$AMD - The Stock Vs. The Sector
Alright, so Micron's $MU earnings triggered some big selloffs across a variety of semi-names. Slowdown, macro, yadda yadda...
Semis are cyclical, as we know. But on the other hand, there are obviously areas of secular growth too. Data, AI, compute power.... all long-term trends. It just so happens that the purchase of these chips etc. ebbs and flows with the macrocycle as buyers allocate resources away from upgrading their servers, data centres, PCs, or consoles.
While all this logically follows, I think it's worth pointing out that this sector impact should be less prominent in a smaller market-share-taking business. Semi-production is a rather inelastic, highly dependent on capacity - kind of sector. So that means it takes time for the objectively best products to sell at scale - as a chip designer (but not manufacturer) like $AMD finds the excess capacity to meet demand squeezing out the lemons from fabs like Taiwan Semiconductors $TSM. The company has only recently developed some lemon squeezing mega momentum, that's been constrained by supply chain and capacity issues this past year. The AMD chips were arguably better in 2019, yet servers and data centre purchasers are only accelerating into their products now. That brings me to the point of this article:
I don't think AMD will be as demand struck on a sector downcycle as the price drop might imply.
3-month return performance - AMD & SOXX (Semi ETF) vs Benchmarks.
To add to this, I'd like to present sales metrics on trailing twelve-month sales. Intel's the primary competitor for CPUs here, so it's worth looking at both combined. Consider AMD GPUs to be relatively small here.
- Intel $INTC: 77B in total sales - flat YoY growth
- AMD $AMD: 19B in total sales - ~50%+ YoY organic growth, 65%+ inorganic growth
AMD, in the scheme of the full compute market, remains small.
Despite the rapid growth of AMD, as a result of general product and performance excellence, AMD remains in market share-taking mode, scoring a lofty 50%+ YoY growth excluding the Xilinx acquisition. They've built immense momentum, particularly in the enterprise, embedded and semi-custom segment that has outpaced their other verticals for a few quarters now on growth. Servers and Data Centers are key here. That's typically been Intel's domain which they controlled 90% of perhaps 2 years ago. AMD is now eating that pie up and is widely considered to have superior products for the price, evidenced by their recently massive margins, free cash flow, and most alt data benchmarks on chips. It, therefore, follows, that in a down cycle, Intel ought to take the brunt of the demand drop given its weaker product performance for the price and declining margins. I've read through Intel's ambitious plans, and hold the view that it would take time and a lot of evidence until something special materializes. Big ships with high CapEx are slow to alter trajectory. Meanwhile, Lisa Su at AMD's helm continues to deliver excellence quarter after quarter.
In my view, AMD should be relatively protected here from macro as it continues to deliver on promises. Intel seems like the bloated demand shock absorber here, as semi-capacity constraints put a ceiling on AMD growth. This is exactly the sort of environment where weakness tends to evaporate away - and strength stands through. Unlike previous down-cycles, Taiwan Semiconductors has seen unmeetable demand for existing and newer product portfolios - such as autos in the last two years. If end buyers are taking it easy on their other more directly cyclical purchases, then that ought to free up capacity in Taiwan for more AMD chip printing.
This isn't exactly a unique thesis. Well capitalized market share takers have - by rule of thumb - taken market share and come out stronger, especially through downcycles. When demand bottoms out, it's the stagnant enterprises that see business fall through while the takers maintain or thrive. The differential in size is still significant for AMD compared to Intel, and they should sail through relatively fine. Nvidia is the whole GPU market. AMD is only a fifth of their total market opportunity. This distinction is important to make when accounting for macro forces across a whole opportunity.
The stock should theoretically score stronger than the sector on fundamentals.
I'm long $AMD. Assuming it does $6B in Free Cash Flow this year - a very muted number considering they're planning for 60% YoY growth anyway to 26B in FY22 and just scored a 30%+ FCF margin last quarter.
It's now trading less than 20x FCF for FY22. Similar numbers on a P/E basis:
Price and Earnings Multiples.
This valuation makes it an easy buy for me for a quality leader with room for growth, chasing a $130B TAM. I've been long for ages, but I'm tempted to now add to my position.
The opportunity here arises from the market unfairly punishing the stock with the rest of the sector without accounting for its individual attributes.
For a hedge, maybe go short $SOXX if that's your cup of tea.
AMD has built big sales momentum
INTC still controls most of sales in the CPU market
Part of AMD's growth was muted by capacity constraints while demand has been ever-rising - an indication that it won't be too demand constrained going forward.
So when overall CPU demand goes down, AMD ought to continue to sell through its planned capacity, while Intel takes the hit. Intel remains most of the market, and AMD is a quarter the size in sales.
Now cheap, get it for less than 20x FCF for FY22