's foundry and then-emerging TFT-LCD sectors given the levels of patient capital needed to attain the scale economies necessary to support continuous innovation in DRAM. The key problem is the mismatch between these capital requirements needed for a successful entry into the DRAM industry and Taiwan's financial system, which is geared more to supporting smaller scale investments with shorter time horizons (link to our early working paper here--the final version was published in Industry and Innovation later in 2003 due to the lightning speed at which academic journals review manuscript). Today Taiwan's government still won't face the facts and give up on this asset-destroying sector. Instead, the state has rolled over loans to the DRAM makers as it contemplates more comprehensive aid packages (to be detailed in my next post).

Of course, the Koreans are not completely blameless either. Some believe that the Korean government's willingness to back Samsung's massive investments irrespective of market prospects is what caused the DRAM glut in the first place (link here). Of course what Korea did Taiwan did as well except on a smaller more fragmented scale.

One has to feel sorry for Micron as it finds itself in the middle of this mess. According to an electrical engineering prof at a major West Coast university who has consulted with Micron intensively over the years, the firm has really made great strides in its manufacturing efficiency in recent years, but this consultant fears that this progress will be all for naught as Micron's overseas competitors receive massive state subsidies and push it out of the DRAM business. Then again, with rumors circulating that Micron is looking to buy Spansion, maybe Micron can still diversify away from this firestorm instead of being consumed by it.

TY Chiu and Other Details of the Expected Hua Hong NEC-Grace Merger

The news on the rumored merger of Grace and Hua Hong NEC is mixed. On the one hand, the media reports one unnamed Grace official as saying "There is already no road back from integration" i.e. it's a done deal (link here). On the other hand, there are still reports of several major issues holding it up. First, the shareholding of the two companies is complex and two state-owned entities,CEC and Shanghai Municipal SASAC, are the respective controlling shareholders in Hua Hong NEC and Grace. Of course, the shareholding structure of Grace is especially opaque so Shanghai SASAC or controlled investment arms of it may not own a controlling stake in Grace. Second, Grace has a lot of assets (fabs basically) but Hua Hong NEC has higher revenue so there are bound to be fights over how to value each firm's contribution to the new entity created by the merger.

In one potentially very good piece of news, it is also rumored that TY Chiu, currently President and COO of Silterra, will return to Shanghai to help run the new firm. If true, this piece of news is most welcome as TY is an excellent manager of fabs. Of course, the question is really whether or not TY will be given the freedom to run the firm as he sees fit. On this matter, there should be plenty of doubt for those who know Hua Hong's past well. After all, TY's last brief stint at Hua Hong as COO was not a successful one given the powerful constraints placed on Chiu and the rest of David N.K. Wang's management team by Hua Hong's state owners.