"The Bull Market climbs a wall of worry." The old stock market adage seems equally applicable to Wall St. Analysts making estimates about Micron Technology Inc. (MU). Perhaps they don't remember what a profit making Micron looks like? In any case we will all remember on June 19, at 4:30 EST when the company discusses their third quarter which ended in May. This article discusses some underlying trends, some Wall St. sell side earnings estimates, and invites you to leave your own estimates-- and the rationale behind them-- in the comments below.
As SA's inestimable Russ Fischer has urged, Micron shareholders and option holders should check out DRAMexchange every day. It's a real luxury to be able to weigh in on a company's pricing on a daily basis and remove that variable from the earnings estimate equation. Here is a reprint of the spot and contract prices from a recent Citibank brokerage report:
With about 40% of Micron's revenue attributable to DRAM, this is a key chart. How much of the price rise do you think the company has captured. Bear in mind since this report was written, the DRAM prices have risen about another 10%. There were several times in the last 10 days this chip ran up more than 1% in a day. The larger 4Gb DRAM was up 2.11% on 6/10/2013 alone.
Here is a quote from a June 10, 2013 report from Wells Fargo which typifies the caution they are using with regard to pricing:
Memory pricing risk. Both DRAM and NAND prices have increase[d] substantially since their trough points in the September-November 2012 time frame. DRAM prices have approximately doubled and NAND prices risen by about 50% from their bottoms. At some point we expect memory price trends might level off and begin to fall, at least in line with the long-term trend of cost and price declines.
Well that sounds reasonable, but what is going to cause the prices to go down?
Despite this rather clear updraft and a market which is already tight, some analysts have downgraded Micron stock and/or insisted that average selling prices (ASPs) for these chips will certainly head down-- presumably because they always have. To me, this makes no sense: it takes two years and several billion dollars to build a new fab and nobody is doing so. There are a raft of new product introductions on the horizon which will consume more flash and DRAM in an already tight market: a new iPhone and a cheaper "entry level" iPhone, a wearable product from Apple, perhaps a new rev of the iPad and almost certainly a retina version of the mini, just to list rumored offerings for one manufacturer.
Most of the sell side firms present detailed industry supply/demand forecasts. In a recent Citibank report (May 27, 2013), here is their prediction for DRAM:
and here is their prediction for FLASH from the same report:
Remarkably the same report shows the ASP for DRAM as follows -18%, 8%, -15%. And the ASP for FLASH as -42%, -20%, -20%. These declines make no sense with supply not meeting demand, and compared to the spot and contract price rises in the first chart, above. Russ Fischer has suggested in comments to various articles that there is a disconnect between those who predict pricing at the bit level (Micron for instance) and those who predict pricing at the chip level, with the attendant bit level price declines due to geometry shrinks. In such a case, even an apparent price decline may result in increased company revenue. I have trouble justifying the ASPs most Wall Street analysts seem to be using.
One earnings estimate. Wells Fargo is one of the more bullish firms on the sell side and published a report on June 10, 2013 with a Q3 (May) estimate of $0.11, raised from $0.05. No firm seems to be offering guidance on what the Elpida transaction will contribute once it closes. They have raised their year estimates as follows: 2013 ($0.26), up from ($0.37), and 2014 $1.00, up from $0.84.
Wells Fargo has this summary on their estimate:
We estimate that Micron's contract price negotiations may have resulted in DRAM contract prices increasing about 26% through the May quarter, and NAND prices nearly 10% through the quarter.
Yahoo reports on the estimates of 30 analysts and says that consensus for the May quarter is $0.01. The consensus for the year is ($0.43) and for 2014 is $0.66.
My (guess)timate. For the May quarter, I believe standalone (pre-Elpida) Micron will report earnings in the range of $0.20 to $0.30 per share, prior to one-time and extraordinary events (last quarter included an asset sale and significant hedging losses). I am significantly more bullish than the consensus and the Wells Fargo estimate for the following reasons:
- I believe Micron has done a better job capturing the price rise
- losses reported on the equity line will be lower than last quarter since Innotera achieved profitability in the quarter.
- Micron supplies quarterly guidance on its calls. I believe they will have strained every nerve to do better than the bit growth of "mid to high single digits" they predicted for DRAM, and "up a couple of %" they predicted for FLASH.
Conclusions and Next Steps. As some analysts have predicted, this should be the quarter Micron enters the black. Put down your own earnings estimate in the comments below, and give some rationale. If you want to hazard a guess on what Elpida will mean on both a continuing and one-time basis, break those estimates out separately.
Tune in on June 19th for what promises to be a very interesting point in Micron's history!