Today, 4-16-2021, we add Snowflake (SNOW) as a Spotlight Top Pick but only with an add below price of $200.
We know full well it may never hit that price at that’s OK with us.
The stock is currently trading at $230.
We detail the valuation and choice of an add below in the section near the conclusion entitled “Valuation.”
This places Snowflake in the second circle of this logic diagram (if and only if the add below price hits):
Officially, Snowflake’s corporate profile reads like this:
Snowflake, Inc. provides cloud data warehousing software.
It provides SQL data warehouse, zero management, and broad ecosystem products.
It offers data warehouse modernization, accelerating analytics, enabling developers and monitoring and security analysis solutions to federal government, financial services, healthcare, media and entertainment, retail and CPG, gaming, education and technology industries.
Snowflake’s “Data Cloud” is a global network where thousands of organizations mobilize data with near-unlimited scale, concurrency, and performance. Inside the Data Cloud, organizations unite their siloed data, easily discover and securely share governed data, and execute diverse analytic workloads.
In English, enterprises store their data on the central cloud using Snowflake’s software and from that point forward, it’s nearly impossible to leave.
Unofficially, as we see it, not only does Snowflake act as a data warehouse, but it it also acts as (i) the connector between enterprise data and third party the data analytics tools, and (ii) the first and only semi-platform that may act as not just the intermediary, but the primary enabler of the public cloud to enterprises without them having a direct relationship with one of the three mega caps — Amazon’s AWS, Microsoft’s Azure, and Google’s Google Cloud.
As the connector to third party data analytics tools SNOW reinforces its extraordinarily high switching costs, after all it wouldn’t just be an enterprise moving its data warehouse, but also all of its third party connections.
As the primary point of entry for an enterprise’s cloud management, that switching cost goes from high to nearly impossibly high.
While SNOW itself sits on the cloud infrastructure as built by the giants, some enterprises simply use Snowflake as the interface.
So, while the cloud platforms still earn all the revenue they would normally receive, the bill goes to Snowflake and that means the cloud platforms may not have as strong of a relationship with the enterprise as they normally would. That would make SNOW the first and only intermediary between enterprises and their cloud provider that we have ever seen at scale.
It’s an interesting theory that we hold although it’s not necessarily crucial to the bullish thesis. We do see it materializing in the future in some way, we’re just not smart enough to know how and when. So, we’ll say, “we think this matters,” and leave it at that for now.
As we move through what SNOW does and the thematics behind it, we also do feel that at some point in the future, SNOW will add data analytics capabilities.
Now let’s turn to what it does and what thematics push it forward.
Thematics and What it Does
We said above “enterprises store their data on the central cloud using Snowflake’s software and from that point forward, it’s nearly impossible to leave.”
Officially the company lists its solutions as these:
But for Snowflake it gets even better. This data thing turns out to be a thing — as we noted back in 2015, data isn’t just exploding, it’s exploding on top of an explosion.
There is an explosion of users creating data, an explosion of devices creating data, and explosion of use cases creating data, an explosion of software (and software companies) creating data, and then an explosion in the people and software creating analytics and machine learning models taking the data and asking for more to increase utility.
We have heard others say that data is oil. We think data is air.
Let’s turn to an old CML Pro image of a thematic and then update it to the present day to show not just how ambitious the forecasts were back in 2016, but how much larger they have become.
We’ll look at data from mobile. Back in 2016 we shared this image of mobile smartphone subscription growth projections:
Projections called for nearly 75% growth in mobile subscriptions within six-years through 2022.
If you look at the 5G projection, it called for 540 million by 2022. That number is now forecast to be 1.18 billion by 2022, so more than a double in the original forecast.
Yep, a double from the forecast made just four-years ago.
We could share half a dozen images like this, but instead, we’ll keep it (a little) pithy and just look at the newest projections for data creation. That 2016 to 2021 comparison was intended to be instructive — to be empirical about what we mean when we say “exploding on top of an explosion.”
When commenting on the forecasts surrounding data creation, we said in 2016 “it will be bigger.”
We say again in 2021, when looking at forecasts of data creation, “it will be bigger.”
The break in forecasts accuracy, as we see it, is an underappreciation for the power of machine learning, its ubiquity, and the tautology that more machine learning on data actually creates… more data.
Now, let’s look at other thematics.
First, we look at public cloud services end-user spending worldwide from Statista.
The worldwide public cloud computing market continues to grow and is expected to reach an estimated 362 billion U.S. dollars in 2022. This encompasses business processes, platform, infrastructure, software, management, security, and advertising services delivered by public cloud services.
As cloud usage grows, so room for data expands as does the machinery needed to run data analytics.
Next we look at public cloud application services/software as a service (SaaS) end-user spending from Statista.
Data comes from many places, and software services are data creation machines. The statistic shows the size of the public cloud application services (SaaS) market worldwide from 2015 to 2022. You can note the impact of COVID.
Next we can turn to the end result — the volume of data created, captured, copied, and consumed worldwide.
The total amount of data created, captured, copied, and consumed in the world is forecast to increase rapidly, reaching 59 zettabytes in 2020. The rapid development of digitalization contributes to the ever-growing global data sphere.
Back in 2010 then Google CEO Eric Schmidt said “From the dawn of civilization to 2003, five exabytes of data were created. The same amount [of data] was created in the last two days.”
Well, in 2020 59 zettabytes of data were created, which is 59,000 exabytes.
If “[f]rom the dawn of civilization to 2003, five exabytes of data were created,” then in 2020 five exabytes were created every 45 minutes.
Yeah, the same amount data created from the dawn of civilization to 2003 was created every 45 minutes in 2020.
Like we said, ” exploding on top of an explosion.”
That data needs to be stored and analyzed.
Much of the storing part is going to happen on the central cloud, and on the central cloud, it’s going to happen using Snowflake.
As for the data analytics part — called “big data and analytics revenue,” Statistasees that revenue generating $274 billion by next year (2022).
Finally, Snowflake, like many cloud centric software companies, has the great advantage of being available across the mega cloud providers which not only gives it access to many customers, but most importantly renders the competition from anyone cloud provider somewhat neutralized.
The multi-cloud infrastructure design is by far and away the clear thematic moving forward.
Few mega enterprises will rely on just one provider as a risk mitigation strategy.
Here is a chart from Statista that illustrates enterprise cloud strategy worldwide.
By 2021, 92% of enterprises have a hybrid cloud (companies combining a private cloud with one more more public clouds) or a multi-cloud strategy. These 92% of firms will needs software that is cloud agnostic, meaning that a solution from just Amazon (or whatever), will likely not suffice.
Finally, our last chart from Statista demonstrates how much a problem data management presents.
From 2020 to 2022, managing data in multicloud environments is expected to be the most challenging data management challenge for enterprises, as more than 60 percent of respondents reported this to be challenging or extremely challenging.
There are so many more numbers and charts we could throw out, but it would leave us with the same conclusion: “data isn’t just exploding, it’s exploding on top of an explosion.”
These thematics, from the proliferation of devices from the Internet of Things (IoT), to the cloud, to data creation, to data analytics, all of them are the tailwinds for Snowflake.
And since that’s the case, yes, we are about to see some tongue wagging financial numbers leading to a huge valuation.
More on Snowflake — and Financials
Now we can, at least in part, turn to some of Snowflake’s presented information.
Snowflake’s financial performance has been nothing short of outstanding, as we would expect with the thematics in play.
In the last quarter reported (Q4 2020) the company delivered:
* 116% year over year (yoy) product revenue growth. The company grew 124% for calendar 2020 versus 2019.
* 168% net revenue retention
* 93% of revenue is consumption-based (not subscription based)
* 70% Non-GAAP gross margin per cent
* Total customer growth of 73% and 88% customer growth in those generating over $1 million in revenue.
* The company claims to have 186 of the Fortune 500, up 46% year over year.
* Adjusted free cash flow as a per cent of revenue has climbed from -75% two years ago to -12% last year. For this current year, the company expects to roughly breakeven in adjusted free cash flow.
While it’s easy to gravitate to the revenue growth, let us not be so remiss to see the thematics impacting the company directly.
The net retention rate means that companies are spending 68% more than they did last year.
The company is quite vocal about not running as a SaaS (software as a service), but rather consumption based.
In a world where we can guarantee very little, we can guarantee there will be more data, and the consumption based model is far more lucrative than a subscription based service.
This billing mode of operation allows SNOW to benefit from ever growing data, every second of every day.
Beyond SNOW’s realized $592 million in total revenue, the company announced $1.338 billion in customer commitments, which we get from Remaining Performance Obligations (RPO).
RPO represents contracted future revenue not yet recognized.
Analysts have forecast over $1 billion in revenue for the current fiscal year or 85% growth.
Even further, analysts have forecast 64% revenue growth for calendar year 2022 and 58% for 2023, reaching just over $2.8 billion in revenue by the end of calendar 2023.
Without even getting out our pencil, eraser, and spreadsheet, we see yet more upside to these numbers.
Snowflake actually just went public on 9-16-2020, so about seven-months ago.
The first public traded price was $250, which is below the current price.
We were well aware of Snowflake well before it went public, and it has been a potential Top Pick for some time.
We waited to see the company report two public quarters and to hear calendar 2021 guidance.
We saved 7-months of pedestrian (or negative) returns and now feel comfortable with an add below price for Top Picks.
While we are very long-term focused, in the arithmetic to follow you can at least see our rational for an add below for the tracking portfolio.
The back of the envelope arithmetic we have done works out to this:
* From calendar 2024 through 2028 (five-years), 38% CAGR in revenue yielding nearly $14 billion in revenue.
* Long-term adjusted EBIDTA margin of 35%, or about $4.9 billion in EBITDA by 2028.
* An enterprise value (EV) to EBITDA of 35:1, or 35x $4.9 billion ~ $170 billion market cap.
As of today, the market cap is $70 billion.
That would yield 13.5% CAGR in the stock price over seven-years (in 2028), or ~140% over seven years.
While we are all used to absurd CAGR in stocks, that is not reality in the future.
We see the potential for great things from Snowflake, but still 13.5% CAGR is not quite our long-term goal for Top Picks.
So, we have an add below price for Snowflake that is about 15% below $230, or we’ll call it $200.
The stock may never hit that price and that’s OK with us.
With a $200 stock price, we would see 15%-16% CAGR in the stock price if those rather sophomorically mapped out forecasts come true.
Those are very good numbers. Those would be Top Pick numbers.
With a company priced to perfection we have great risk that “very good” will leave stock investors with “not very good” returns.
Even a greater risk is that mediocre or poor financial performance could leave the stock decimated — let’s see even 50% lower than it’s current market value.
Other risks include:
* Competition – the explosion of exploding data is hardly a secret. There will be meaningful competitors, and they will likely have some impact on Snowflake’s growth.
* Not cash flow positive means… not cash flow positive. While the company has a pile of cash (over $3.9 billion) and little debt (about $200 million), at some point it will need to turn cash flow positive. It does appear that the company is well on its way.
* Some unforeseen event – a change in technology, cloud infrastructure, geopolitical risk, and the “unknown,” as well.
In total, we see the greatest risk to Snowflake coming from an inability to perform at its stratospheric levels. The risk is hardly existential, but as an investor, it could feel existential.
Today we add Snowflake (SNOW) as a Top Pick with an add below price of $200.
We know full well it may never hit that price at that’s OK with us.
Thanks for reading, friends.
The author has no position in Snowflake (SNOW) at the time of this writing.