Fastly's Portfolio Fit

Long FSLY, for now as I’m still undecided on what to do. - Griz

I believe this statement perfectly reflects the lingering FSLY discomfort in most of these posts. Fastly threw us a curve last week, and I think the board has done a tremendous job working through it. The story has clearly changed even if opinions differ on exactly how much. It seems most of us have landed somewhere on the hold/trim/sell spectrum (any buyers here?). Ultimately, hindsight will make it clear what we should have done. It always does.

Rather than rehash the announcement itself, I thought it might be worthwhile to outline the thought process the led to my decision to sell. First, my initial thoughts on the news can be found here: https://boards.fool.com/a-couple-of-you-reached-out-on-fsly-…. Second, I agree 110% with every point Bear made here: https://boards.fool.com/fsly-and-updating-beliefs-34643625.a….

Like Saul, I tend to group my holdings by conviction level. As noted in my linked post, my conviction level on FSLY changed. I might not know the exact level yet, but I am sure it is lower. The question becomes do I have enough information to act on it?

Stealing Bear’s format:

Prior Belief – FSLY has an intriguing new product in beta. As I am waiting for that to hopefully kick in, the company is supporting itself with a core business at mid-to-high 50’s growth (using the prior guide with a reasonable beat and not last Q’s 62%). I know about the low enterprise growth and concentration, but FSLY seems to be managing it so far. Besides, the Signal Sciences acquisition should help here. The prior CEO moved to CTO, which seems to better fit his skill set. The new guy is green but seems OK otherwise. FSLY has a pretty attractive thesis.

Update – FSLY has an intriguing new product in beta. As I am waiting for that to hopefully kick in, the company is now supporting itself with a core business at low-to-mid 40’s growth (using the new guide with a lesser beat). That’s a big drop off. The low enterprise growth and concentration might be more of a concern than I thought. If nothing else, it puts more pressure on Signal Sciences contributing immediately. Fixing this misstep will be handled by a new CEO and a former CEO who prefers the technical weeds instead. This isn’t a terrible thesis, but it is one with considerably more risk.

My portfolio entering that fateful day looked like this based on allocation size:

ZM
CRWD
DDOG
FSLY
NET
DOCU
OKTA
TWLO
PTON (top watch list name)

The second my FSLY conviction changes it becomes time to reassess its spot in my portfolio. This is the same process I used in June when my FSLY conviction rose high enough to add it to my portfolio in place of WORK. In this case:

FSLY vs NET – This comparison has been beaten to death, and they were close for me with Fastly being a step ahead coming into this comparison. NET recently followed its Birthday Week with Zero Trust Week. Between the two events, Cloudflare released somewhere in the neighborhood of 74 million new products and enhancements (give or take, of course). Meanwhile, FSLY shot itself in the foot and was surprisingly quiet with damage control. So, NET > FSLY.

FSLY vs DOCU – DOCU is trending strongly upward with next quarter almost certainly delivering 50%+ growth and favorable profitability metrics. Meanwhile, FSLY shot itself in the foot and was surprisingly quiet with damage control. So, DOCU > FSLY.

FSLY vs OKTA/TWLO – OKTA and TWLO are both mid-to-high 40’s grower with higher run rates than FSLY, more stable profitability profiles and MUCH more established leaders. I previously liked FSLY’s growth prospects better, but that advantage is no longer certain. In fact, TWLO just pulled off a big acquisition the market loved and might put itself back in the 50%+ growth category next quarter (hmm, I may need to revisit TWLO). Besides, have I mentioned that FSLY shot itself in the foot and was surprisingly quiet with damage control? So, OKTA/TWLO > FSLY.

That puts FSLY’s conviction level at the bottom of my portfolio and forces a showdown with PTON.

FSLY vs PTON – Peloton just posted a year of 99.6% growth with strongly positive EBITDA and cash flow. It estimates 99.9% growth this year (which will almost certainly be beaten) along with a stronger EBITDA margin. I fully acknowledge PTON selling consumer-based widgets makes it a bit tougher to follow and likely a lot more fickle than a software company. However, it also has an impressive subscription component growing 100% at close to a $500M run rate. FSLY is…well…something much more uncertain than it was before. So, PTON > FSLY.

Staying disciplined to how I am trying to run my portfolio, I have a decision to make right here. This reassessment just earned PTON a spot in my portfolio. Notice this development by itself is not just a FSLY decision but involves PTON as well. PTON just cleared my inclusion bar. That would not have been the case if FSLY had stopped higher up. Now the question becomes whether I want to hold myself to eight positions or increase to nine by say swapping some FSLY into PTON. After thinking about it, I decided to stick with eight. Once I make that decision, FSLY automatically exits my portfolio. It doesn’t matter whether I’ve got FSLY’s new market status exactly right. All I need to know is I no longer have enough conviction to own it in my 8-stock portfolio. Nothing personal Fastly, and I’ll gladly watch you from afar. That being said, don’t let the door hit you in the ass on the way out.

To be fair, this all happened VERY quickly during the first 30 minutes or so after market close Wednesday. I sold after hours immediately upon making the decision, giving me ~9% cash. Conveniently enough I had just finished my PTON breakdown, so I already felt good about its spot at the top of my watch list. I also implicitly trust much of the work done by others on this board on PTON’s behalf (isn’t crowdsourcing great?). My initial comfort-level blink on PTON puts it in the 5-6% allocation range. With PTON sitting at all-time highs, I wasn’t entirely sure about biting off a 5%+ chunk all at once. So, knowing I would double check my due diligence this weekend, I entered limit orders Thursday at 3%, 5%, 7% and 9% below the high just in case I could grab some a little cheaper. The first hit Friday, so I now have a 1.6% position to start. The other orders remain intact heading into Monday. I plan on filling out the position over the next day or two even if I end up buying a little higher. The remainder most likely finds its way into some combo of DOCU, TWLO and potentially another name or two.

I know this post trends more toward portfolio management than analyzing an individual company, but I do believe it is relevant to the FSLY topic at hand. I hope it helps. I know it helped me to write it out.

Hope everyone is having a good weekend.

138 Likes

Stock novice, there is a lot of talk about edge computing.

There was also a lot of talk about ZPA at ZScaler.

There was also a lot of talk about Flex at Twilio.

There was also a lot of talk about Xi at Nutanix.

In my case there was a lot of hope for international at Roku.
That is clearly becoming a multi-year thing and not a this year thing (they entered U.K. December last year for the first time).

There was a lot of talk about Apple TV.

I could go on of course. My point is in my book FSLY is a next generation CDN growing at 35% a year that got bid up big time because growth went to 62%, and they’re working on edge computing. Until FSLY sees success at edge it’s a hypothetical. Just like every single biotech company with a product in Phase III. Not only do you have to have a working product (or FDA approval) but adaption in the marketplace.

I can buy hypothetical growth stocks all day because there is an abundance of stories out there. And unlike Fastly, they didn’t see a 5x price increase in a few months because growth temporarily spurted.

My question is how did it temporarily spurt? Who were these customers and what were they doing that drove the growth? Was it obvious it was temporary? We may never know.

7 Likes

stocknovice wrote, in part:
Update – FSLY has an intriguing new product in beta. As I am waiting for that to hopefully kick in, the company is now supporting itself with a core business at low-to-mid 40’s growth (using the new guide with a lesser beat). That’s a big drop off. The low enterprise growth and concentration might be more of a concern than I thought. If nothing else, it puts more pressure on Signal Sciences contributing immediately. Fixing this misstep will be handled by a new CEO and a former CEO who prefers the technical weeds instead. This isn’t a terrible thesis, but it is one with considerably more risk.

I think that’s a fair assessment. To pull from a post of mine back in Fastly’s glory days (early August): https://discussion.fool.com/let39s-start-by-clearing-things-up-i…

I’m actually not invested in Fastly for its CDN. Yes, it’s a great CDN and has advantages over other CDNs, especially on the push side, but Fastly’s CDN business alone is not going to provide the kind of high-growth that we look for here. The future for Fastly is Compute@Edge, which is still several months away from launch and the TikTok news has no bearing on that.

What we’ve seen is that the legacy, mostly commodity, CDN business isn’t growing like we had seen in Q2, but the company is making the right moves for its future edge computing business. Such moves such as acquiring Signal Sciences and just recently, hiring the entire Wasmtime team from Mozilla.

One thing no-one here has discussed (but it was mentioned in some reddit threads), is: What is the effect of the Wasmtime team acqui-hiring by Fastly on Cloudflare? Cloudflare has had WASM support for a while now. Here’s what Cloudflare said about WASM just a couple weeks ago:

WASM is closer to machine code, and thus faster than JavaScript (JS) or JIT. Compiler optimizations, already done ahead of time, reduce the overhead in fetching and parsing application code. Today, WASM offers the flexibility and portability of JS at the near-optimum performance of compiled machine code.

AI models have notoriously large memory usage demands because configuring them requires high parameter counts. Cloudflare already extends support for WASM using their Wrangler CLI, and we chose to use it for our exploration. https://blog.cloudflare.com/exploring-webassembly-ai-service…

Does this mean that Cloudflare will de-emphasize support for WASM? They’ve always touted their JavaScript support (since JS is probably the most popular programming language today), but since there are performance penalties Cloudflare has to offer alternatives. Is Fastly’s acquisition a business smart move as well as a way to get additional technical expertise and double-down on future performance advantages?

I do understand people trimming or selling out of Fastly completely. It makes sense, depending on your investment thesis and understanding of the technologies and especially in relation to other opportunities. Even Cramer the day of the pre-announcement said that Fastly was likely to decline much further, and was right. If you sold, congrats. If you didn’t sell, think about what your investment thesis is moving forward from here.

18 Likes

What is the effect of the Wasmtime team acqui-hiring by Fastly on Cloudflare? Cloudflare has had WASM support for a while now.

Does this mean that Cloudflare will de-emphasize support for WASM? They’ve always touted their JavaScript support (since JS is probably the most popular programming language today), but since there are performance penalties Cloudflare has to offer alternatives. Is Fastly’s acquisition a business smart move as well as a way to get additional technical expertise and double-down on future performance advantages?

  • Smorg (bolding mine)

This is some underrated thinking. And this is an underrated post: https://discussion.fool.com/stocknovice-wrote-in-part-update-fsl…

Thanks, Smorg. Obviously the hype-train has backed down on FSLY but not on NET, and I get it – NET hasn’t disappointed yet, and also already has a lot of other irons in the fire than just CDN. Still, I’d love to know how this move by FSLY could close that gap, or otherwise negatively effect NET and its non-CDN offerings. I’m not able to make these connections or understand the tech, so I appreciate you mentioning it and would love to hear more.

Bear
Still about 5% NET and 1% FSLY, but thinking those should perhaps converge. Or both go away. But humbled by my real lack of ability to understand this space.

10 Likes

Obviously the hype-train has backed down on FSLY but not on NET, and I get it – NET hasn’t disappointed yet, …

Some people have a short memory. Not I.

Former Uber chief security officer charged with covering up hack
https://www.datacenterdynamics.com/en/news/former-uber-chief…

This was why Cloudflare was trading cheaper in the first place if I’m not mistaken. Didn’t Twilio also get beaten with the ugly stick for accounting irregularities? Sometimes companies go through things, but better ones will get out of it. Sometimes they won’t.

Dominic

1 Like

That’s the rub though isn’t it? If the investment thesis is that CDN is not a competitive advantage but edge compute is, then FSLY is a story stock. It could work out very well for those invested in it, or it could go the way of TWLO (after the Sendgrid acquisition) and do nothing for a long period of time.

Since edge compute seems so far off in the future in terms of generating revenue (the revenue impact is not being quantified or discussed on even a pro forma basis at NET/FSLY/LLNW) what is the thesis that revenue will again accelerate in the near term? The anemic new customer growth and existing customer spending expansion does not add up to 40% YoY. In the meantime you’re missing out on the companies that will do well for your portfolio for the next year. It seems like a lot of hope to me.

The current performance just doesn’t add up to the 50%+ growing investable cohorts that we have available (ZM, CRWD, DDOG, PTON, TDOC/LVGO, ETSY, etc.). We’ve all mentioned opportunity costs recently and while it’s easy to say that TWLO came back after a patient wait of a year+, just look at the performance of alternatives. In the end it’s still market beating returns (which I think FSLY will do, 40% is still no joke), but it’s not beating a portfolio of the best companies.

3 Likes

Aphalite-
‘It seems like a lot of hope to me.’.

Other than the one quarter usage based bump in revs, which didn’t make a trend, I agree with the above statement about Fastly.

I disagree with your assessment-
‘Since edge compute seems so far off in the future in terms of generating revenue (the revenue impact is not being quantified or discussed on even a pro forma basis at NET/FSLY/LLNW)…’

Net has had Teams out in the wild for over a year and was only giving it away for free until this September. So that’s why I hold NET, anyway.

————

I’m not going to explain myself well about my buying Fastly at $29 and selling at $88 before the said rev bump and why. I’ve already done that, it was written here quoting Tinkers lesson to me specifically about not investing in hope.

I did get back in due to all the tech talk here and Elsewhere.
So not saying I’m immune to the hard wired difficulty had when investing not only money but also time (learning)into a company and how difficult it is when existential change is needed given the realization of a mistake.

2 Likes