All Quiet on the DRAM Front: Can Micron Survive the War of Attrition?

It's difficult to imagine that the memory market can stay in its present condition much longer, particularly after the blood baths (i.e. the negative earnings) that have been decimating the industry recently. Competitors in the DRAM and NAND flash markets have been engaged in what amounts to a war of attrition; a stalemate of sorts that will simply last until a few companies run out of resources. With the economy in a slump, it appears that the markets for memory will only weaken more over the next few quarters or longer, exacerbating things even further for companies barely holding on right now.

Yet this grim outlook is precisely the reason to keep an eye on Micron (MU). It's not that Micron looks to be on the verge of returning to profitability. Rather, it's that at least one of its competitors looks to be on the verge of succumbing. Qimonda's (QI) Balance Sheet and Cash Flows have looked extremely shaky to me for quite a while, particularly in comparison with a competitor like Micron that looks to be in much better shape to survive the storm.

Moreover, the most recent reports from the front suggest things are only getting worse for Qimonda as Seeking Alpha blogger Eric Savitz reports that Qimonda is heading towards insolvency. Meanwhile, the rumor is that Micron might purchase Qimonda and some analysts have downgraded MU based on fears of shareholder dilution from a potential deal. To top it off, analyst Daniel Berenbaum increased MU's loss estimate for the current year from $1.17 per share to $1.21 per share. Based on this negative outlook, Micron has been trading as low as $4 per share recently.

As the market continues to be somewhat sour on Micron, I am beginning to find it more and more intriguing. On one hand, it should suffer even more in the short-term as the economy takes a downturn. On the other hand, Micron might not be in such a bad shape to survive the Great Memory War. Has Micron now become a buy or is it still a sell? Before answering that, we need to take a look at the numbers.

The Balance Sheet and Cash Flows

Since we've already established that the memory sector is a business-world equivalent to World War I, let's look at Micron's liquidity before delving into Income Statement trends. Micron has a debt-to-value ratio of 39.8%. While not spectacular, it is better than Qimonda's 51.4% ratio. This figure does not provide all that much illumination, however, when you consider that there are companies in the memory market (such as Spansion) with debt-to-value ratios of 60% plus!

Micron has a working capital ratio ("Current Assets" over "Current Liabilities") of 2.31, which is actually fairly impressive to me given the fact that it has had six consecutive unprofitable quarters. Compare that to Qimonda's 1.06 ratio and you start to see how Micron is in significantly better shape.

I have a few personal measures I always like to apply to balance sheet accounts. One is what I call "Liquid" over "Current Liabilities." "Liquid" is my term for the types of liquid accounts that could be used to pay the bills (e.g. "Cash", "Receivables", and "Marketable Securities"). Applying this figure to Micron, I end up with a 1.45 ratio, significantly better than Qimonda's 0.64.

Finally, to account for current cash flows a little better, I like to look at "Liquid" plus "Cash Flows from Operations" and divide the result by "Current Liabilities." This results in a ratio of 1.88 for Micron and 0.52 for Qimonda. These last two measures were the primary basis for my belief that Qimonda was on its deathbed a few months ago. Micron is not in fantastic shape, but it can stand to benefit by simply surviving and it appears to be in a well enough position to do that.

The Income Statement

Unfortunately, while Micron's Balance Sheet looks amazingly good for a company that has been taking sizable losses for a significant period of time, those sizable losses are still rather frightening. Over the past two years, it has had dismal 2.1% sales growth and its gross margins have shrunk from a high of nearly 29% (Q1 of 2007) down to roughly nil!

Starting with Q4 of 2007 and moving forward to Q3 of 2008, Micron's past four quarters have consisted of per share losses of $0.21, $0.34, $1.01, and $0.31. The $1.01 looks particularly frightening on the surface, but actually includes a 60 cent goodwill impairment charge. All the same, Micron has to return to profitability in order to survive, so we never like to see such ugly numbers with few signs pointing towards improvement. However, it at least appears to be doing much better than Qimonda whose most recent quarters have included per-share losses of $1.75, $1.41, and $1.17.

Forecasting a Possible Future?

I do not necessarily feel like I have enough expertise in the memory sector to create an accurate forecast for Micron over the next few years. Moreover, there are so many industry and general economic variables that could (and probably will) change, it becomes an almost meaningless exercise. Instead, I decided to use a different approach called "The Ideal Quarter." Using this method, I simply attempt to forecast Micron's profitability in a quarter based on numerous different scenarios. With that information, I try to extrapolate what might be a fair price for the stock based on these scenarios.

The goal here is not to create a valuation for Micron. Rather, what we are attempting to do is understand how attractive Micron might appear given certain improvements in performance. We already know if Micron continues to be unprofitable that it's not a stock to buy, so there's no reason to delve into worst-case scenarios for this exercise.

The two main variables I change around in these scenarios are "Net Sales" and "Gross Margin." I decide to use a steady SG&A charge of $115 million and R&D of $170 million in each scenario, which seems appropriate to me since these costs appear to be more fixed than variable.

Scenario #1: Sales Revenue of $1.6 Billion

In this scenario, I assume sales revenue reaches $1.6 billion. Using gross margins of 5%, 10%, 15%, 20%, and 25%, I get the following earnings for our ideal quarter. All figures in parentheses indicate a loss:

• 5% = ($0.24)
• 10% = (0.13)
• 15% = (0.03)
• 20% = 0.07
• 25% = 0.18

Scenario #2: Sales Revenue of $1.75 Billion

• 5% = (0.23)
• 10% = (0.12)
• 15% = Break Even
• 20% = 0.11
• 25% = 0.22

Scenario #3: Sales Revenue $1.9 Billion

• 5% = (0.22)
• 10% = (0.10)
• 15% = 0.03
• 20% = 0.15
• 25% = 0.27

These three scenarios don't particularly give me great reason for optimism with Micron. It would appear that Micron would need to produce margins in the 20%+ range to become a worthwhile investment. Considering its gross margin in the most previous quarter was 3.2%, that seems like a far ways off!

However, if things were to radically improve in the memory markets, we can see how this stock would shoot upwards. In the last scenario with $1.9B in revenues and a 25% margin, Micron would seem to have a yearly EPS around $1.08. If we assume a Price-to-Earnings (P/E) ratio of about 10 for the stock, that means that it could conceivably be trading in the $10-11 range, a nice 150-175% increase over the current $4 price!

Yet, take a look at Scenario #2 and even with a 20% margin, we would only have a yearly EPS of roughly $0.44, which might conceivably produce a $4 to $5 stock price --- which is precisely where it's trading right now. From this, it becomes easy to see why buying into Micron is essentially betting on high margins again at some point in the future.

Conclusions

Even if we get mixed signals from our analysis, I would still keep an eye on Micron. At some point this stock becomes a tempting buy due to the potential for high rewards. If Micron does end up acquiring Qimonda at "fire sale prices", it's not completely unforeseeable that it could spring back at some point in the next two or three years with both significantly higher revenues and significantly higher margins. Just don't expect any miracles in the short-term. This is a long-term buy and hold until the market (and consumer spending) picks back up.

What is the right price for it right now? It is a tempting buy at $4 but due to the recessionary environment and almost certain losses that will continue to pile up over the next few quarters (possibly extending till the end of 2009 or longer!), it might see even further price erosion yet! If it dips down into the $3.00 to 3.50 territory, it might start to look very attractive. Mind you, I would not buy this stock and bet the whole bank on it, but it might be worth a very worthwhile gamble in a high-growth (and hence, high-risk) portfolio, even at $4.

Based on a 3-year holding period, I would give this stock an upside potential in the $15 range and a short-term downside risk of about $2. Given the nature of the memory market, I would take into consideration the fact that Micron might be playing the role of Germany in this Great War and give it a long-term downside risk of $0.

However, I see its prospects more in line with the Allied side: it'll survive the "war of attrition" and things improve some, but it does not win a dramatic and decisive victory. Based on that, I'll give it a 3-year target price of $6.

Stock position: None.

Comments

[Rolv E. Heg...:]

Your article neglects to mention that Micron is currently in a suit against Rambus where, if found to infringe Rambus' patents, Micron could be ordered to pay up to $1b in back-damages and a royalty of 3-10% going forward. There's another suit coming in January where Micron is sued for Antitrust together with Samsung, and Hynix - the latter two companies have pleaded guilty of antitrust and paid 100's of millions in fines to the DOJ. Micron seems to have been the ring-leader which, if proven, would nullify their snitch-agreement with the DOJ.
I also don't think Micron has much of a chance with a CEO that has been voted the worst CEO in the country for many years running and who has been in charge while Micron has lost billions of dollars.

2008 Oct 02 09:11 AM

[mbr:]

Good analysis except off balance sheet items such as lawsuits are always to be considered. Also the human factor in leadership, but, apart from Micron's results, what can be said about its CEO? Could you answer Rolv E.?

2008 Oct 03 04:04 AM

[rd4sndk:]

I can not believe someone is trying to put lipstick on this pig of an investment. You should try a list of worstening scenarios that suggest under what conditions would micron become insolvent. As I see it, smart money will start to exit micron no later than 1Q09 and that micron will receive a bit somewhere in the 2Q09. At a current market cap of $3.4B (4.4 pps), it's p/b at 50% would suggest that with a little further deteriation in its share price, one can pick up at least two 300mm mfg plants for the fire sale price of $4 Billion. Throw in the other operations and a hefty 25% layoff after acquisition micron is ripe for the picken. Buying micron would benefit Samsung or Hynix well as all two are in the 3 basic businesses that Micron operates. Sandisk would also benefit in buying Micron because they would eliminate another major nand competitor while improving its royalty stream, and more importantly give Sandisk an upper edge in removing Intel from the nand business. If Sandisk were to buy Micron, this could usher in a greater collaboration between Intel and Sandisk and would go further to solidify Sandisk/Toshiba partnership as the defacto standard for nand chips and cards.

2008 Oct 04 10:09 AM

[H.J. Huneycutt:]

I don't really disagree with any of the comments about the risks involved with Micron --- which is why I explicitly stated this is not something you want to bet the bank on. It's a potential high-reward stock that should be looked at for a high-risk portfolio.

Also to rd4sndk, I totally agree on SanDisk. They have a stronger balance sheet than just about anyone in the industry.

2008 Oct 06 09:53 AM

[anand:]

Can anyone talk about what the future of flash post MLC will be (only 1-2 year away)? Will SPSN's patent portfolio give it an edge? Many say that INTC/MU are well positioned compared to Toshiba/SNDK and Samsung.

2008 Oct 16 11:54 PM

[anand:]

Of course SNDK/Toshiba might soon become SNDK/Samsung ;-)

2008 Oct 16 11:54 PM