Elpida - What happens next?

 

Elpida filed for bankruptcy in Japan today. The phoney war is coming to a close & the real action is about to start.

The first thing to say is that Elpida's business model has been unsustainable ever since PC sales & their memory content began to stall. Once DRAM embarked on its current down cycle in Q2 2010, it has increasingly revealed just how Elpida have consistently overpaid for assets & undersold their product and technology.



I also think their early leadership in mobile DRAM concealed failings and inaction elsewhere in the business. No doubt its profitability here subsidised other Elpida customers and channels that delivered plenty in terms of volume but little in the way of value.




Investors will be hoping to see less of this in future

Today's filing for bankruptcy protection (two weeks before they needed to make a decision) is obviously a negotiation ploy by Elpida but regardless their options remain limited. The Japanese government are unwilling or unable to bail them out.


Most of the other Japanese semiconductor companies proposed as partners are bigger basket cases than Elpida themselves. The most logical deal has always been a tie up with Micron and Nanya (assuming they can bring Formosa along).

That said though, logic is often trumped by politics and personality. But if we look at the facts: 1. Micron is the industry's serial consolidator. 2. A combined entity would vault Hynix and become a force to be reckoned with by Samsung. 3. They would drive the DRAM industry into a 3 party game & follow the HDD guys into a much stronger and more stable future.

On the politics side, could "Japan" accept selling their company to "America"? And on the personality side, will Elpida's executive management push for a deal where they will be quickly sidelined?

The easy out would be to get Toshiba on board somehow. Elpida (or at least most of it) would remain Japanese and their key executives would stand a better chance of surviving in the new entity. But while Toshiba are no basket case, they don't exactly stir that much fear amongst their competitors. 

My guess is that everybody in Seoul would sleep much easier at night if they felt Elpida's centre of gravity remained in Tokyo, a didn't move to Boise.

It's structural, stupid

There is also the possibility that Hynix could get involved. SK Telecom have certainly got the capability and ambition. Also they have most to lose if Micron did the deal. That said, the politics of Japanese-Korean deal makes the politics of a Micron deal look like a piece of cake.
I gueess, at the very least if Hynix did get involved, they could get a look at Elpida's books and drive up the price for everyone else.

But whatever way this plays out, 2012 is still going to be the year the DRAM industry steps into mature middle age. What we have been witnessing since Q2 2010 is not just another cyclical phase, but a fundamental structural change.

Things are never going back to the way they were for DRAM. The industry is already adapting to this but many companies, models and people will struggle to do likewise. Jack Welch said "change before you have to", Elpida failed to do so but they won't be the last.

Please feel free to post your comments below.

If you want to discuss anything in more detail you can email me at memoryman@notsodimm.com

Next Post: The impact of DRAM consolidation on the module makers.