Micron And China's Semiconductor Dreams

Summary

• Tsinghua Unigroup is unlikely to ever acquire Micron or even form a major joint venture with Micron to build a new fab.
• The true replacement cost of Micron is significantly more than the $26/share that John Pitzer suggested as the replacement cost of Micron's fabs.
• China will spend 1 trillion renminbi over the next 5-10 years to develop its semiconductor industry, but this investment is unlikely to result in a new DRAM fab from China.

The two most important economies in the world today are those of China and the United States. China's annual GDP growth rate from 2010-2014 has ranged from 7.4%-10.6%. In the same period, annual GDP growth in the US ranged from 1.6%-2.5%. China's GDP (in USD) in 1980 was $303.76 billion and has grown to $10.36 trillion in 2014.

Over the last year, fears of a massive slowdown in China's economy have escalated dramatically. Former Treasury Secretary Hank Paulson stated China's economy has "run out of steam." In an effort to keep the economy growing over the next few decades, Chinese policy has shifted its focus to the technology sector. In 2014, China developed a policy in which it will invest up to 1 trillion renminbi (~$157 billion USD) over the next 5-10 years across the semiconductor industry.

In July 2015, rumors of Tsinghua Unigroup bidding $23 billion to acquire Micron (NASDAQ: NASDAQ:MU) surfaced. There is no reason to believe that a bid at that price from a Chinese company is going to result in a deal. The DOJ and FTC might not like such a deal. Senators John McCain and Chuck Schumer certainly aren't fans of a potential acquisition of Micron by a Chinese company for national security reasons. Shareholders of Micron are highly unlikely to agree to sell their shares for $21 apiece. There was never really a chance of this deal occurring. I suggested in this article that Intel is significantly more likely than Tsinghua Unigroup to acquire Micron. However, even an Intel-Micron deal is not all that likely.

It's clear that Tsinghua Unigroup is not acquiring Micron. In recent weeks, rumors of some sort of Micron and Tsinghua Unigroup partnership or joint venture have emerged. It seems almost as unlikely as a Micron-Tsinghua merger. Micron does not need capital from China and it already has arguably the best partner around in Intel (NASDAQ: INTC) to develop and build NAND, XPoint and whatever next generation memory emerges. So why would Micron enter into any sort of relationship with Tsinghua UniGroup?

The first thought would be that China's stance on developing its semiconductor industry at any cost could pose a competitive threat to Micron. China is committed to invest $157 billion over 5-10 years to develop its semiconductor industry. A large percentage of that money could be allocated to build a fab to manufacture DRAM (and/or NAND) if China insists on becoming a major manufacturer of memory. Samsung (OTCPK:SSNLF) is planning on spending $23 billion to build a very large fab that will begin production some time in 2017. At a Raymond James conference on September 22, Micron's Kipp Bedard stated that he believed it would cost a Chinese company $40 billion-$50 billion to get started and at least 5-10 years to build a viable DRAM fab on its own that could compete with Samsung, SK Hynix and Micron. Bedard stated it would be "extremely, extremely difficult" for China to organically get into the DRAM market as the intellectual property is inaccessible to anyone outside of the current DRAM manufacturers.

That's spending potentially over 30% of the $157 billion budget over that timeframe on a DRAM fab that would increase wafer capacity and possibly cause prices of DRAM to decline dramatically. That sort of new capacity would cause the Chinese fab to incur annual losses (in the billions) until the market finds a way to absorb the new wafer capacity. I doubt the Chinese government would be willing to sink that much money into a project that is so likely to fail. Therefore, I believe there is close to no chance that a Chinese owned DRAM fab will emerge without heavy involvement from Micron, SK Hynix (OTC:HXSCF) or Samsung.

Micron does not need capital from China. No Chinese company currently offers intellectual property that could further Micron's interests in next generation memory. Micron can be highly successful without the aid of Tsinghua Unigroup. However, there is still potential for some sort of joint venture or partnership. Someday (in the not too distant future), Micron will need new capacity that can only be achieved by building a big brand new state-of-the-art fab. Kipp Bedard expressed at the Raymond James conference that Intel may look to build 3D XPoint outside of its partnership with Micron as Micron will just not have anywhere close to enough fab space to build enough wafers of XPoint to meet what seems to be a potentially enormous amount of demand for the new high performance non-volatile memory.

With this set of information, it seems possible that Tsinghua Unigroup could jointly build a fab (or fabs) with Micron in China to produce DRAM (and/or 3D NAND and maybe even XPoint). Because Micron would be negotiating from an incredible position of strength, if the two wanted to form a 50/50 joint venture (or something close to that), Tsinghua Unigroup would have to commit a significantly larger amount of capital to fund the new fab(s) than Micron would. Let's assume a new fab will cost $30 billion when this theoretical project begins in a few years. It's hard to know what a fair contribution would be from Tsinghua Unigroup for 50% interest in the joint venture. It might be a $20 billion cash contribution from the Chinese firm and $10 billion from Micron. It might be $30 billion cash from Tsinghua and $0 from Micron. It might even be as crazy as Tsinghua paying the entire $30 billion construction cost plus a hefty fee to Micron to buy into the joint venture. Suffice it to say, terms would have to be very favorable towards Micron in order for a deal to occur. I wrote an article on this site that explains how Tower Semiconductor has been able to breakout of the traditional foundry business model by acquiring capacity at way below market price. Micron could steal a page from Tower's playbook.

I do not believe there could be a simple deal that would be a sale of Micron's DRAM business to Tsinghua Unigroup (as government intervention is just too likely). I also do not believe a deal where Micron could provide its intellectual property and engineering prowess in exchange for a perpetual royalty from Tsinghua Unigroup is feasible either. Any deal that leads to a new competitor emerging is highly undesirable in the long run. It's also unlikely that Micron (or the rest of the DRAM industry) would like a Chinese firm to have complete discretion in the amount of DRAM bit output it put into the market each year. The profit motivated Micron, Samsung and SK Hynix may put themselves at risk of huge losses if they face a competitor who is increasing DRAM output for motivations that are not purely profit driven.

I don't believe a replacement cost analysis is the correct way to analyze the fair value of Micron, but it might be useful in giving a general idea of how valuable the firm is. John Pitzer of Credit Suisse suggested the replacement cost of Micron's fabs is over $26/share. Kipp Bedard's comments make me believe the DRAM business has a replacement value far higher than that. I imagine Micron's DRAM business alone has a replacement cost of well over $40 billion (based on Bedard's statements). To be fair, Pitzer estimated the replacement cost of just the fabs and implicitly stated the value of the intellectual property would make the true replacement cost more than the $26/share he estimated. Factoring in 3D NAND, 3D XPoint and in-progress development of future memory technologies, an all encompassing replacement cost of Micron is almost certain to exceed $50 billion and probably significantly more than that.

In the end, I do not think a merger will happen between Micron and Tsinghua Unigroup (or any other Chinese company). Frankly, I don't think a joint venture between Micron and Tsinghua is likely even with incredibly favorable deal terms. Micron's fate is either a successful standalone entity or a merger with Intel. The former scenario is more likely. Maybe I have lost my mind and objectivity when it comes to Micron, but I still firmly believe Micron investors will significantly more than double their money over the next few years.

Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.

Disclosure: I am/we are long MU. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.