Advanced Micro Devices (AMD) is a rather uniquely positioned fabless chip designer. Through its acquisition of ATI Technologies, it is one of the two major vendors of integrated and discrete graphics processing units (GPUs). It is also the last remaining significant player in the x86 processor market other than the dominant Intel (INTC). I believe AMD is unfairly valued here and that there exists significant upside potential in the second half of the year.
First and foremost, I would like to address the company's recent earnings. A quick glance at the company's most recent earnings reports shows that the company has seen a loss in the last two quarters. However, these losses were due entirely to one time charges. In Q4 2011, the company lost $209M due to an impairment in their investment in Global Foundries. They incurred an additional charge of $98M due to restructuring charges. This led to a GAAP net loss of $177M, but excluding charges the company managed to beat analyst estimates and earned $0.19/share, versus consensus of $0.16/share.
Then, in Q1 2012, AMD again swung to a much more painful net loss of $590M. However, this was after taking a $703M charge (including divesting itself of its remaining 8.8% stake in Global Foundries) in order to get out of an exclusivity agreement for the manufacture of AMD's 28nm CPUs/APUs going forward. Excluding these charges, AMD beat the analyst estimates again, posting a $0.12/share profit versus $0.09/share consensus.
The financial yoke of Global Foundries is not quite released, even after these heavy charges -- AMD still owes the chip manufacturer $275M that will be paid over the year. Until these one time charges are completely resolved, it is advisable that investors evaluating AMD as a potential investment concern themselves with the non-GAAP numbers that exclude these charges.
Now, suppose that we look at the price to earnings ratio for the trailing twelve months and use the non-GAAP numbers for Q4 2011 and Q1 2012. This gives us an EPS of $0.55/share, which as of the last closing price of $5.73, yields a P/E of 10.41. Forward P/E for the fiscal year ending on December 31, 2013 sits at a mere 6.59. I feel that this is extremely cheap. Why?
It's no secret that any company that's very entrenched in the traditional PC market (which is widely viewed as contracting in the face of smartphones and tablets) is seeing low valuations. Hewlett Packard (HPQ) is trading at 7.89 times past earnings and 4.61 times forward earnings. Dell (DELL) is trading at just 6.97 times past earnings and 6.01 times forward earnings. Even the mighty Intel is trading at just over 11x past earnings. I believe, however, that AMD has significant potential for growth, especially given its size (and therefore its agility).
AMD is very tactfully moving away from fighting the high end CPU performance and process technology battle with Intel -- it's very, very difficult to compete with Intel there, especially as a fabless design house. However, instead of fighting this losing battle, AMD instead seems to be leveraging its significant graphics expertise as well as its position as the only other major x86 CPU designer to produce compelling products at competitive price points in an attempt to grow its market share.
In fact, it seems that the "APU" (CPU + GPU on one die) chips from AMD are helping it take share from Intel: in Q1, AMD's market share in the x86 processor market rose to 19.1% up from 18.2% on a year on year basis. Even if all of the "doom and gloom" of the PC industry growing slowly is true, AMD still has plenty of opportunity to grab more of the existing market from rival Intel. In fact, let us examine AMD's market share situation more closely.
Breaking it down, we see that AMD actually owns 43% of the desktop, so I wouldn't expect too much more growth there. On the notebook side, AMD holds 16% of the market. The APU offerings are significantly more attractive to OEMs due to the strong integrated graphics processor, so I suspect that AMD will be able to take more share in this segment. Finally, AMD's share in the server segment is a paltry 5.5%. AMD should be able to take some share here with their latest "Interlagos" chips, but the really exciting stuff will happen when AMD begins to release APUs for the server segment in 2013. Many HPC applications can take advantage of high performance GPU cores, so in these applications, AMD's offerings should offer an interesting and nontrivial advantage over competing, "pure CPU" solutions.
I also believe that as x86 tablets become more prevalent, AMD's low power APUs will play a significant role in driving revenue and profit growth. It will be interesting to see how such APUs will fare in terms of performance per watt, battery life, and graphics performance. I suspect that these chips will allow a very full windows experience in the tablet form factor for fairly cheap, especially if the integrated graphics allows significant advantages in games and other multimedia content over the ARM (ARMH) and Intel solutions.
Finally, I believe the graphics segment is valuable and will grow going forward. Despite the very low end discrete graphics being gobbled up by Intel and AMD's integrated offerings, there is still very strong demand for higher performance discrete graphics cards. Gaming, professional workstation, and high performance computing applications still require as much brute compute as possible. AMD and NVIDIA (NVDA)'s GPUs are very well suited to take advantage of these markets going forward.
AMD is not a dying company -- they are a company in flux and a company with a lot of good technology that, under the right management, can be monetized quite significantly and can bring significant shareholder value.
