A year and a half ago, I wrote an article on how the DRAM glut would long term help Micron (NASDAQ:MU). At that point, the stock was trading at about $22 and I believed it would jump to back over its previous high of $35. Just a short 18 months later we watched Micron drop from its all-time high excluding the tech bubble in the year 2000, and “settle” in at around $57. All the news about Micron is positive, so what's causing the stock to correct and is it done correcting?
To start we will look at the most recent news as of May 31st. According to a statement released by the company, they are being investigated by the Chinese government for what the Chinese media believes to be over inflated chip prices. These visits also took place for Micron’s two biggest rivals, SK Hynix and Samsung (OTC:SSNLF). During this precarious time of trade disputes, any action by the Chinese government could put a large dent in Micron’s earnings as China accounts for over half the company’ sales.
While this had the potential to be relevant, it is not the real reason Micron is, and should be, trading down. The biggest reason for the company to trade below its current price is the fact that its ability to earn a profit is directly related to market factors that they don’t control. We can see this historically as during the Samsung-led DRAM glut in 2016, the company had margins of about 1%. In their current state, Micron has operating margins of nearly 50% with their heavily DRAM focused segments topping 60%.
Before I get into my models and valuations, there are a few background ideas that need to be established for it all to make sense. The first is that the company does not release gross margin for its individual products. The only margin information the company releases is operating margin which means deciphering R&D is more guesswork than it is explicitly stated. This means that in order to properly forecast my model by segment, I took expected operating margin and worked backward. I chose this method because different segments have different growth potential and revenue potential and thus forecasting them individually gave a more accurate overall forecast.
The next important concept is that despite what people think, the revenue growth for the company is almost completely driven by margin growth. Between 2016 and 2017, revenue rose 64%, while operating margin rose 71%. This implies that the majority of the growth came from increased margins as opposed to volume growth. While volume growth is still a factor, as can be seen from the 10-Q, the increase in selling price for their products has had a larger impact than their increase of gigabits sold. For the company to continue high income growth, which is the basis of the bullish sentiment, they would either need margins to increase significantly again, which is unsustainable for consumers, or volume growth to increase significantly, which would be hard to keep up with from a manufacturing standpoint as wafer fabrication plants are slow to build.
Before the correction, I created a few models for Micron. One was a bull scenario, one was a bear scenario, and the last was a target scenario. After creating these, it became incredibly evident why Micron should correct. To get a good idea of the value of the firm, we must take all scenarios into account and assign our own weightings to them to find a fair price.
First, I will describe the bull scenario which is the basis behind the incredible rise Micron has experienced this year. The bull scenario takes two major concepts into account. The first is that DRAM prices remain inflated, and second, that NAND prices begin to inflate as demand grows. As stated before, I used operating margin by segment to forecast earnings and thus I kept Micron’s CNBU unit, the most heavily DRAM dependent and the largest of the four, at a 63% operating margin over time. Next, I made the MBU (Mobile Business Unit) and EBU (Embedded Business Unit) units rising to 54% from 44% and 51% from 44% operating margin respectively by 2022. These are because the MBU is dependent highly on DRAM but with a solid portion of its revenue coming from NAND related products, thus an increase due to the increase in NAND prices. The EBU is highly composed of NAND-based products which in this scenario I have rising to 51% operating margin. Using these values and different weightings of the units’ CAGRs (compound annual growth rate) I came to a valuation of $78.
The second model I created was a bear model. This model took the negative possibilities into account. I didn’t want to make it extreme as this would inaccurately exaggerate the downside, but even historically reasonable downturns in market conditions had a huge affect on stock price. Historically, between 2013 and 2017, Micron had an average operating margin of just 14%. Yet, in my bear model I created a scenario where the company has operating margins drop to 30% and the valuation still plummets. As we saw from the 2016 DRAM glut and in 2013 when they struggled to make a profit as seen below from Micron’s 2014 10K, this situation is very possible. In the situation of operating margins dropping to 30%, the company’s valuation drops to just $32.5 per share.
The last model I created was the target model. This took what I expected to happen into account. In this model I maintained significant growth in their MBU and CNBU, their two most successful units with room to grow, and had margins drop only slightly as the high earnings potential is likely to cause supply growth from other companies such as Samsung and SK Hynix. It also will likely cause increases in supply and volume sales for Micron which should largely offset the decrease in margins but will likely stunt growth. At operating margins still in the mid to high 40’s% range, the company is still in favorable market conditions, just not the perfect storm it finds itself in now. This model gave a share price of $53.58, about 7.5% under current share price.
While I find the bull [note: an earlier version of this article mistakenly used the word "bear" here] and the target scenarios to be the most likely to occur with promise for 3D-XPoint NAND technology to become a major factor for Artificial Intelligence and with the MBU seeing an increase in demand as smart phones incorporate more memory, the downside risk is too large and too possible for the inflated valuations seen in the last month.
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.
Additional disclosure: I am long short-term call options but still have a long-term bearish outlook.