Micron (MU) has been working to acquire Elpida, the last remaining Japanese DRAM memory manufacturer, out of bankruptcy since last year. Court approvals are in place in Japan and the US and the expiration of a clock on appeals and the issuance of final court orders implementing the transaction are expected any day. Little information has been issued on the current results of Elpida. The Japanese bankruptcy process is not as transparent as that in the US. Micron itself has been muzzled by the process but probably hasn't wanted to attract attention to Elpida's strong results since it has faced competing reorganization plans from Elpida bondholders. A bit of digging and a bit of math indicate Elpida will indeed be accretive to Micron from the moment the transaction closes.
Sell-side brokerage analysts writing on Micron do a hand wave on the Elpida acquisition. Drexel Hamilton wrote, on March 5,
- Japan's final DRAM survivor, Elpida, apparently destined to fall into Micron's lap for $2.5bn at present rates of exchange looks to soon shift its roughly $750mm [author: report says "bn"!] in Sep 2012 quarterly revenues to the U.S. concern."
And Wells Fargo wrote, on March 15,
Our new estimates do not incorporate any potential effect from the completion of Micron's pending acquisition of Elpida, which we think could result in a one-time gain in the quarter in which the deal closes and be accretive to EPS in subsequent quarters.
Bernstein Research came closest to quantifying some of the value that Elpida will bring to Micron when it wrote, on March 22,
We believe that upon closing of the deal in 1H'CY13, Micron will benefit significantly from a lower Elpida PP&E appraisal (even lower than the latest $1.3B number disclosed concurrently with the convertible note offering in February) and hence lower depreciation. We believe our model currently does not fully capture this potential upside yet.
Once Elpida gets fully consolidated (we currently assume Micron fully consolidates Elpida's bits starting in FQ4'13), the acquisition should represent a ~45% increase in Micron's total trade memory capacity (equivalent to about double Micron's current capacity in DRAM), but only a 15% increase in depreciation costs. This is based on Yen/USD exchange rates from November time frame. We currently forecast overall DRAM bits sold to more than double in FQ4'13 versus FQ3'13, so depreciation/bits should come down significantly.
Admittedly, getting information hasn't been easy. Last fall Judge Sontchi, the US Judge on this Japanese bankruptcy said, "I don't have a problem tanking a case," expressing his concern over the lack of transparency on the whole Elpida process and their results. But there are some clues if you dig. On February 5, Micron filed a Form 8-K with the SEC which shows a pro forma roll up of Elpida into Micron, using the September 2012 results of Elpida.
Here is a condensed summary of the 8K proforma:
Below the gross margin, the 8-K details Elpida actuals and proforma adjustments totaling 182 million in expenses to get to a Net Income attributable to Parent of $22 million. Dividing that number by the 1.017 billion shares forecast to be outstanding, one gets $0.02 per share of earnings attributable to Elpida had the transaction closed as of the 9/12 proforma date. Note: the proforma doesn't deal with the differing year ends of Elpida and Micron and merely "staples" Elpida's 9/12 quarterly to Microns 11/12 quarter. Since the purpose of this proforma is to show how Elpida would have looked as part of Micron as of a certain date, hold your nose, ignore the differing dates, and be happy that Elpida would have been accretive as of the date[s] of the proforma.
Sensitivity Analysis: Of course, readers of MU articles on Seeking Alpha are aware that the prices of DRAM have more than doubled since the beginning of Micron's fiscal 2Q. These prices are not reflected in the proforma above. The forecast suggests that at the time of the earnings call on 3/21/2013, management had already experienced a 5% increase in DRAM ASP as of the Quarter to Date. For every additional $11 million in Elpida revenue, Micron would realize an additional penny per share.
Something interesting is happening in Depreciation: Note D to the proforma in the 8-K details the depreciation effect that Bernstein was alluding to in its recent research report. Note D shows $188 million in eliminated depreciation during the 9/12 quarter, due to "push down" accounting and the lowering of Elpida's PP&E. But it also shows $1010 million in eliminated depreciation for the proforma year detailed elsewhere in the 8-K. The fact that the most recent quarter is less than 1/4 of the year shown indicates that depreciation is rolling off quickly - even without the pushdown - and that margins should get better due to this fact alone.
Conclusion. Elpida was profitable down to the proforma net income line at 9/12, when Micron was still printing massive amounts of red ink due to terrible DRAM pricing. Elpida participates in the higher margin mobile memory sector. Make your own conclusions about how much of the huge spot market increase may have rolled into Elpida's results since 9/2012. An already accretive acquisition is becoming more so.
After this article was originally written, the company announced the Elpida acquisition was being appealed in the Tokyo District Court. While this will delay the full operational integration of Elpida, and could end up costing the company more in acquisition price and certainly in legal fees, Elpida's results seem to be getting better: further roll-off in depreciation? further price increases in DRAM and firmness in mobile memory? Some prognosticators feel further weakness in the Yen may lower the ultimate purchase price if the acquisition closes later.
Disclosure: I am 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.