3 More Reasons To Stay Long AMD

| About: Advanced Micro (AMD)


Market capitalization is well below 1 times annual revenue.

Margins are likely to improve.

Gaming consoles to ramp in China in 2015.

Advanced Micro Devices (NYSE:AMD) is a fabless semiconductor company focused on desktop and notebook PCs, servers, embedded solutions, and graphics. During the last two years it has changed its strategy, notably securing a near monopoly in supplying CPU/GPU chips to the game console market. AMD revenues increased in 2013 despite a decline in the global PC market.

Based on revenues, AMD's share price is low compared to other major semiconductor makers, including rivals Nvidia (NASDAQ:NVDA) and Intel (NASDAQ:INTC). This article will examine the Price to Sales (P/S) ratio, (which equals market capitalization to trailing annual revenue), whether that ratio is likely to rise, and the possible resulting increase in stock price.

I. Price to Sales Ratio

The first table shows how AMD's market capitalization to annual revenue ratio (which equals P/S) has varied for the years 2010 through 2013:

Year Ending Price Market Cap
2010 $8.18 $6,233 $6,494 0.96
2011 $5.40 $4,115 $6,568 0.63
2012 $2.40 $1,829 $5,422 0.34
2013 $3.87 $2,949 $5,299 0.56
Click to enlarge

The P/S ratio was near 1 at the end of 2010, then it declined steeply to 0.34 at the end of 2012. At the end of 2013 it had risen back up to 0.56.

P/S varies greatly by industry and within industries. If earnings represent 5% of revenue, and a stock has a P/E of 20 (representing modest growth in the current low-interest rate environment), it will have a P/S of 1. P/S is particularly useful when a startup company has ramped sales but has not reached profitability, as it shows how much investors are willing to pay for revenue.

Now compare the ratio to a few companies in the semiconductor space:

At market close on June 20, 2014, using 2013 revenue
  Price Market Cap Annual Revenue
AMD $4.10 $3,124 $5,299 P/S
Intel $30.20 $150,355 $52,708 2.85
Marvell $14.83 $7,534 $3,404 2.21
Nvidia $18.93 $10,562 $4,130 2.56
Qualcomm $79.86 $134,793 $24,866 5.42
Click to enlarge

Intel is an AMD rival in the x86 computer space. Nvidia is a rival in discrete graphics chips. Marvell (NASDAQ:MRVL) is a rival in tablet computers, but gets about half of its revenue from storage chip controllers; it will be a rival in ARM-based servers. Qualcomm (NASDAQ:QCOM) is not a direct rival, but is the leader in the cell phone and ARM space.

AMD is doing terribly in P/S. Intel, Marvell, and Nvidia are clumped in the 2 to 3 range, while Qualcomm stands out at 5.42. It would appear that semiconductor companies with P/S below 2 have some explaining to do.

Market capitalization, which is the stock price times the number of shares (either outstanding or diluted), encapsulates many factors, but the prime factor is earnings. If two companies have the same revenue, but one has higher margins and resulting earnings, that will likely be reflected in the stock price, which results in a higher P/S.

Marvell and AMD illustrate this. AMD had substantially higher revenues in 2013 than Marvell, $5.3 billion versus $3.4 billion. But Marvell made more profit on its revenue, and so has a much higher market capitalization and P/S.

If AMD could bring its P/S up to what looks (again, rule of thumb) like the bottom of the more typical range for semiconductors, 2.0, its stock price would rise to $13.67 per share. Even lifting it to 1, which is slightly above where it was in 2010, the stock price would be at $6.95.

AMD revenue looks to be considerably higher in 2014 than 2013. Q1 2014 revenue was $1.4 billion, up 28% from Q1 2013. If revenue rises and P/S stays flat, that would still raise the price per share.

But, of course, there are reasons for AMD's low margins and low market capitalization compared to revenues. A CFO can't just waive an accounting wand and change the ratio.

The revenue still makes a juicy target. If AMD can work up to the kind of margins and P/S Marvell or Intel get, the stock should be worth a lot more to investors.

II. Margins are Likely to Improve

Why are AMD margins so low, and how could they be raised? The answers have to do with history, technology, competition, and strategic decision making.

Since near the beginning of the PC revolution, AMD's key competitor has been Intel. Other competitors that made CPU chips for PCs, like Motorola, Via, and IBM, dropped out. AMD had some moments of glory, as when it introduced Opteron server chips, but over time Intel ground AMD down on multiple fronts.

Since AMD acquired ATI, it has also competed head-to-head with the leading GPU maker, Nvidia. AMD had a good theory about the ATI acquisition: it would put both the CPU and the GPU on the same die, thus getting out in front of both competitors. That did not succeed as a strategy largely because AMD executed slowly and was always behind Intel in process technology (the amount of circuitry that could be put in a single chip). Intel did a fair enough job putting GPUs on dies with CPUs to brunt the AMD effort.

Low margins at AMD came from higher per-chip production costs than Intel, lower sales prices of chips to OEMs due to perceived inferiority, and operating costs spread out over a much smaller revenue base.

Margins are likely to improve going forward largely because AMD is competing less directly with Nvidia and Intel. The most important example is the AMD chips for gaming consoles. There is no competition from Nvidia or Intel there. The same should be true as AMD wins embedded and specialty chip designs. This should allow for better pricing, which is a key margin mover.

An important aspect of margin improvement will be in sales and marketing spend. Having won the game console slots, AMD does not have to do any kind of sales spend on those units. This makes it more like Marvell, which sells to OEMs, but does not need to spend money on direct-to-consumer advertising.

Because AMD sales are now seasonal, higher and Q3 and Q4 than in Q1 and Q2, I will use Q4 2013 as a reference point for margins. Note that margins will be worse in Q1 and Q2 as roughly the same expenses get spread over seasonally slower revenues.

Using AMD's published GAAP figures (which included non-cash expenses):

Margins, Q4 2013
Revenue $1,590  
COGS $1,036 65.2%
gross margin $553 34.8%
R&D $293 18.4%
MG&A $169 10.6%
Amortization $4 0.3%
Legal -$48 -3.0%
operating margin $135 8.5%
interest & other $46 2.9%
taxes $1 0.1%
net income $89 5.6%
Click to enlarge

The biggest impact on margins is from COGS (Cost of Goods Sold). This basically represents payments to foundries. It could improve as yields improve within a process node, but it can also decline when a new node is introduced. It is highly dependent on unit volume. COGS leaves gross margin near 35%. If revenue continues to ramp, either because of unit growth or average price increases, this margin will improve.

Operating expenses are mainly research and development (R&D) and marketing, general and administrative (MG&A). Here is where improving unit volume and revenue can result in better margins. These costs should be able to cover higher unit sales. Then GAAP operating margin can rise. But note in Q4 there was a $48 million legal expense benefit, which will not be repeated (and is excluded from non-GAAP numbers).

AMD's debt level is about $2 billion, and the interest rates it pays are high. The net income margin could also increase if the debt could be paid down, or refinanced at lower interest rates. Before addressing that, there is one driver coming in 2015 that is highly likely to increase unit sales without increasing operating costs.

III. China legalizes gaming console sales

China has legalized the sales of gaming consoles. Both Sony and Microsoft have announced plans to introduce their consoles into China. Each console contains an AMD APU chip. This should increase revenue without increasing AMD's R&D or marketing costs. It may be the main driver of AMD's stock price in 2015. But it is anyone's guess how many gaming consoles will sell in China. Given the spending proclivities and love of technology in China, I would be willing to bet on the word "significant."

IV. Debt, Interest, and Cash Flow

In Q4 2013 AMD's cash flow from operations was only $21 million, and free cash flow was zero. Cash flow from operations was negative $204 million in Q1 2014, largely because AMD made a large payment to GlobalFoundries to finish paying for a 2012 purchase agreement. Interest payments were $47 million on long-term debt of $2.0 billion.

AMD does have about $1 billion in cash. If AMD can generate free cash flow and pay down its debt, it could eventually take as much as $47 million per quarter off its expenses. That would result in a significant earnings increase, but it would likely be a multi-year process.

V. Conclusion

AMD could see a significant increase in its stock price if its P/S ratio moves higher, especially if it approaches what rivals achieve. That in turn would require a continued revenue ramp combined with better margins. While there are several places where margins could be improved, in 2015 the best bet for increased margins is through sales of gaming consoles in China.

AMD will report Q2 results on July 17. This should give investors better visibility as to sales and margin trends this year, even though it is typically a seasonally-slow quarter.

Disclosure: The author is long AMD, MRVL. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.