I will be completely upfront and say that my view on ARM Holdings (ARMH) has evolved rather dramatically over the past year. While I do think that the company trades for a sky-high valuation, and while there is risk that Intel (INTC) or MIPS via Imagination (OTCPK:IGNMF) will gain some pretty meaningful processor share, it's tough to deny just how well run ARM is.
ARM develops CPU IP, GPU IP, system-on-chip interconnect IP, and physical IP. The company has been fantastic at enabling its chip partners, putting out a broad portfolio of IP for just about every conceivable use case, and just executing as a best-in-class semiconductor IP. Warren East really did a fantastic job of setting the stage for future growth and I believe that the current CEO Simon Segars knows exactly what he needs to do in order to keep the business growing nicely.
That Said, ARM Trades at a Very Fragile Valuation
Many people don't understand that there's a world of difference between liking a stock and liking the company. I think ARM Holdings the company is absolutely fantastic. However, the problem here is that the company's valuation is so incredibly rich no matter how you slice it that it is very vulnerable to negative news flow.
See, that's the thing with high growth tech stocks. While you can do a discounted cash-flow analysis until you're blue in the face based on whatever assumptions you want, the truth is that during the "growth phase" of these companies, the share price will be very volatile and very sensitive to news. This is because nobody really has any idea what the long-term growth rate actually looks like and the range of potential outcomes (and thus valuations) is pretty wide. When things are "looking good", people assume the "best case" and when things are "looking bad," people assume the "worst case." The truth is very often somewhere in the middle.
The Latest Report Wasn't Bad, But...
ARM's latest earnings report, despite the one-off charges, wasn't actually that terrible. However, the real problem is that they're not beating anymore. The company is simply coming "in-line" with analyst expectations and, unfortunately, due to the saturation of the high-end smartphone market, the imapct of the mix shift from feature-phones to smartphones is not only nearly complete (which has been a big driver of ARM's royalty growth), but the remainder isn't anywhere near as lucrative as the initial part of the transition.
Now, ARM does have expansion into networking and communications to look forward to as both ARM and X86 displace PowerPC and MIPS to some degree, but the unit volumes there — even with the juicy royalties per unit — probably won't drive anywhere near the level of growth that smartphones/tablets once did. At any rate, ARM's growth seems to be slowing and the company — after a pretty epic run from 2009 — seems to finally be "maturing".
That Said, I Can't Get Too Bearish
While I think that ARM-the-stock is probably "worth" about $20 on a discounted cash flow basis today, one needs to be very careful in assuming that the most recent drop is the "end of the line" for the stock's run. The stock has shown rather remarkable resilience and has recovered from these fairly large drops multiple times in the past and as long as ARM can play up the Internet-of-Things and as long as Intel keeps fumbling in mobile, investors are unlikely to be so keen to unload their shares.
I also see some potential positive news-flow helping to prop the stock up from all of the "server" fanfare that has been made. While I do think it will be extremely difficult for ARM's crew of small licensees to break into Intel's server stronghold, I do expect that until more of these ARM server players begin to drop like Calxeda did that the "dream will be alive." It's tough to short such a widely-held "dream" when the popular media still thinks that Intel doesn't have a competitive micro-server part (it does).
ARM is a great company, but it's extremely expensive and basically trades on news-flow. I don't see the news-flow getting too bad over the next year (unless, of course, Intel does make some meaningful inroads in mobile in which case, all bets are off), but it's also tough to see it getting too much better. At this point, I'd imagine that ARM will trade within a $40 to $50 range for a while until some truly meaningful catalysts show up.
Whether this means that ARM will break down and head towards $20 or rocket to new highs, I wouldn't be able to tell you today and I'm personally not willing to take a position either way today. However, I wouldn't underestimate Intel longer term in mobile, nor would I place undue faith in the "ARM server ecosystem." I do think, however, that the internet-of-things story (which is really just micro-controllers which has already been a great business for ARM) could be interesting, although probably not enough to drive the same kind of revenue/profit growth that mobile was able to.