Cloud Hosting

This article was highlighted for PRO+ subscribers, Seeking Alpha's service for professional investors. This article was highlighted under our Tech vertical. Find out how you can get access to the best content on Seeking Alpha here.

The shares have Arista Networks (ANET) have experienced woeful performance until the last few weeks. Overall, they lost 42% of their value from peak to trough, with the latest trough occurring in early November after the release of Q3 earnings, which were fine and guidance which was dreadful. Lately the shares have bounced around 15% since the start of the year based on hopes that guidance was a bit of a sandbag and regarding the possibility of Arista acquiring Big Switch, one of the key movers in the networking world. (Industry press has confirmed that the deal closed, but there has yet to be a formal press release from either party. I assume a press release is most likely when Arista announces its quarterly earnings on February 13th. The deal is said to be on very favorable terms (a song according to the linked article, and is more about Arista acquiring people and technology.)

Should readers consider taking a position in the shares at these levels? I think a strong case can be made for entering the name at this price and at this time. There are plenty of demand growth drivers that have yet to be felt by Arista and which could turn current estimates upside down. The acquisition of Big Switch could mark a renewed period of growth for Arista based on the combination of technologies. And the Intel (INTC) announcement regarding strength in its business from the equivalent of Aristas cloud titan segment has to be taken positively. But most of all, in a sea of high valued names, ANET shares are painfully cheap, particularly on an EV/S basis, if only the growth, forecast to be negative for the next two quarters, resumes at a meaningful pace.

So yes, this is a call to buy the shares, and I intend to do so opportunistically, perhaps starting this week.

In the course of editing this article, I became a ware of a downgrade of Arista shares by the analyst at Barclay's from buy to hold. The principle issue raised by the analyst is his contention that Arista is losing market share or its market share gains have stalled in the cloud titan space. That is such an inflammatory comment that I needed to reference it in what is a rather long article. I am not sure as to the basis comment by the Barclay's analyst. Needless to say, it is one that I feel is way off base. At the end of the day, given there aren't very specific statistics by quarter that can be used to determine market share data with regards to cloud data center switching revenues, one is left to rely on proxy data. As I will detail in this article, Arista has indicated that its forecast both for the quarter to be reported, and the for 2020 are predicated on flat to down capex for switching in its key cloud titan vertical. The evidence we have is that is exactly what is happening. Later in this article, I have illustrated from company data that Microsoft has indeed cut its capex. The business results of FB also suggest a company that is likely reducing its capex in an effort to size its infrastructure growth to the growth in revenues. That is precisely the forecast that was made by the Arista management.

Because this is a key point for me, I think it reasonable to look at the precise answers made by company management on the market share issue. "

Alex Henderson -- Needham & Company -- Analyst

Great, thank you very much. I was hoping you could spend a little bit of time relative to this cloud issue, to what extent you're confident that there is no competitive incursion here that's causing it, and that in fact you have sustained share at that customer. How can we judge that -- how do you get your arms around, clarity around that point?

Jayshree Ullal -- Director, President and Chief Executive Officer

Alex, that's a very good question. From our perspective, the competitive dynamics have not changed in the cloud or in general as we always have aggressive competition and we will continue to see aggression there. But what gives us confidence the cloud titans are delaying their spend or distributing their capex differently is, as you know, we always pride ourselves in a close partnership and relationship with cloud titans. And generally, especially in the case of Facebook and Microsoft, they have been not only a vendor customer relationship, but really a core development that requires the kind of partnership which is engineering to engineering, it's not just business.

So, when you look at that, there is no evidence that competitively or white-box wise, there has been any change, there has been a process processing, there is better inventory management, there is better procurement, optimization etc.. And you can always expect these cloud customers of ours to want to be multi-sourced, but it isn't any different than we've seen in the past in behavior, in relationship, in our innovation we have 10, 400 gig products, and lot of then in styles . So, relationship and the technology partnership couldn't be better. Anshul, do you want to add to that?

Anshul Sadana -- Chief Operating Officer

Sure. Thanks, Jayshree. Alex, we work very closely with these customers to a point where we are working on this 2021 roadmap along with these customers right now. And quite well aware of the thesis they are making with the architecture as well and have very direct feedback from customers as well that there is no alternate that's pleasing us, it's simply the demand has gone down and we are very confident of our share when that demand comes back as well since we collaborate with these customers. So, we're not worried about it. And the customers are pretty direct as well, this is not our share going to someone else. Their demand reduce.

Alex Henderson -- Needham & Company -- Analyst

Okay. Thank you very much."

One can either accept the specifics of management commentary on market share, or accept the unsubstantiated assertions of the Barclay's analyst. The Barclay's analyst has apparently had some issues in calling this name properly having initiated Arista as a buy last August before the company forecast its problems with the cloud titans. I might also observe that the analyst is positive about Cisco. At some level, it is not really possible to be positive about Cisco and positive about Arista. Either one company has staunched market share losses and is suffering because of macro demand issues, or its smaller rival is continuing to take share. I will take the latter side of that argument based on the industry sources, such as they are, that seemingly suggest that Arista's competitive position remains positive.

Arista Networks was founded to cater to users who needed high capacity switches that took advantage of the latest technology-an oversimplified description bound to irk some readers but one that captures the essence of what Arista does. One of its co-founders, Andy Bechtolsheim, is well known as the co-founder of Sun and a founder of Granite, a company that became the core of Ciscos (CSCO) Gigabit systems unit. And he was one of the early investors in Google. (GOOG).

The companys CEO, Jayshree Ullal has held that role since 2008. Before that, Ms. Ullal had held senior management roles at Cisco and was responsible for the companys switching portfolio and its offering of unified communications technology.

The company was a pioneer in software defined networking and in offering multi-layer switching technology. The companys Extensible Operating System (EOS) is the core of its technology. The company has always used merchant silicon as part of its technology, allowing it both to take advantage of advances in technology and to achieve very high gross margins which are greater than 64% on a non-GAAP basis. The low latency of the companys switches has appealed to important users, and in particular, what the company calls Cloud Titans, which are firms such as Facebook (FB), Google (GOOG), Amazon (AMZN) and Microsoft (MSFT).

For some years, the story was a happy one with but few distractions other than a hard fought IP lawsuit between this company and Cisco. The company was wildly successful and very profitable selling its offerings to both the Cloud Titans identified earlier, to smaller cloud vendors, to financial services companies in need of the low latency and to service providers and enterprises. The company went public back in 2014 at a price of $43/share and first traded at around $58/share.

The year the company went public, ANET had revenues of $392 million, EPS of $1.29 on a GAAP basis and operating cash flow of $132 million. This year (2020) the company is forecast to have revenues of $2.4 billion with non-GAAP EPS of $9.06. So, over a span of several years, valuation has compressed noticeably. Is ANET a value name-not quite, but it has an EV/S of less than 3 on a forward basis and it has a 30% free cash flow margin.

Over the past year, the ANET story has been shaken by issues regarding the growth in demand for some of its cloud titan customers. Different titans at different times have slowed down their procurement of network infrastructure. The issue hasnt been loss of share, or even an overall slowdown in growth of the cloud titans, but essentially changing architecture that has allowed some of the titans to achieve their latency goals without continuously upgrading servers and switches.

And some of the companys growth initiatives such as the 400 gb switch and the companys campus switch offering have simply not taken off-or taken off more slowly than the rate that some commentators, including this writer, had once hoped for.

Overall, revenue growth which had been 45% as recently as 2017, is now forecast to have been 12% last year and the First Call consensus estimate for 2020 shows a forecast of no growth at all. EPS is forecast to decline by more than 6% at this point, which actually would be a strong result if revenues dont grow at all. These next two quarters, essentially congruent with company guidance, forecast the nadir of revenue decline, with March quarter revenues expected to fall by 12%.

With those kinds of expectations, it is little wonder that Aristas share price has suffered grievously. Last year, after hitting a high of $328/share in April, the shares imploded to as low as $188 by the start of November. Since then, there has been a strong rally and the shares are now back to the mid-$230 range.

Starting earlier this month, Arista shares have rallied very strongly, rising from about $203 at the start of the year to Fridays close of $238 last Friday, prior to the current panic regarding the Coronavirus and its possible impact on world economic growth. Over that span, the companys Chief Customer Officer resigned, and then rumors started to fly about the company winning the race to acquire Big Switch. In addition, the earnings release of Intel (INTC), which spoke about robust demand from cloud service provider customers, has lead to some speculation that what Intel saw, Arista is likely to see at some point in the near term.

Not all of the potential positives are likely to emerge in one fell swoop. But I do believe there are enough positive potentials to suggest that the risk/rewards of establishing a position at this level makes sense in a portfolio that is willing to own some speculative names.

The acquisition of Big Switch by Arista is all about a new direction for the company that involves a much greater focus on the enterprise, and a much lower level of customer concentration. While we do not know just how much revenues the 4 cloud titans have accounted for at Arista, we do know that their slowdown in network switch acquisition has brought the companys growth to a halt. By acquiring Big Switch, Arista is changing the center of gravity of its go-to-market effort and should become perceived as more of a software company. I do not think that any hyper-growth vendor can survive at this point simply peddling boxes, even if the boxes come with lots of software. I think it is premature for me to speculate precisely how the numbers are going to work. And I think that commentators that might look at historical revenues of Big Switch in evaluating this transaction will be looking at it upside down-this is a merger about changing the future of Arista by buying the technology leader in software defined networking. Just how the revenue synergies play out is really indeterminable at this point and any guesses I make would be just that-guesses

At this point, it hard to speculate just how the relationship between Dell and Big Switch might evolve. Presumably that has been a component of Aristas evaluation. Overall, the acquisition is likely to help Big Switch revenues maintain triple digit growth for several years into the future, but perhaps of more importance it is likely to help Arista sell its high performance switches to enterprise vendors who may have been reluctant to buy from Arista before it acquired this capability.

As mentioned, there has yet to be a formal press release detailing the purported transaction. So much of the following is speculative-although even after the deal is announced, there will be much white space and speculation in figuring out how the two companies will combine and achieve the revenue synergies required to pay for what was inevitably an expensive transaction.

How much Arista might have paid for Big Switch is not known-although it is said to be "a song." Big Switch is thought to have had multiple suitors over the last several months and of course some of these suitors like Dell have significant financial resources. Arista had gross cash of more than $2.4 billion on its balance sheet at the end of September, and is generating cash at a rate in excess of $800 million so it should not have had trouble in financing the deal. Big Switch may have been burning cash lately; its last capital raise was back in 2017 and just the working capital needed to support an enterprise of this size and with triple digit growth might suggest that the company needed to find a home where capital constraints were not an issue.

Big Switch has been a private company which has been growing at triple digit rates for some time now. As opposed to much of Aristas focus on the cloud titans, this company is focused its sales effort on selling to the Fortune 100 and it has some installations in 30% of those users. While there is no real estimate for revenues, the company has raised about $120 million in venture funding. Some analysts believe that the company has a current revenue run rate of as much as $150 million-but obviously Arista is not buying Big Switch for its current revenues-this is deal about major revenue synergies and one that is likely to redirect some of Aristas focus.

Big Switch was founded by some of the research team that developed software defined networking. There are some observers who have maintained that its founders were just too far ahead of its time. Just to be clear, it is a software company and it has partnered with hardware vendors, and in particular Dell and HPE, with the Dell relationship said to be the most meaningful. It also has relationships with Nutanix and VMW. Big Switch only sells through channel partners at this point. The company uses and contributes to many different open-source communities. I have linked to a couple of evaluations of the Big Switch technology by current users.

The company currently offers 2 primary fabrics. The Big Cloud Fabric is built with cloud networking design principles and delivers cloud-Network as-a service performance for private cloud platforms such as those offered by VMW, Nutanix, Microsoft and Red Hat. It is another visibility platform, but one that is designed to operate in a multi-cloud environment offering one-click trouble shooting.

Its other major product is called Big Monitoring Fabric. It is known by users as Big Mon-and I have had nothing whatsoever to do with making up that name. That platform is all about providing users with what is known as pervasive visibility which includes elements of predictive analytics, security monitoring, connection tracking and correlations between events and applications.

Users have adopted this technology because it lowers TCO by 50% over 5 years, it apparently speeds up new service enablement dramatically and improves change management. There are many competitors in this space, and most larger networking vendors have bought a specialist vendor so they can offer their users something in terms of network management. Arista has lacked a significant competitive capability in attempting to compete in the enterprise with both its network and campus switches by not having this technology and will inevitably be able to enter far more competitive procurement processes than had heretofore been the case.

In closing this segment, I think it worth noting that on last quarters call, Arista talked about its introduction of CloudVision 2019, a platform designed to bring cloud principles to network operators across Places in the Cloud. This is very similar to some of the offerings of Big Switch and I expect to see the Big Switch technology offered as part of Aristas CloudVision platform as an important offering for the companys enterprise and financial users.

As mentioned earlier, part of the reason for Aristas share price decline at the time of its last earnings release was its guidance and its comments about demand from cloud titans. Two of Aristas customers, Microsoft and Facebook are each more than 10% of revenue. Both of these users have seen some variable capex and indeed, this afternoon, MSFT announced that its capex in the December ending quarter had fallen, rather than risen, and that was enough to drag shares of Arista down after hours.

I have chosen to provide readers with the exact quote from the call rather than interposing my own filter. Here is the part of script that is relevant, After we experienced the pause of a specific Cloud Titans orders in Q2 2019. We were expecting a recovery in second half 2019 for cloud titan spend. In fact, Q3 2019 is a good evidence of that. However, we were recently informed of a shift in procurement strategy with a material reduction in demand from a second cloud titan, reducing their forecast dramatically from original projections for both Q4 2019 and for calendar 2020. Naturally, this type of volatility brings a sudden and severe impact to our Q4 guidance.

Given the step in forecast and volatility of this cloud segment, we believe the cloud titan forecast should be modeled as flat to down in calendar 2020.

The following quote is from the CFO and is perhaps somewhat different in terms of providing an outlook than the quote from the CEO,

All indications or are these actions do not represent a loss of positioning our share for Arista at these customers, but will likely effect -- will likely result in demand from this part of the business being flat to down on a year-over-year basis for the remainder of 2019 and into 2020. While we are not at this point in a position to provide overall guidance for 2020, we did want to make the following points. at this point, we believe this trend combined with typical Q1 seasonality and the recent updates to cloud forecast described above may result in revenues for the first quarter of 2020. They are approximately 5% below Q4 2019 level.

The current consensus forecast for Q4 and Q1 are almost exactly consistent with the guidance that was in those passages. The guidance as provided would lead to Arista showing revenue declines in Q3 and Q4 by 7.5% and 12% respectively.

The CEO elaborated that her concern related to the fact that she believed that cloud titan capex was likely to be flat to down, and that the titans would pivot their capex from networking to infrastructure. Further, the CEO said that the 400gb product line, while shipping to customers, is not seeing adoption at scale tracking the results that were seen when the 100gb product line were introduced. Adoption is slower, although why that might be is not quite clear.

Intel, however, in its latest earnings report, covering the period down through 12/31/19 had a different experience and a different outlook as can be seen in the following quote from that company's conference call. Intel's collection of data-centric businesses achieved record revenue in the fourth quarter, led by record Data Center Group (DCG) revenue. DCG revenue grew 19 percent YoY in the fourth quarter, driven by robust demand from cloud service provider customers.

Intel is a much larger company than Arista, and even the DCG segment is several times the size of Arista. Intel and Arista are not competitors; they often partner and collaborate. And it seems unlikely, simply based on the overall business trends for the cloud titans, that they would make a decision to slow capex growth. While there obviously is not a one for one correlation between revenue growth and capex spending, and cloud titans have many capex priorities, it seems unreasonable to believe that the titans are actually pausing capex or can pause it for very long, and far more reasonable to expect that cloud titans are investing heavily in their infrastructure.

It is much more difficult, at least for me, to determine whether the comments of the Arista CEO regarding the reprioritization of capex at the titans is still valid. Apparently, the one titan that essentially shut down their procurement of Arista switches did so in the wake of a shutdown in the procurement of servers. So, at least in this one instance, capex for servers and capex for switches has been correlated. If capex for servers is now rising again, per the comment in the Intel earnings report, than so too, should capex for switches. I think it is evident, given the rather dramatic changes in the forecast being seen for the titans and their procurement of either infrastructure or switches, that there is a material lack of visibility.

It isnt that Arista, and its management had any specific data when they last reported numbers, but the numbing effect of being shut down at one particular user that unnerved management, and lead to what I believe to be hyper-cautious guidance and commentary. My own guess, and I freely acknowledge that is all it is or can be, is that Arista will report upside in revenues, and a return to a more benign demand environment for their cloud titan segment over the course of 2020-although this trend is likely to take a couple of additional quarters before it becomes totally apparent. Overall, I think the likelihood that Arista can achieve double digit revenue growth in 2020 is better than 50/50 and despite the run-up in the shares, the valuation does not reflect that kind of outcome, in my opinion.

My thesis regarding investing in Arista is long term. Aristas success is not specifically about calling a turn in demand for high performance switches from cloud titans. The company itself doesnt have a perfect record in terms of understanding how cloud titan demand wanes and waxes-and that is putting it kindly. I would be foolish to suggest to readers that I have specific knowledge regarding the outlook for the quarter that will be reported or the quarter that has recently started. And even more emphatically, I cant foretell what guidance this company might provide for the full year.

In one sense, if you are a long term investor, it doesnt matter greatly. Of course I like to get quarters right, and depending on what is reported and what is forecast the shares will react materially. If the guidance for the full year is still marginal, and still calls for a revenue decline, the shares will go down, and I would far rather buy for less than more. So buying before this quarterly earnings is a gamble-although less of a gamble than it might appear because of the companys still compressed valuation.

My contention is that this company can and will return to mid-teens growth over the next 12-18 months as several demand generating factors start to generate business. I do think it is reasonable to believe that cloud titan capex, and the demand of cloud titans for network switches will revive. But investors should also take account of the other factors that will influence growth.

Of course the acquisition of Big Switch is a component of that, and indeed, if it works out as I have tried to suggest might be the case, the revenue synergies from the deal alone would go a long way to change the overall demand curve for Arista. But beyond that, over the past couple of years, the company has made several conspicuous product announcements that are likely to be factors in generating growth for the company outside of the cloud titan area by the second half of this year.

Probably the most significant is Aristas line of Campus Switches. Arista indicated that volumes for the product family had reached or were reaching an annualized rate of $400 million or about 15% of current revenues. I think that the campus switch offering from Arista has some unique features and is quite differentiated from the offerings of competitors. Importantly, about 50% of customers are new to Arista. It is apparent that the campus switch offering has driven Aristas enterprise segment into position as the companys second largest revenue segment. It doesnt seem farfetched to expect campus switch revenues to provide a 1000 basis point tailwind to total revenues both in 2020 and beyond.

The company has launched the next generation of switches, its 400GB -7500 product family It is now about 16 month since the launch, and the company hasnt seen the kind of acceptance it initially had been expecting. Many observers, including this writer, had expected a demand growth cadence similar to that which had been seen during the introduction of the 100GB product line. Why it has taken longer for the 400GB architecture to get accepted is not readily knowable. Many in the industry have different theories. Arista has talked about the optical components, necessary for deployment, having moved out such that initial deployments of the switch on a mass basis will come in 2H of this year.

The introduction of the 100GB solution set made Arista. It was the single most salient factor in the companys hyper-growth and market share gains in 2016-17. The reasons users have migrated progressively to higher performance switches havent changed. Users need more bandwidth to support digital transformation applications. The cadence of the migration will not determine the amplitude.

Does Arista enjoy the same level of perceived product differentiation withing the 400 GB switch category as it has enjoyed heretofore. That is another issue that is hard for me to determine with any degree of confidence. I have every reason to believe that the advantages that Arista has enjoyed havent changed much in this category when compared to the 100GB switch. My guess, too, is that when the apparent logjam of implementation breaks, the upside in demand for Arista will be far greater than anyone might prudently forecast. I have no reason to believe that the growth in demand caused by the advance in technology will not be of a similar magnitude to the growth seen in 2016-17, and that is obviously not what is being factored into forecasts or valuation analysis.

Finally, it might be well to suggest that some growth for Arista is likely to be coming from Aristas other two acquisitions, Metamako and Mojo. Both of these are smaller companies, but both of them have solutions that fit well for Aristas strategy. Can they produce a few hundred basis points of revenue growth. Both of them fit the thesis of pivoting this business to one with a more balanced revenue profile: Metamakos solutions are focused on users in financial services with ultra-low latency requirements and Mojo invented the Cognitive WiFi offering that has facilitated the development of cloud managed wireless networking.

Recommending ANET shares at this point involves a certain amount of a leap of faith or a leap into the dark if one prefers. I chose to write this article at this time to take advantage of that uncertainty and to comment positively on the potential ANET has to use its Big Switch acquisition to emerge as a different company from one exclusively based on selling high-performance switches to Cloud Titans.

There are certainly straws in the wind suggesting that some of the demand obstacles that Arista commented about last quarter have abated. Amongst these would be multiple comments from various companies talking about strength in enterprise IT spending, comments from Intel that its business with cloud providers was very strong in the December quarter and commentary from some 3rd party industry analysts suggesting that the pullback in IT spending is, or will shortly abate. Fundamentally, the growth of the cloud titans, and their need to manage data hasnt changed; whether the timing of their capex, or its priorities may have changed they still have to manage data in their centers to provide their end-customers low latency service.

On the other hand, MSFTs quarterly report out this afternoon as I write this, showed a noticeable fall in Q4 capex, confirming in part the rather subdued guidance for Arista about which I commented earlier. That said, MSFT is not likely to be able to continue to grow its cloud revenues by over 60% and to grow other components of its business at elevated rates without the necessity of upping its capex at some point over the coming quarters.

Arista talked on its latest call about demand issues with its two smallest verticals, Tier 2 Cloud Vendors and Service Providers. I imagine the Service Provider space will continue to be a negative for Arista as it has been for most other vendors selling to service providers . I think the negative sentiment expressed by the management of Arista on its latest call regarding Tier 2 cloud vendors may have been overstated. The overall growth in traffic in the cloud is not slowing, even if that is not always reflected in revenues.

I wrote extensively about Big Switch and its potential revenue synergies as well as its acquisition changing the center of gravity of this company. I suppose we are going to have to wait until February 13 before we see an announcement, and just how many details will be revealed at that point is not knowable by this writer. Still, I think that taking the core capabilities of Arista, and moving them into the broader networking space is likely to produce a highly favorable outcome that has yet to be fully discounted.

While ANET shares have bounced noticeably from the levels they fell to after the last conference call, the valuation metrics that I use are still materially constrained. Even after a rally of about 26% from the companys trough, the EV/S ratio is still less than 3X and the free cash flow margin is at or over 30%. I have used a 3 year 10% growth rate in finding a growth cohort for Arista. I obviously believe that Arista has a potential to substantially exceed that kind of growth over the next several years, but regardless the shares are more than 40% below the average EV/S for 10% growth. In addition, the shares have a free cash flow margin that is 30% above average for the 10% growth cohort I think the management team is particularly strong as I have detailed in prior articles on the company. Not all of the positives are going to emerge in a single quarter or on a single conference call. But I think the risk/rewards are very strongly stacked in favor of a commitment to the shares at the current time and at the current price.

Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in ANET over the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Read more from the original source:
Arista Networks - Is This A Phoenix And The Implications Of The Apparent Big Switch Acquisition - Seeking Alpha