Cloudflare’s capex heavy business model

[wsm007]

I’ve been worrying for a while about the fact that Cloudflare has a fairly heavy capex model and needs to build, maintain and upgrade a global network in order to keep up with the demand of their customers. Whereas most SaaS companies can essentially determine where they want their margins to land, by curtailing spend, Cloudflare has to keep spending on its network in order to maintain it’s service levels. Much like a telco has to…So I decided to delve a bit deeper into the capex dynamics, especially in light of the concerns raised by some about Cloudflare’s lack of progress towards profitability.

Capex as % of revenue

Last quarter’s capex as a % of revenue for some of the companies in my portfolio:

S - 0.6%
BILL - 0.8%
DDOG - 0.9%
SNOW - 1%
UPST - 1.1%
ZS - 2.4% (last 6 months)

And here is NET:

	Q1	Q2	Q3	Q4
2019			16%	15%
2020	13%	19%	13%	8%
2021	16%	9%	17%	15%
2022	12%			

So where most of our SaaS companies have extremely light capex spend - around 1% of revenue - Cloudflare has to spend around 14% of revenue (average over the past 3 years) on an ongoing basis. The CFO said recently that they will aim to have this at around 11% long-term, but for the next couple of quarters, we’re in for 12-14% which they state is “elevated”. However it’s not elevated in my mind: it’s in line with what they’ve been doing for the last three years. Given their business model, traffic growth patterns and ambitions I would argue that they will need to continue to invest in their network at these levels, or even higher, basically forever - like telco’s need to. It won’t suddenly come to a stop and is therefore part and parcel of their business model.

Depreciation as % of revenue
DDOG - 1.5%
BILL - 1.9%
S - 0.9%
ZS - 3.9%

And NET:

	Q1	Q2	Q3	Q4
2019			8%	8%
2020	8%	8%	8%	8%
2021	7%	7%	7%	7%
2022	7%			

Again, where capital-light SaaS companies’s operating margin is relatively unaffected by depreciation - at between 1% and 2% of revenue, Cloudflare’s operating margin is depressed by 7-8% on a pretty consistent basis as their capex spend gets expensed.

So what?

If Cloudflare had a capex-light business model and profile similar to other SaaS companies (say at about 1% of revenue), they would have been free cash flow positive for all quarters except one last year, and positive for the full year. They would have generated $43m of free cash flow last year, in stead of negative $43m.

And they would have shown operating leverage in stead of hovering at around zero margin. If their depreciation charge was sitting more in line with other SaaS companies, at around 2% of revenue, operating margin would have been positive from Q3 2020 onwards and would have steadily increased to around 7% in the last quarter.

This is not to say there isn’t a big free cash flow bonanza out there somewhere in the future, but it does mean that operating leverage and especially free cash flow generation is more difficult to achieve, and is farther out in the future than SaaS companies without heavy capex requirements. Cloudflare’s business model dynamics seem to me to be more akin to a capex heavy telco than a capex light SaaS company. And in this environment that may matter more than it used to.

-WSM

[FreePuncher]

115%
116%
119%
123%
124%
124%
125%
127%
These are the DBNRR for NET for the past 8 Quarters. They continue to attract more customers, keep them and their customers keep adding to NET’s revenue stream. If this is what CAPEX does for a company, I do not mind the CAPEX intensive model.

They have $1.725B cash and equivalents and $1.433B debt. Plenty of cash to continue model and absorb the negative FCF which is listed below for the last 8 quarters.
$-20.2m
$-17.85m
$-23.73m
$-2.22m
$-9.77m
$-39.7m
$+3.03m
$-68.8m

[catsunited]

WSM,

Yes that’s right. in contrast with capital-light SaaS, NET’s FCF have trailed their operating margin by 10 pts a year historically. That’s what is called “low quality” earnings. Contrast that with capital-light SaaS where the FCF exceeds their operating margin by 10-20 pts a year.

At some point far far away in the future, as NET’s capex comes down as they mature, the two will converge. But as I wrote in my other thread, if they intend to keep their operating margin at breakeven for years (as they say they will) and continue to spend wantonly on capex like drunken sailors, they will keep burning cash (say, an FCF margin of -10% to -15%) for the next few years. Their TTM FCF margin is -15%.

and this company is not debt free. they have debt of $1.4 bil and cash of 1.7 bil. Given their capex-heavy and cash burning model, they will need substantial cash on hand. At some point, they will then have to come to the market to raise equity - and with their depressed stock price, that will only happen with substantial dilution.

---

FreePuncher,

Are you still thinking in terms of revenue growth without profits ? Sorry but The ship has long sailed already.

Cats

[buynholdisdead]

and this company is not debt free. they have debt of $1.4 bil and cash of 1.7 bil. Given their capex-heavy and cash burning model, they will need substantial cash on hand. At some point, they will then have to come to the market to raise equity - and with their depressed stock price, that will only happen with substantial dilution.

To be clear the debt is convertible senior notes and can in the companies discretion be redeemed with cash or shares or a portion of each. The senior notes that are redeemable in 2026 are 0% coupon. Yes not interest. Their cash per share has gone up every quarter and is now sitting at $2.26. Their total paying customers accelerated last quarter and was up 10 percent sequentially and 29.28 percent YoY.

They are building out an unique network and it is something that the big cloud companies can’t do because how they have been designed. With Net everything works across the whole network. With the others, Azure, Google, AWS they have it set up into regions, not everything works in every region, At least that is my understanding. Also here is one of the statements they made on the last earnings report.

One last story. A Fortune 1000 gaming company signed a $3.3 million, 3-year contract. I love this story. They were using AWS, but found their security couldn’t prevent the attacks they were seeing. After struggling to keep their application online, AWS’s team eventually told them, “You should just use Cloudflare.” And so they did. [indiscernible] that let them and lots of other customers know that R2 is progressing to open beta next week, and we expect we’ll be able to save them lots of AWS egress fees as well.

For anyone who wants a secure network it seems they will be moving to Net.

Andy

[bnh91]

Cats,

Cloudflare is going after a very large TAM in the multiple 100s of billions of $. I don’t view the Capex they are spending at this time to improve their network performance as a bad thing. This only helps improve the moat of the business (which is very strong IMO). There are a couple of other high quality extremally valuable businesses (AWS/Azure) that have a similar, if not heavier, capex profile as Cloudflare’s. These two have very little chance of being disrupted partially because of this.

Below is a quote from the CFO in the Q4 earnings call. They have been very transparent about this. So I am not sure why anyone/the market would be surprised by the FCF print during this this Q…

We are very pleased to have achieved our first free cash flow positive quarter as a public company. Going forward, we expect to see some variability in the first half of 2022 as we continue to invest in building refurbishments and network capacity, but we do expect to return to positive free cash flow in the second half of 2022.

So FCF will take a hit during 1H and return positive in the 2H. It isn’t a guarantee Cloudflare will need to raise $, but they just raised $1.125 convertible debt offering at 0% interest last year that was oversubscribed. This company is not going to have any trouble raising debt at fair/favorable terms. I understand it is very in vogue to rip apart any business that isn’t returning capital/printing FCFs right now, but lets take a deep breath and maybe a few steps back from the ledge.

Bnh
Disclosure: Long NET

[catsunited]

Andy,

“To be clear the debt is convertible senior notes and can in the companies discretion be redeemed with cash or shares or a portion of each”

The conversion price of their debt is $193 (vs $56 now) due 2025/2026. if their stock price doesn’t go above that, they then have to repay the debt in cash.

Generally Cloudflare is a very binary investment case - reflecting in the “best-in-class” volatility of its stock price.

[buynholdisdead]

The conversion price of their debt is $193 (vs $56 now) due 2025/2026. if their stock price doesn’t go above that, they then have to repay the debt in cash.

That is not the way I read it Cats. Here is the wording off of their 10Q. The bolded part is where I am directing your attention.

August 2021, the Company issued $1,293.8 million aggregate principal amount of the 2026 Notes in a private offering to qualified institutional buyers pursuant to Rule 144A promulgated under the Securities Act, including the initial purchasers’ exercise in full of their option to purchase an additional $168.8 million aggregate principal amounts of the 2026 Notes. The total proceeds from the issuance of the 2026 Notes, net of initial purchaser discounts and commissions and debt issuance costs, were $1,274.0 million.

The 2026 Notes are senior unsecured obligations of the Company and will mature on August 15, 2026, unless earlier redeemed, repurchased, or converted, and are governed by the terms of the indenture dated August 13, 2021 (the 2026 Indenture). The 2026 Notes are 0% convertible senior notes and therefore do not bear regular cash interest.

The 2026 Notes are convertible at an initial conversion rate of 5.2263 shares of the Company’s Class A common stock per $1,000 principal amount of the 2026 Notes, which is equivalent to an initial conversion price of approximately $191.34 per share, subject to adjustment upon the occurrence of specified events in accordance with the terms of the 2026 Indenture.

The 2026 Notes may be converted at any time on or after May 15, 2026 until the close of business on the second scheduled trading day immediately preceding the maturity date. Holders of the 2026 Notes may convert all or any portion of their 2026 Notes at their option at any time prior to the close of business on the business day immediately preceding May 15, 2026 only under the following circumstances:

(1) during any calendar quarter commencing after the calendar quarter ending on December 31, 2021 (and only during such calendar quarter), if the last reported sale price of the Company’s Class A common stock for at least 20 trading days (whether or not consecutive) during a period of 30 consecutive trading days ending on, and including, the last trading day of the immediately preceding calendar quarter is greater than or equal to 130% of the conversion price on each applicable trading day;
(2) during the five business day period after any five consecutive trading day period in which the trading price per $1,000 principal amount of the 2026 Notes for each trading day of the measurement period was less than 98% of the product of the last reported sale price of the Company’s Class A common stock and the conversion rate on each such trading day;
21
Table of contents
(3) if the Company calls such 2026 Notes for redemption, at any time prior to the close of business on the second scheduled trading day immediately preceding the redemption date; or
(4) upon the occurrence of specified corporate events.
None of the circumstances described in the paragraphs above were met during the first quarter of 2022.
In addition, if the 2026 Notes are converted prior to the maturity date following certain specified corporate events or because the Company issues a notice of redemption, the Company will increase the conversion rate for such 2026 Notes converted in connection with such a corporate event or during the related redemption period, as the case may be, in certain circumstances set forth in the 2026 Indenture.

Upon conversion, the Company will pay or deliver, as the case may be, cash, shares of the Company’s Class A common stock, or a combination of cash and shares of the Company’s Class A common stock, at the Company’s election. It is the Company’s current intent to settle the principal amount of 2026 Notes in cash.

The Company may not redeem the 2026 Notes prior to August 20, 2024. The Company may redeem for cash all or any portion of the 2026 Notes (subject to the partial redemption limitation (as defined below)), at its option, on or after August 20, 2024, if the last reported sale price of the Company’s Class A common stock has been at least 130% of the conversion price then in effect for at least 20 trading days (whether or not consecutive) during any 30 consecutive trading day period (including the last trading day of such period) ending on, and including, the trading day preceding the date on which the Company provides notice of redemption at a redemption price equal to 100% of the principal amount of the 2026 Notes to be redeemed, plus any accrued and unpaid special interest to, but excluding, the redemption date. If the Company elects to redeem fewer than all of the outstanding 2026 Notes, at least $100.0 million aggregate principal amount of 2026 Notes must be outstanding and not subject to redemption as of the relevant redemption date. No sinking fund is provided for the 2026 Notes.

If the Company undergoes a fundamental change (as defined in the 2026 Indenture), holders of the 2026 Notes may require the Company to repurchase for cash all or any portion of their notes at a repurchase price equal to 100% of the principal amount of the 2026 Notes to be repurchased, plus accrued and unpaid special interest to, but excluding, the fundamental change repurchase date.
Based on the closing price of the Company’s Class A common stock of $119.70 on March 31, 2022, the if-converted value of the 2026 Notes does not exceed its principal amount. The remaining life of the 2026 Notes was approximately 53 months as of March 31, 2022.

So if I am reading that correctly cats, and I may not be, but the company can elect to pay it off with cash and shares depending on the share price at the time they come due. While the company wants to pay cash the option is theirs. Plus this is money loaned out at a 0 coupon. How great is that for the company?

Andy

[CrazyCzech]

Andy

that is not how convertible bonds ever work. Convertible bonds are a form of hybrid debt which combines borrowing money with defacto selling call on company’s stock at a certain strike price - if price of stock is above this price, holders can redeem the bond for stock or cash. Cats is right that IF the price of the stock is below conversion price on the bond at maturity, holders will be paid in cash the original face value of the bond. What companies get is cheap source of funding (0 in this case). What holder of the bonds get in return is a call option on the company’s stock.

For the record, NET said very clearly (and stated in current earnings presentation) they expect to return to cash flow positive in second half. They also advertised that first half will be cash flow negative during Q4 conference call so this, to me, is not a surprise. I certainly expect them to generate lots of cash. They said they will, they did and continue to say they will. I have no reason to doubt this management.

This topic is addressed in their most recent investor presentation very clearly. I suggest anyone interested to look at it.

[rliu]

Hi all,

Because this is OP, so I just simply explain it and shortly. CB is a bond+option, after few months they may sell their options separately or directly asked turn it into shares. So I believe that’s not a issue here. The issue here is when they need more money, they might have to issue a [friendly exercise price] CB.

[buynholdisdead]

This topic is addressed in their most recent investor presentation very clearly. I suggest anyone interested to look at it.

Thank you CrazyCzech, do you mind posting a link of what your are talking about? I am interested in looking at it.

Andy

[FinallyFoolin]

I’d like to also point out that on the Q4 earnings call, there was a question on how they handle their CapEx. They stated that their model allows for spending on CapEx in an on-demand style rather than “if you build it, they will come”. That tells me that if we see CapEx going up, the demand for that is also going up. Note though that it might be a quarter or two before the revenue hits the top line. I say this because of things like they said in the most recent call about some stuff they’re working on with some very large clients. I suspect their dev focused edge network allows them to build out special things the likes of ZS & PANW can not. Then they can take these things and pitch them to other clients.

In any case, the CapEx is part of the moat. I don’t see any competitor, even hyperscalers, that can provide companies to do similar things. In the security space, sure ZS is best of breed, but now NET can tout their ability to thwart the best of the best hackers via their work in Ukraine. This levels that playing field a bit. In addition, they can say that your developers can do things within this same edge network that is also providing security that ZS & PANW simply can not do. I think this is why in the most recent call, they stated a few times about companies favoring vendor consolidation.

One of my 2 cents on NET.

Oh… and I’m STOKED that Prince responded on twitter to my response on his. He asked what they should announce tomorrow (which is now today) and I said D2 with d=database (R2-D2). He said, “Not tomorrow”; perhaps that means later this week and if so… look out! I’d expect it to be a document database similar to Mongo and selfishly would prefer it to be Atlas built into their edge but that’s not really their MO. In any case, DB build into the edge would allow for much closer competing v the likes of the hyperscalers. ok ok, this went off the main topic but its the 2nd of my 2 cents.

[SaulR80683]

I also don’t like companies that don’t want to make a profit, but have agreed that expanding their customer base now is a worthy alternative as the customers that they sign up will be recurring revenue pretty much forever.

While thinking about this I remembered exiting AMZN in the $900’s about 5 years ago because nowhere did Bezos even mention any interest in making a profit (they were barreling along at about breakeven plus or minus one percent). However they hit $3500 before this tech crash came along!!! And started making a profit along the way.

And Morningstar, which always loved all my companies, but always said they were grossly overvalued, now shows Cloudflare as hugely undervalued, even though they also would like to see more emphasis on the bottom line.

Best,

Saul

[Robear1020]

Re: “And Morningstar, which always loved all my companies, but always said they were grossly overvalued, now shows Cloudflare as hugely undervalued, even though they also would like to see more emphasis on the bottom line.”

Saul, thank you for sharing this!!! As I read through this thread I thought that you might follow WSM’s lead and sell off a large portion of your Cloudflare like he apparently did (going from 13% in March to 2.4% in April). That might have been a good move for the time but now certainly feels like a good time to hold (or add) especially if it’s hugely undervalued.

By the way, this article from yesterday also sounds pretty encouraging…
“Cloudflare Just Made Every Application on the Internet More Programmable”
http://newsfile.refinitiv.com/getnewsfile/v1/story?guid=urn:…

Thank you for building an incredible community!

Best regards,

Robear

[WillO2028]

Peter Offringa updated his recent post reviewing Cloudflare’s Recent Quarter results on May 10th at https://softwarestackinvesting.com/first-look-at-cloudflares……

He updated it at the bottom of what he’d previously written specifically to give his opinion, as always supported by superior analysis IMO, regarding Cloudflare’s being Cap Ex heavy relative to Other Saas companies.

At this point I tried to find something in what Peter wrote to add here as a nugget of his wisdom and I fell short. The entirety of what he wrote was all just too much to the point. Points I think I should not try to truncate.

I think, as I indicated with my choosing the word ‘superior’ as an adjective to describe Peter’s analysis shows, reading his work on the subject, it is absolutely worth the time to read in its entirety, for anyone here how owns or is thinking to buy Cloudflare (now would be the time, IMO).

Thank you Mr. Offringa,

Jason
(I maintain ~13% of my portfolio in Cloudflare and am looking to add)