This article was highlighted for PRO subscribers, Seeking Alpha’s service for professional investors. Find out how you can get the best content on Seeking Alpha here.
Fastly’s Quarter-It was as fast as a speeding bullet-but not everyone likes speeding bullets
Fastly (FSLY) reported the results of its latest quarter on August 5th, 2020. The share had previously spiked into the numbers reaching a high point of $116. At that point, the shares had appreciated by nearly 6X from where they had started the year, and were up by almost 150%, just since June 1st. The latest spike of the shares had come in the wake of the strong results posted by one of Fastly’s significant customers, Shopify (SHOP). The Fastly solution has proven to be immensely attractive to the requirements of Shopify, to the point where Shopify has become a poster child for Fastly’s technology: Shopify : Case study | Fastly.
While Fastly has more than a few customers, investors decided to make a bet on a Fastly upside, based on its relationship with Shopify and the future opportunities that such a relationship suggests. That said, one rarely sees the level of appreciation described above based on company operations, and thus the shares were vulnerable to any kind of disappointment.
The company’s short interest is a relatively moderate 9.33%. For the most part, volume has spiked relatively less than is typical in the immediate reaction to the implosion of the shares.
For those unfamiliar with this story, Fastly shares have been a significant holding of a hedge fund Abdiel. As of the last report on Abdiel’s holdings, it had a position of about 8.7 million FSLY shares and it had been a buyer in Q1. 8.7 million FSLY shares today would be worth $661 million, or perhaps 25% of Abdiel’s portfolio. It may well be that Abdiel has been selling shares in order to better manage its portfolio, and that it will need to continue to do so.
While the quarter was in no way a disappointment, forward guidance reflected an issue that was not previously recognized by investors; the fact that TikTok, the Chinese based short-video application had become a 12% customer for Fastly. The issue, of course, is the demand by the Trump administration that TikTok’s owner, ByteDance should be banned from US operations effective September 15th in order to preserve data security from the possible inroads of Chinese intelligence agencies. I obviously haven’t the domain expertise to determine the validity of the Trump claim or the likelihood that its threat can be realized. Forcing a sale, and then insisting on some seigniorage payment for forcing a transaction is somewhat of a novel approach to international relations.
This is likely to prove to be a complex story with many twists and turns. Here is the latest “news” on the subject, although just how much credence can be placed in an article from the “South China Morning Post” which cites unnamed sources and some kind of weird probability regarding the outcome is difficult for this writer to handicap. ByteDance to escalate fight against Trump’s TikTok ban, sources say. To a certain extent, I am inclined to imagine that this article is based on the current official Chinese government position, and it may or may not turn out to be accurate.
In my own experience, rarely does the worst happen in these kinds of situations, but as I will explain later, Fastly management, in what is perhaps an abundance of caution has chosen to provide guidance that essentially excises any contribution from the use by TikTok of the Fastly CDN.
This quarter was one that I imagine will be hard to replicate in terms of percentage growth any time soon, and that is regardless of the outcome of the TikTok controversy. The company achieved 62% revenue growth for the quarter; sequential growth was just less than 20%. This company does not have a bookings metric because the company derives a significant component of its revenues from the usage of its CDN. The company acquired 114 net new customers last quarter, the largest new customer acquisition since the IPO of the company’s shares. The company’s enterprise customer count continued its slow uptrend rising by about a bit over 2% last quarter, while the revenue per enterprise customer rose by 12% sequentially, showing just how strongly the company’s CDN usage grew last quarter, which was coupled with existing customers adding additional Fastly modules as part of their overall build out of a Fastly-centric infrastructure.
The strong revenue growth afforded the company the ability to show a high level of leverage. Gross margins, in particular rose by more than 500 bps last quarter year on year, and actually rose by more than 600 basis points on a non-GAAP basis. That kind of expansion has much to do with the incremental margin based on usage being close to 100%. Overall, the company’s net expansion rate rose to 138% this past quarter and that compares to 130% the prior quarter.
This was the first positive quarter for Fastly in terms of adjusted EBITDA. The company saw some improvement in non-GAAP sales and marketing and research and development expense, but general and administrative costs spiked quite a bit to levels that will almost surely be remediated in future quarters. Overall, non-GAAP operating expenses were 59% of revenues last quarter, compared to 78% of revenues in the quarter same quarter the prior year. While some of this improvement was based on the sharp increase in stock based comp., overall the improvement might have been larger still, except for the 100% spike in GAAP general and administrative costs.
While the company burned a little bit of cash last quarter, the cash burn was more than entirely a function of the increase in accounts receivable. With revenue growth at very high levels, and much of the growth coming from usage trends at existing customers, the spike in receivables was essentially unexceptional. The company has plenty of liquidity, and last quarter its capex, constrained to a degree by the decision of Fastly not to move into new territories due to the impact of Covid-19, fell to just $2 million.
So, What’s the problem
Looked at objectively, Fastly continues to grow at accelerated rates, and with strong unit economics. The company has a risk with its 12% customer that may or may not happen, but it has plenty of opportunities to back fill the revenues it has secured thus far from TikTok. The company has enormous opportunities to work with a variety of larger clients to build market share in the edge computing area, and there was nothing in the Q2 earnings release that might suggest otherwise.
That said, there are a couple of issues that have probably upended Fastly’s upward share price trajectory-at least for the time being. One of those-and the one that is of most concern to this writer-is obviously valuation and the share price melt-up of the last several months. The fact is that very few stocks have seen the upside of Fastly since it took off in the wake of its earnings release for Q1. How much “should” Fastly shares appreciated. Was 4X in literally less than 4 months too steep an ascent for valuations at those levels to persist?
Prior to March, which seems like an eon ago, these days, Fastly’s long-term growth expectation had been around 30% or even less. There was no particular reason for that number so far as I could tell, other than analysts decided it felt right. Frankly, I never did feel that a 30% growth objective was anywhere close to the opportunities presented by advanced edge computing Then came a quarter announced on May 6th, with 18% sequential revenue growth, and a material guide-up, and the rest, as they say is history. This current quarter showed sequential growth of 19%, bringing the year on year growth to 62%.
Now, the company, while beating revenue and earnings expectations materially, as outlined above, and still guiding up further, has presented investors with a bit of a conundrum in terms of sequential guidance. What Fastly sells is on the frontier of computing-I will discuss some of that assertion below. That is not to say that there wasn’t edge computing before Fastly, but the Fastly CDN offers both extra speed-hence the name-and several other key features that users find to be of great value. In that regard, I suggest that the link I provided to Shopify’s testimonial is as good a place as any to understand the advantages that Fastly offers to enterprise users.
But the guidance that the company provided at this point which shows Q4 year over year growth of “only” 45%-50%. I personally do not find that kind of year over year growth disappointing. It would be hard to imagine any company at this point, given the uncertainties in the economy, and the uncertainties relating to network usage, to try to project greater than 50% growth. And Fastly in particular, is trying to figure out the direction of revenues from its 12% customer TikTok.
I have used a 3 year growth projection of 42% in looking at Fastly in a growth matrix. That is almost 1000 basis points greater than the published estimate provided on 1st Call for growth in 2021. My revenue estimate for Fastly is $354 million on a 12 month forward basis, and that yields an EV/S of 19X. That estimate is just a couple of percent greater than the consensus estimate for the same time period. That kind of valuation is about 20% below the current average for a low 40% growth cohort. But it is quite a bit above average for the published consensus growth of 33%. And for a variety of reasons, Fastly is not currently, and is unlikely to be a cashflow story-it is all about growth, although profitability is likely to steadily improve at scale, in my opinion.
One of the issues that has happened in the course of the upside down market brought on, in part , by the pandemic, is the desire of companies to be prudent in providing guidance, and the seeming inability on the part of investors to try to evaluate precisely what companies are suggesting. To reiterate, Fastly reported a halcyon quarter; in part a function of continued high usage of their CDN. Management has been cautious in forecasting that sequential growth in usage from current levels might continue higher. Investors want to be told something that prudent managements are simply unwilling to articulate.
And the fact is, there is no rational basis a company such as Fastly might have for forecasting a continued surge in the usage of its CDN. The company obviously doesn’t report changes in usage on the part of all of its clients. Many observers have speculated that the strong jump in usage in Q2 was in whole or part, a function of the very strong jump in usage by Fastly’s client Shopify and presumably other Fastly clients such as Etsy (ETSY), Foursquare, Hearst. Pinterest (PINS) and Opera (OPRA). I have no reason to question that kind of speculation, other than to suggest that Fastly has 304 enterprise clients, and I am surely not in a position to know which of that group added modules or increased their CDN usage during Q2. I might further observe that Q2 marked an all-time high for net customer adds, and often these customers who start at small levels, ultimate graduate to the enterprise level-a typical land/expand strategy.
In particular, Shopify saw its Merchant Solutions revenues grow by 84% sequentially in the quarter, and its growth in merchandise volume on the web was greater than that. But I really do not think that it is quite reasonable to take that kind of volume growth sequentially, and think that Shopify’s usage of Fastly platform rose by anything like that amount.
Since this company went public, and long before the advent of the pandemic, Fastly has made it a practice to attempt to minimize the guidance it is willing to forecast because of CDN usage spikes. As long ago as last year’s Q4, the company went out of its way to deny it would see a material usage spike in that quarter and what followed was 18% sequential growth in revenues in the quarter, a far cry from what had been forecasted, or the sequential growth of 8% the prior quarter.
The company is now forecasting essentially no sequential growth in the current quarter, followed by 16% sequential growth in Q4 in which usage will benefit from both seasonal factors and onetime factors such as traffic engendered by our election. Fastly should also begin to see some noticeable revenues for the many new customers that became clients in Q2, coupled with its sales of new modules to existing customers such as Cloud Optimizer and the company’s DDoS Protection solution. While the company wasn’t entirely explicit, it seems to me as though the guidance provided assumes that a few companies such SHOP which had huge usage growth quarters will flatten out some over the balance of the year, and that the company will see revenues from TikTok decline or even disappear. Just given how the math works, I think it would be difficult to envision a scenario in which Fastly might have forecasted for stronger growth than the company is guiding. 19% sequential growth (its growth in the just reported quarter) per quarter will double a sum in the course of a year. I do not think there is anyone who has actually imagined that Fastly revenues could double in a year. So, the issue about inadequate guidance is really a bum rap.
Neither I or anyone else is in a position to know how the issues with TikTok will work out. Will TikTok reach a deal with MSFT? Will it continue to be able to offer its service in the US and in other geos. Will it be able to restart its effort in India where it has been banned for essentially the same reasons as it might be expelled from operating in the US.
For what it might be worth, I think that the current guidance provided by Fastly includes an assumption that it loses a significant component of revenue from TikTok which will limit top line growth. But the trends toward edge computing and this company’s competitive position is such that I expect it will be able to replace any lost revenue from TikTok and will most likely continue to exceed forecasted revenue levels.
Edge Computing-How does it work and why is Fastly the leader?
While edge computing has been around for some time now, it has become more important in a digitally oriented world. Basically, edge computing is a paradigm in which both compute resources and storage are located at the edge of a network where users actually need it, rather than in the center. The concept is all about saving both bandwidth and response time. The buyers of edge computing solutions want to minimize the need for long distance communications between clients and a central server. A high-level view of edge computing can be had in the linked article: What is edge computing and why it matters
The opportunities in edge computing can be hard to over emphasize. According to the article linked here, the space is forecast to reach $43 billion of revenues, with a CAGR of 37% through 2027: Edge Computing Market Worth $43.4 Billion By 2027 Sometime in writing articles, I find myself doubting the methodology and results of the surveys I cite. In this case, there is far less ambivalence. From my perspective, and simply put-No edge, no digital transformation. The old saw about you can’t get there from here is alive and well in considering why edge is becoming such a standard paradigm-if a user wants to achieve decent performance from a new digital transformation, then the use of edge seems inevitable.
To reiterate, the reason for that kind of growth being forecast in the linked article is simply that it is impossible for many of the digital transformation initiatives to exist without utilizing edge computing. The most obvious case is, of course, that of Shopify itself, which never would be able to support the spike in transactions on its network had it not deployed an architecture built on edge computing.
Fastly is still a small company in a space with lots of giants such as Akamai (AKAM). Why is it likely to be successful? To paraphrase the famous slogan of one of our former presidents, “it is the architecture, stupid.” If you want something more substantive to chew on, but still at a relatively high level, I recommend the following article: Fastly Edge Compute Explained – Software Stack Investing There are a number of salient points in this article, but one of those that is straightforward enough relates to an analogy between the competition between convenience stores and supermarkets. Fastly’s architecture is essentially based on supermarkets, so to speak.
From time to time when I write articles, readers wonder why it is that I tend to wind up writing about high-multiple, high-growth names. One reader wondered, in a different context, why I might recommend Atlassian (TEAM) at its elevated EV/S ratio. The issue to me, is that these newer companies have developed something unique and different. I probably wouldn’t choose to recommend a buggy whip vendor with a small line of automobiles, when I had the choice of recommending a company in the forefront of automobile development and manufacture. I have quoted here a relatively lengthy passage from the linked article, and some readers will not find it useful. But if a reader wants to know why Fastly has won, and should continue to win, and why the growth rate of Fastly is likely to remain elevated for some time now, this might be of most interest: Just to make clear, the legacy CDN’s are the vendors such as Akamai who are the largest vendor in the space, but also the most vulnerable to competition from Fastly.
“ While a convenience store is generally closer to a person’s home, it has a limited set of items for sale. If the person drives a few more miles to the supermarket, they could get all of their groceries in one trip. In this case, convenience stores represent the approach taken by the legacy CDN’s with many local POPs and the supermarkets represent Fastly’s approach with fewer, larger ones.
This is why Fastly still has a limited number of POPs across the globe (72 total as of June), versus thousands advertised by competitors. Fastly POPs are thoughtfully placed at network crossroads, where they provide proximity to geographic regions, but have the storage capacity to enable much higher hit ratios for user requests.
Additionally, Fastly chose to utilize SSDs (relatively new at the time) to store the cached data. SSDs are more expensive than standard disks, but offer much faster retrieval times, on par with RAM. Based on Fastly’s research, a typical hard drive could perform approximately 450 IOPS when reading and 300-400 IOPS when writing (in a test with 4kb files). The SSDs Fastly uses, however, execute somewhere in the region of 75,000 IOPS reading and 11,500 when writing. Making this design choice further contributed to reducing Fastly’s average response times, by making data retrieval within the POPs super fast (furthering the supermarket analogy, like having all your groceries sitting at the front door). At the time, each server in a POP had 384 GB of RAM, and 6 TB of SSD space ( https://www.cnn.com/2020/08/06/business/fastly-stock-falls-tiktok/index.html made up from 12x 500 GB drives), and each CPU has 25 MB of L3 cache. A typical POP had 32 machines spec’ed like this.u
But Fastly didn’t stop there. They also wrote a custom storage engine for their POP servers to bypass the file system and squeeze every last drop of performance out of the SSDs, cramming as much data on them as possible. They employ various algorithms to keep commonly used data in the 384 GB of RAM, making it even faster. For some assets, such as “Like” or “Share” buttons that never change and are requested millions of times a second, they go even further by serving them out of the L3 cache directly from the processor.
As a third problem with existing CDNs, Fastly examined the experience for developers and engineers within customer organizations who were responsible for delivering the application. These individuals were frustrated by a lack of tools and controls to manage changes to the content they were delivering. Common complaints were that legacy providers required technical support personnel to roll out changes and that purging content could take hours. Additionally, they had limited ability to programmatically cache dynamic content or run custom logic on Fastly’s edge nodes to evaluate user requests. In response to this, Fastly added programmability to content control through the Varnish Configuration Language. Varnish is an open source web accelerator designed for high performance delivery. Fastly uses a customized version of Varnish 2.1 and Fastly engineers have continued to contribute to the general open source project.”
Another key opportunity that should keep Fastly growing is their new Compute@Edge platform. It is in beta now, and it will not produce revenue until 2021. Compute@Edge, other than being an awkward name to type, is an architecture designed to enhance personalization and security at the edge. When many investors think of Edge security today, they focus on Z-Scaler (ZS) and Cloudflare (NET). Fastly really doesn’t compete with those names in any significant fashion and both of those companies are and will be leaders in their space. But Compute@Edge will give the company a presence of some strength in the web security space and could provide a substantial increase in the company’s growth rate. Over time, there will be some convergence in terms of functionality between what all 3 vendors offer to users.
Currently, to some extent, Cloudflare’s Worker’s product is designed to facilitate what is called serverless computing, a solution that will also be part of Compute@Edge. Just to be clear, I do not think it is necessary to bash Cloudflare’s approach in terms of serverless computing, in order to forecast that Compute@ Edge will be a successful entrant in that space . It will be a large enough market and there are other, though lesser vendors who will compete. There is no reason why investors can’t allocate funds to both of these names in a portfolio, and I haven’t any strong preference at this point between the two. Just for clarity, I own neither of these names at the present, but I expect I will return to a position in Fastly shares in the future.
There are going to be many significant use cases and new paradigms enabled by the Compute@Edge platform-at this stage, suffice to say that it is a key consideration in evaluating the company’s growth potential over the coming years.
Fastly has lived up to its name both as a company in terms of the technology it offers and as a stock. After appreciating almost 6X in little over 3 months, it has pulled back sharply, losing 35% of its value since the day before it announced earnings on August 5th. Is that enough of a pullback or are the share still to high.
I am not trying to write a trading column here. The fact is that in the last few weeks, investors have moved noticeably to a risk-off stance, and have sold the higher valued, higher growth names that were of interest from the middle of March. Just how long that trade will persist is not a question I can answer, and I have I have little doubt that until investors return to what some call a “risk-on” trade, Fastly shares will probably lag.
I also am not in a position to evaluate the realities of the denouement of the TikTok situation as it might pertain to the relationship it has with Fastly. It most recently has been a 12% customer. Will the business Fastly enjoys from TikTok purchases plummet, stay consistent with recent results, or something else. And if TikTok winds up selling its US business to Microsoft, how that might impact revenues for fastly generated by TikTok (although, it should be noted that Microsoft is already a customer of Fastly) is not knowable.
It is my belief, based on the latest interview on the subject from Fastly’s CEO, that the current forecast provided by Fastly is based on some significant degradation in the business it will derive from TikTok. That said, some negative announcement regarding TikTok and its relationship with Fastly will almost surely drive the shares in the short term.
As mentioned earlier, Fastly’s EV/S is almost 25% below the average valuation for a company with low 40% growth. I have tried to point out why I believe that the company will be able to sustain that growth rate. I might further point out, that what I believe to be the closest analog to Fastly, Cloudflare, has a valuation that I currently estimate to be more than 27X, or 42% above the EV/S I ascribe to Fastly. Both companies are near cash flow break-even.
I think investors in a portfolio of high-growth IT names will most likely want to include Fastly as part of their holdings. Great technology, strong space, a decent business model all combine to make this a desirable investment, in my opinion. The implosion in valuation provides investors with a reasonable entry point, although the uncertainty with TikTok isn’t likely to be quickly resolved.
My advice-start a position at current levels-just shy of $77. Scale into the name if the opportunity is presented, or gradually add more.
Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in FSLY 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.