Valuation & Exponential Growth: Flexible Mindset

Investment Analysis Clubs Saul’s Investing Discussions
1

Understanding Exponential Growth
13 months ago, when the pandemic was kicking in, I listened to a podcast where Daniel Kahneman explained why we underestimated Covid-19. “We see things doubling every two days, three days, four days…and people, including myself, don’t seem to be able to think straight about exponential growth. I knew that there were a hundred cases in France, and I was about to fly to France. I also knew that epidemics are exponential. But I didn’t even consider the fact that if the rate of infection doubles every 3 days, then in a month it would increase by a factor of 1,000x. What we see today are infections that occurred 2-3 weeks ago. All of this is beyond the intuituitive comprehension – we’re in a situation that we’re simply not equipped to understand.”[1]

I recently remembered Kahneman’s quote, when a friend remarked that despite the recent drop in Snowflake’s share price, he wouldn’t invest in it, as it is still valued at 100 P/S (LTM). I know that valuation is a contentious topic, but I do believe that it is one of the attributes that we ought to evaluate as part of our comprehensive analysis of hypergrowers.

The Guidance Game
As we know, companies generally provide conservative guidance so they can “beat and raise” every quarter and keep old-school investors happy. The average revenue guidance beat across a cohort of this board’s companies is 6.5%. [2]

Old-school investors are not that naïve to not fall into this game, so they incorporate modest beats to their estimates of a company’s future metrics. Regardless, this cohort tends to beat consensus estimates by ~2%. This may not seem like much, but let’s remember that companies report four times per year. So, annualizing 6.5% becomes 29% of growth unaccounted for in guidance. Similarly, annualizing 2% becomes 8% in growth unaccounted for in consensus estimates.

The Guidance Game & Exponential Growth
So, why does all of this matter if this is how “the game” has worked forever, and how it will likely continue working for the foreseeable future? Well, it matters because the effect of this "game” compounds as a function of growth. And if this effect is “beyond intuitive comprehension” for a Nobel prize winner, surely it’s way beyond comprehension for us mortals here. So let’s use an example to put it in perspective.


Company		Consensus Rev Estimates (CY’21)[3][4]		Consensus Rev Growth Estimate 		Estimates Accounting for Beats (CY’21)		Rev Growth Accounting for Beats	
Okta			$1,321M							30%					$1,177M						41%
Docusign		$1,932M							36%					$2,093M						46%
ZScaler			$733M							37%					$793M						48%
Zoominfo		$818M							41%					$728M						58%
Cloudfare		$616M							43%					$667M						55%
Zoom			$3,800							43%					$4,113M						55%
Twilio			$$2,540M						44%					$2,749M						56%
Datadog			$889M							47%					$962M						59%
Crowdstrike		$1,319							51%					$1,428M						63%
Snowflake		$1,093							85%					$1,182M						100%

Now, assuming that the share prices from this cohort remain the same for the next year, this is how they would be valued on 10 May 2022.[5]


Company		May’22 EV/S (LTM)		May’22 EV/S (NTM)[6]
Okta			24.0x				21.1x
Docusign		16.2x				14.1x
ZScaler			25.4x				21.8x
Zoominfo		20.6x				16.2x
Cloudfare		28.6x				23.9x
Zoom			19.1x				14.4x
Twilio			15.8x				12.9x
Datadog			22.1x				18.1x
Crowdstrike		25.0x				20.0x
Snowflake		39.5x				32.2x

Tying it all Together
Here is where the realization comes into play. Snowflake might certainly look expensive today using conventional valuation techniques. But (1) visualizing the effect of compound growth, (2) reading into “the game” that’s played with old-school investors, and (3) knowing that these techniques are highly fallible for these companies yields a different perspective.

Do we really think that Snowflake will be valued at ~32x EV/S (NTM) next year when it’s expected to grow at ~85% the following year? For context, despite the recent meltdown in the share prices of these companies, Crowdstrike’s average EV/S (NTM) for 2021 has been 35x. And this is when consensus estimates for Crowdstrike are for 51% growth. If Snowflake expects >80% growth, we can only assume that “the market” would attribute a higher valuation to Snowflake – which would translate to share price appreciation over the next year. The same can be said about many (if not all) of the companies in this cohort.

Disclaimer
This is NOT to say that Snowflake is cheap. Or any of these companies for that matter. This is also not to say that I can predict the future results of any of these companies better than others. I don’t have a crystal ball to read into the future, and I’m not going to pretend that I can be even mediocre at estimating it. Rather, the purpose of this post is to try to equip ourselves with an understanding of the effect that compound growth can have. It is important to understand these companies holistically – the technology, the financials, the management, the execution…and valuation is just one of those factors. The more we sharpen these, the better we will become.

-RMTZP
Protect the identity of this board and maximize your learning by visiting https://discussion.fool.com/for-board-newcomers-all-others-34765… before participating

References:
[1] https://podcasts.apple.com/mk/podcast/why-we-underestimated-…
[2] https://discussion.fool.com/q1392021-earnings-preview-34819176.a…
[3] CY refers to “calendar year” to normalize for the differences in company’s fiscal year-ends
[4] We could spend a lot of time debating what a “realistic” expectation is, but for simplicity I’ve assumed a 3% beat on each quarter to consensus estimates
[5] I know that this is unrealistic, but this helps explain the example
[6] Assuming growth continues at 85% the pace of previous year’s growth rate

68 Likes
2

There is a mistake on the “Consensus Rev Estimates (CY’21)” column.

Okta should be $1,087M and Zoominfo should be $673M as opposed to $1,321M and $818M, respectively. Apologies!

3

And this is when consensus estimates for Crowdstrike are for 51% growth. If Snowflake expects >80% growth, we can only assume that “the market” would attribute a higher valuation to Snowflake

But if the market already knows that Snowflake has told it to expect that level of growth, why do you think it’s not already priced in appropriately at NFY p/s? Because Wall St can’t take LTM and multiply by 1.85x? One of year of 85% growth is MUCH easier to value than doubling every 2-3 days for a month.

It seems like you don’t think the future should be discounted at all in today’s stock prices, but the market never works that way.

Assuming all these firms beat and raise each Q all year is fine until they start missing/not beating. That’s when high-valuation stocks get massacred. Even though SNOW and DDOG are amazing firms.
TTD has been cut in half since late Dec, as just one example.

It’s not realistic to say they will all beat/raise & then only slow by 15% when DDOG just told us they expect to slow by 30% this year.

Agree in advance that those firms that do continue to beat and raise will likely do well over the longer term. But we’re not going to see the 2020-Covid bump in growth rates again.

15 Likes
4

But we’re not going to see the 2020-Covid bump in growth rates again.

Covid was a detriment to growth for DDOG.

AJ

4 Likes
5

Not to put words in the OP’s mouth but I think their point was that Wall Street (or people in general) struggle to grasp the idea of a company growing at 85% for years and years. Not that people can’t multiply by 1.85. I only say this because I’ve seen this argument lately. That Wall Street or investors in general perpetually undervalue high growth stocks. When all you have to do is look at history to see that’s not true. The study of the nifty fifty in The 1970s for example shows investors bid them up too high. As for whether SNOW will be trading at 32x ntm revenue this time next year or whether they would be considered shockingly low, I don’t know, I guess we’ll find out. A few years ago that number would have been shockingly high.

2 Likes
6

I think Covid was a detriment to growth for most software companies unless they were connected to WFH. In hard times ,especially unique hard times ,companies want to delay projects that do not lead to a larger bottom line in the immediate future.
The case for digital transformation is stronger than ever. The demand for digital scientists is hot . Making me wonder if Alteryx is permanently maimed or due for a comeback. Because it is a good example of something that “can wait a little bit until we are more sure about business conditions”.

4 Likes
7

NajdorfSicilian,

I’ll keep my responses short and relevant as I’m afraid we are nearing an off-topic discussion. Happy to continue discussing via email to refrain from polluting this board.

But if the market already knows that Snowflake has told it to expect that level of growth, why do you think it’s not already priced in appropriately at NFY p/s? Because Wall St can’t take LTM and multiply by 1.85x?

I think that it is not priced in because the market has failed to do this for as long as cloud companies have been in existence. Take a look at BVP’s Emerging Cloud Index (https://cloudindex.bvp.com/) – if markets were efficient this would not have been possible. It is true that many of these companies have seen significant multiple expansion over the past year – but even ignoring the last year we observe vast overperformance from these companies as a result of them constantly surprising the market time and time again.

Of course I’m not saying that Wall St can’t multiply 1.85x! I am by no means pretending to be smarter than anyone. I included in my disclaimer that “This is also not to say that I can predict the future results of any of these companies better than others.” I do think that by analyzing these companies exhaustively as we do here, we sometimes have a slight edge. The whole purpose of this is to try to think outside the box from conventional techniques.

Assuming all these firms beat and raise each Q all year is fine until they start missing/not beating. That’s when high-valuation stocks get massacred.

Sure, missing/not beating leads to share price massacre. But as I pointed out, this cohort beats expectations by 6.5% on average. I just verified, and confirmed that there hasn’t been a single quarter where any of these companies missed their guidance. Can it happen? Of course it can! But that’s not the point. The point is to show a potential scenario if they keep beating and raising – which is historically what tends to happen.

It’s not realistic to say they will all beat/raise & then only slow by 15% when DDOG just told us they expect to slow by 30% this year.

I cannot find that part of the call, but would appreciate if you could point it out. Datadog just reported 11.8% sequential growth, which annualizes to 56%. Their guidance for next quarter would represent 51% YoY growth. I don’t see how they can slow to 30% this year.

Agree in advance that those firms that do continue to beat and raise will likely do well over the longer term. But we’re not going to see the 2020-Covid bump in growth rates again.

As Phoolio18 indicated, the pandemic was a detriment to Datadog’s business. Perhaps you meant that we won’t see Covid bump in multiples, which is true. As Saul pointed out in his 2020 review, we probably not see share a performance like last year’s ever again. Again, that’s not really the point.

12x,

As for whether SNOW will be trading at 32x ntm revenue this time next year or whether they would be considered shockingly low, I don’t know, I guess we’ll find out. A few years ago that number would have been shockingly high.

A few years ago that might have been shockingly high, but a few years we simply didn’t have companies (like Snowflake) with a ~$760M run-rate growing at >100% with an 88% increase in >$1M customers. At its peak, Salesforce was growing in the mid 60%s at a ~$450M run-rate. Plus, not to mention how imperative these technologies have become across industries compared to 15 years ago.

-RMTZP
Protect the identity of this board and maximize your learning by visiting https://discussion.fool.com/for-board-newcomers-all-others-34765… before participating

16 Likes
8

As for whether SNOW will be trading at 32x ntm revenue this time next year or whether that would be considered shockingly low, I don’t know, I guess we’ll find out. A few years ago that number would have been shockingly high.

But a few years ago we were searching for companies growing revenue 20% per year, and 30% per year was a real find. There was no such thing as a major company growing revenue at 124% per year (which is what Snow grew last 12 months). We wouldn’t even have been able to imagine that a real company could grow that fast, so saying that back then a stock price of 32x revenue would be considered shockingly high is kind of irrelevant. It would have been shockingly high for a company growing revenue at 20% or 30%, but what does that have to do with Snowflake?

Saul

41 Likes
9

I think that it is not priced in because the market has failed to do this for as long as cloud companies have been in existence. Take a look at BVP’s Emerging Cloud Index (https://cloudindex.bvp.com/) – if markets were efficient this would not have been possible. It is true that many of these companies have seen significant multiple expansion over the past year – but even ignoring the last year we observe vast overperformance from these companies as a result of them constantly surprising the market time and time again.

Of course I’m not saying that Wall St can’t multiply 1.85x! I am by no means pretending to be smarter than anyone. I included in my disclaimer that “This is also not to say that I can predict the future results of any of these companies better than others.” I do think that by analyzing these companies exhaustively as we do here, we sometimes have a slight edge. The whole purpose of this is to try to think outside the box from conventional techniques.

I think the issue some people have is that they look at the average or look at a class of companies in a monolithic way. While valuation multiples for IT software companies are generally high, some companies will grow like weeds for many years. Others will stall. Others will flounder. Our task is to try to analyze companies and foretell which companies will continue to grow like weeds and buy those. Our other job is to reject the others. Our third job is to monitor our investments and scan for other, better investments. This monitoring process often leads us to discard companies that we previously thought would grow like weeds (but we changed our minds based on new information). Other times we reduce allocations to raise funds for other investments, again based on new incoming information that leads us to change our view about which companies will continue to grow like weeds.

Using the above process, we are usually/often able to exit companies and redeploy the funds to companies that will end up doing better. There’s lots of information and the decisions are often nuanced (and certainly the hit rate will be below 100%) so following the process is not so easy for the average investor and certainly not easy for people who don’t put in the work. Growth stock investing is not for the faint of heart because there will be occasional big drops. And it’s not for people to hang on to something and can’t change their mind based on new information.

GR

54 Likes
10

A few years ago that might have been shockingly high, but a few years we simply didn’t have companies (like Snowflake) with a ~$760M run-rate growing at >100% with an 88% increase in >$1M customers. At its peak, Salesforce was growing in the mid 60%s at a ~$450M run-rate. Plus, not to mention how imperative these technologies have become across industries compared to 15 years ago.

We don’t have to go back very far to get a similar growth story as Snowflake and seeing how the market has really warmed up to growth stocks.

Crowdstrike went public at 46x sales in Jan 2019, with revenue more than doubling. Had one paid 140x TTM sales as Snowflake IPO day investors had, CRWD would have gone public at $189 a share. Essentially what the stock is today.

There is no denying that price multiples for growth stocks have expanded over the past few years. Which you acknowledge in your post. You then go on to say that even further back than a year(!) we see the same thing, rising prices. A few years and what has happened over the past few years in no way, shape or form could be construed as to how the stock market “normally works.” It is just how the last few years have gone down in the 200+ year history of the stock market. The next few years could be remarkably different.

Sure, missing/not beating leads to share price massacre. But as I pointed out, this cohort beats expectations by 6.5% on average. I just verified, and confirmed that there hasn’t been a single quarter where any of these companies missed their guidance. Can it happen? Of course it can! But that’s not the point. The point is to show a potential scenario if they keep beating and raising – which is historically what tends to happen.

But there already have been. If not missing guidance, surely lowering guidance which had the same impact as an outright miss. See NTNX for example, or AYX, where, when I pointed out that growth was going to fall to 10-15% growth rate, I was told I wasn’t paying attention. When you have nothing but higher and higher prices it really becomes more difficult to outperform when you get these eventual missteps.

There have been people who put their studies on what happens on every fall over the past few years, going back to 2018. Of course, every single time, stocks just shot right back up within months. So people get solace in these reviews over what happened over the past few years. I find none in a study over such a short time span.

4 Likes
11

Now, assuming that the share prices from this cohort remain the same for the next year, this is how they would be valued on 10 May 2022.[5]

Company May’22 EV/S (LTM) May’22 EV/S (NTM)[6]
Okta 24.0x 21.1x
Docusign 16.2x 14.1x
ZScaler 25.4x 21.8x
Zoominfo 20.6x 16.2x
Cloudfare 28.6x 23.9x
Zoom 19.1x 14.4x
Twilio 15.8x 12.9x
Datadog 22.1x 18.1x
Crowdstrike 25.0x 20.0x
Snowflake 39.5x 32.2x

Here’s what actually happened a year later, using Jamin Ball’s report ended 4/29.
Okta 11.3X
Docusign 6.9x
ZScaler 24.2x
Zoominfo 20.4x
Cloudflare 32.8x
Zoom 5.6x
Twilio 4.6x
Datadog 26x
Crowdstrike 21.6x
Snowflake 27.2x

Tying it all Together
Here is where the realization comes into play. Snowflake might certainly look expensive today using conventional valuation techniques. But (1) visualizing the effect of compound growth, (2) reading into “the game” that’s played with old-school investors, and (3) knowing that these techniques are highly fallible for these companies yields a different perspective.

Do we really think that Snowflake will be valued at ~32x EV/S (NTM) next year when it’s expected to grow at ~85% the following year? For context, despite the recent meltdown in the share prices of these companies, Crowdstrike’s average EV/S (NTM) for 2021 has been 35x. And this is when consensus estimates for Crowdstrike are for 51% growth. If Snowflake expects >80% growth, we can only assume that “the market” would attribute a higher valuation to Snowflake – which would translate to share price appreciation over the next year. The same can be said about many (if not all) of the companies in this cohort.

To your question, do we really think that Snowflake will be valued at NTM 32x EV/S right now, I think the issue for anyone who would find it surprising or “shockingly low” as I said a year ago, is ut is possible to have a cognitive bias of “valuation anchoring.” Most people are familiar with “price anchoring” which I am sure you are familiar with since you reference Daniel Kahneman, but it’s also possible to “valuation anchor.” That is, since CRWD EV/S for all of 2021 was 35x, that we can assume that is something we can anchor everything to. For all of the 200 years in the history of the stock market, we can calculate a valuation strategy on what happened with CRWD in 2021. It turns out stock valuations do not stay on a permanently high plateau, valuations can compress from historically high levels, and things can change.

Outlook can change, things change, and right now the market is really penalizing companies if they depend on stock dilution to prop up unsustainable growth.

7 Likes
12

Most people are familiar with “price anchoring” which I am sure you are familiar with since you reference Daniel Kahneman, but it’s also possible to “valuation anchor”…Outlook can change, things change, and right now the market is really penalizing companies if they depend on stock dilution to prop up unsustainable growth.

12x,

Clearly, this has been a very tough lesson for me to learn. While the key point of my post was to try to think about the effect of multi-year exponential growth for these companies, I was clearly anchoring on valuations and underappreciating how quickly things can change. This post was already flirting with off-topic so I’ll ask others to refrain from piling on to respond. But I did want to acknowledge your response, as it shows how fickle valuations can be.

We can control little beyond analyzing the performance of our companies, and while many other forces may affect their share prices, I am deeply appreciative that we have this forum to stay focused on what we can control. Back to analyzing individual companies now to stay on course to doing what we do best here :slight_smile:

-RMTZP
Visit https://discussion.fool.com/rules-of-the-board-revised-edition-3… to maximize your learning of the board before posting

5 Likes