Seeking Alpha
Value, long/short equity, special situations
Profile| Send Message|
( followers)  

Earnings season is coming up for Micron (NASDAQ:MU) which is never a pretty moment given the poor guidance they provide, and the dreadful transparency of their financial statements. I think I will stick by the $3.50 per share for the year, which Kipp Bedard, VP of IR, guided, and just take 1/4 of that and call it my 1Q 2014 estimate -- $0.87. If Kipp said it, it must be true as we all know.

Actually, I think it's instructive to look at two of the best Wall St. analysts side by side and point out what they are up against, and what they feel the upside and downside of their estimates may be. Joe Moore and his team at Morgan Stanley, and Mark Newman et al at Bernstein Research both do a good job, especially given the opaque Micron management. They not only put down their number, they each back it up with a lot of graphs, charts, and underlying assumptions. You may not agree with them, but you know where their numbers are coming from.

Here's a snapshot of the recent estimates from the two firms:

Micron Technology Estimates
Morgan Bernstein
Stanley Research
Oct 11, 2013 Nov 5, 2013
1Q 2014 FY 2014 1Q 2014 FY 2014
Revenue 3,823 14,309 3,861 16,206
COGs 2,323 9,494 2,640 10,835
Margin 1,500 4,815 1,221 5,371
GM % 39% 34% 32% 33%
Non GAAP net 812 2,321 852 3,410
Shares Out 1,128 1,132 1,151 1,184
Non GAAP eps 0.72 2.05 0.74 2.88
GAAP eps 0.01 1.06 0.54 2.43

A couple of points pop out of the side by side comparison:

  • there is a HUGE disparity between GAAP and non-GAAP estimates. Kipp, when you are writing an earnings release spell out the difference and don't bury it back a few pages. When news wires confuse your poorly written releases, have someone contact them and get them changed. You might start with the last quarter's numbers.
  • the two firms are quite close for the upcoming quarter but are considerably apart on FY Revenues and margins. More on this below.
  • the shares outstanding are widely different. How secret is our projected share count? Can't we at least get guidance on that? (nb, the Morgan Stanley numbers had to be derived from eps and non GAAP net income). I'd once again refer readers -- especially you Kipp and Ron Foster -- to the Sandisk (NASDAQ:SNDK) IR web page. There you can find handy PDF's to reconcile GAAP and non-GAAP earnings, and for a Convertible notes dilution calculator.

What color does Morgan Stanley provide? MS has three price targets for three cases. Surely one must be about right! Writing their report when the stock was at $17.27, they had a bull case PT of $27, a base case PT of $19, and a bear case PT of $9.

Here was the overall summary:

Risks to our view:

We assume this is a normal cycle overlaid with modest structural improvement and certain one-time event such as the Hynix fire. Could we be missing more sustainable improvement? We are watching for behavioral changes

And they had this to add about their bullish case:

Very strong memory cycle. DRAM remains in an allocation mode in 2014as slower supply growth and strong demand in mobile and server of weaker PC demand. NAND is slightly better than our base case scenario, which is already bullish; Micron achieves new highs in gross margin by Company behavioral changes make investors believe there are sustainable profits to be had here; companies leave fab utilization below 100% even during relatively profitable periods, slow capacity shrinks further, make no capacity announcements. Elpida goes smoothly with very capital spending requirements, and Micron is able to seamlessly migrate some capacity from DRAM to NAND to improve the business mix.

What color does Bernstein provide? The Bernstein report cited is a general update on the memory industry. As such it has some excellent tables and charts on industry wide supply and demand for both DRAM and NAND. MS does these too, just not in the brief Micron update cited.

Bernstein has a bullish view on what is happening to DRAM pricing, across the industry:

We don't believe concerns that after Wuxi ramps back, DRAM will go back to oversupply. Even with Wuxi back to full output by Q2'14, we still show DRAM capacity down vs. Q2'13. Borrowing of NAND capacity to reduce DRAM shortfall is only temporary.

And they have this to say about Micron:

We rate Micron Outperform with a target price of $24.0. We believe that DRAM should keep recovering in 2013 following Micron's acquisition of Elpida and subsequent supply rationalization. The NAND rebound we also expect to continue through 2013.

A bone to pick. Bernstein and others continue to use price to book as a metric for valuing Micron:

Micron Technology Inc.,

We use both P/E and P/B to derive our $24 TP, which is equivalent to 8x CY 2014 Non-GAAP EPS and 2.1x one year forward BVPS.

Surely they realize there have been more than $4billion in impairment charges and write-offs over the past two years at Elpida and the purchase accounting has skewed this number. Again, Micron management should be working to educate shareholders and analysts on the implications of the Elpida purchase accounting. They are not.

Conclusions. One of the discrepancies between these two analysts cited above is their assumptions on Gross Margin. While MS has a higher margin for Q1 '14, their margin assumption tapers down to 34% for the full year-- approximately where Bernstein is.

With the clarity of a crystal ball I don't possess, Bernstein posits the following margins for industry players in 4Q 2017:

Sandisk 49.6%
Samsung (OTC:SSNLF) 55.9% 47.0%
Toshiba (OTCPK:TOSBF) 45.5%
Hynix (OTC:HXSCF) 53.6% 37.2%
Inotera 45.2%
Micron 42.9% 36.8%

This slow steady upward projection for the entire industry is really at the core of Bernstein's thesis that it is indeed different this time, and that a New Paradigm is emerging in memory.

But I wonder, and hope analysts will wonder out loud at Micron investor presentations, why Micron is at the bottom of the class in terms of Gross Margin? Didn't we just buy a very modern and up to date Elpida at a fire sale price? Aren't we the guys with the largest geographic spread of fabs with the attendant ability to shut an entire fab for a process node improvement or switch to 3D?

Me, I'm gonna stick with whatever Kipp tells me. Kipp tells me we are gonna earn $3.50 for a whole year and I know a quarter of that is around $0.87 for 1Q 2014. Wait, Kipp, was that GAAP or non-GAAP? and how many converts will you have bought in by then? And will you have liquidated any of your capped calls, or are you going to dilute us? Kipp, ohhh Kipp ...

Source: Micron Earnings Estimates