Innovation is happening so rapidly in this space (IT - server, storage, networking, DC virtualization, cloud) that after any good run, I would get out.
I wanted to separate this out from the TAM argument. You’re essentially saying that businesses based on technologies undergoing massive evolution, even disruption, won’t do well over time. I presume that’s because you believe competition will do to them what they did to the previous leaders. Like how DEC disrupted IBM only to be disrupted itself by SunM, which itself was disrupted by personal computers, which itself was disrupted by phones and tablets. So, clearly I’m familiar with the pattern.
But, I take issue with your “after any good run, I would get out” advice. While even companies that disrupt get disrupted, the telltale sign of that happening is typically not a stock price retrenchment. For starters, it’s really hard, impossible for most, to predict whether a stock price retreat signifies the end of high growth for the underlying business or not. I have a friend who sold out of NVDA in 2016 when it tripled to like $70 and then declined a tad. Later, after it broke $100 or so it had a more than 10% decline. I didn’t get out - matter of fact I sold Puts and increased my profits. This because I had confidence that Nvidia’s run wasn’t over. Wouldn’t you have said Nvidia’s 4X run was time to get out?
So, how are we to know today whether ANET’s run will continue or not? If anything, there sure seems to be lots of market share that Arista doesn’t have today that it could get. With new technolgies there are early adopters, majority, and then later adopters. Strikes me that Arista might only just now be entering the majority phase, which tells me that its largest growth is yet to come. I certainly don’t see any signs that Arista is in a late adoption phase or that Arista is going to soon itself be disrupted by something better. If anything, the company is super well positioned to move into higher bandwidth markets as they mature.
In comparing where NVDA and ANET are, I’d also say that NVDA has better competition. Companies like Intel (buying Mobileeye) and Google (TPU) have lots of money and brains to throw at things. Nividia is currently in a sweet spot of its traditional market (gaming chips & boards) still thriving (thanks to crypto currency demand) while its current growth market (servers) is doing OK, and while the hype around its future markets (AI, autonomous driving, etc.) is peaking. But, the crypto craze could end suddenly, servers may not be a large market, and the future may end up being many years off. So, I’m watching NVDA carefully right now.
Arista’s competition is Cisco, and it’s ripe for continued disruption. So, at this point I’m not worried about Artisa’s run coming to an end. There is not newer technology that I can see right now that would disrupt Arista, and like I said, it’s not near being in the late adoption stage.
So, the “after any good run, I’d get out” thing: that makes no sense to me at all with Arista. As you said, you’re not evaluating about the company or even its growth prospects, it’s just a stock price valuation criteria. Actually, the recent stock price decline is because of people with viewpoints similar to yours, and similar to my friend who sold NVDA at $70 something after its “good run.” This is simply a case of people who don’t understand the company and the business being scared because the CEO is a well grounded leader who isn’t promising the moon. She has a history of underpromising and the company overdelivering.
Finally, it’s difficult to apply valuation metrics to growth companies. A company like McDonald’s is trading with a PE of about 25. Nvidia is twice that. But, it seems an easy argument to make that Nvidia will grow twice as quickly as MickeyDs. And, even with ANET now trading at a PE of 52, it seems obvious to me that Arista will grow faster than both Nvidia and MCD. Now if you want to talk overpriced, we can go back to my train wreck favorite, CMG, which is also has just about a PE of 50. New CEO or not, there are better places for my money.
And this is something I’ve learned over the years - that when a company is destined for success (like ANET), many many people recognize that. The question becomes whether its eventual success will be even greater than the market has already priced in. A stock can be priced high with expectations for 30% YoY growth, but if it’s actually going to achieve 40% YoY growth, then it’s underpriced.
I don’t “get out” after “good runs.” I get out when the business prospects for the company have changed. I don’t see that today for Arista at all, and am a buyer (and seller of Puts) today.