The stock prices of hypergrowth software companies have gotten ridiculously cheap, and I’m going to prove it.
I recognize it is hard to convince folks (including ourselves, sometimes) of this. Just because a stock’s price is half of what it once was, does not mean it is suddenly a bargain—perhaps it simply went from “ludicrous” to “unreasonable.” This is the pervasive cognitive dissonance we’re experiencing now, trying to reconcile selling out of companies we know to be exemplary.
Despite growing awareness of the positive differentiation that a SaaS business model provides, alarmist headlines around “ridiculous Price-to-Sales multiples” make for easy press, gives temporary validation to those who have underperformed us for years to say “I told you so,” and provides pundits and passers-by license to dismiss these high-quality businesses as nothing more than stratospherically-priced story and momentum stocks that have become disconnected from their fundamentals. Nothing could be further from the truth, of course, but without a way to demonstrate this point numerically, these negative assertions are sustained.
We see this now on the board with such off-handed comments like, “there’s literally no bottom” or “P/S of 20x-30x should actually be 5x-10x.” Such generalizations, usage of irrelevant ratios like price-to-sales, and apples/oranges comparisons to legacy valuation techniques rule the day in a discourse where seemingly the only person of prominence taking the other side of the argument is Cathie Wood. I’m going to challenge this FUD head-on, but first allow me to tell a story…
Jewels vs. Jokes – A Lesson in Price-to-Gross Profit
There are two stores in the town of Profitville: “Jewels” and “Jokes.” Jewels sells lovely gemstones for $100 each. Unfortunately for Jewels, a gemstone must be mined, cut, and polished by experts. For Jewels, getting a gemstone from under the ground to on the shelf comes at a high cost–$99 to be exact, good for a 1% gross margin For each gemstone sold, Jewels makes just $1 of gross profit ($100 Revenue - $99 Cost of Revenue).
Jokes, on the other hand, sells jokes to those looking for a quick, cheap laugh. For just $1, a customer can hear a witty anecdote to brighten their day. Seeing as jokes can be conjured from thin air, the cost to produce a single joke is $0 (100% gross margin) meaning for each joke sold, Jokes nets all $1 revenue as gross profit. This is something that Jewels and Jokes have in common—a gemstone and a joke both yield the same gross profit to the company’s owner. Another thing these two companies have in common is their market cap—both are valued at exactly $100. Finally, business this year was terrible, Jewels only sold ONE GEMSTONE, and Jokes only sold ONE JOKE.
Which company is “cheaper?”
Well, looking at the Price/Sales ratio, it must be Jewels. I mean, Jewels P/S is 1x, while Jokes is trading as an astronomical P/S of 100x.
Of course, I’m being facetious, because “P/S” for these companies, and any other real-life company, is largely irrelevant. What matters far more is gross profit; gross profit is the money that actually flows into the business once costs to produce the revenue are paid. Gross profit is then used to pay for overhead (SG&A and S&M) and R&D. This extreme example shows why Price-to-Gross Profit (P/GP)—while still very imperfect—is a far better measure than Price-to-Sales for making valuation comparisons. If I’m a homebuilder, I can artificially drive revenue by increasing the cost of each home by $50k and just advertising that every house comes with a bitcoin in the front yard. Never mind that it doesn’t do anything for my gross profit, but my revenue sure did jump!
As I move into the next section, keep in mind that gross profit, based on industry and company, is extremely wide-ranging. From Ford and GM at or below 10%, to Monday, ZoomInfo, or Upstart above 85%. An understanding of gross profit is exceedingly important to understand; much more so than sales.
A Comparison of Price-to-Gross Profit
These past couple of months have seen a painful rotation out of high-multiple hypergrowth stocks and into blue chip industrials, megacap tech behemoths, financials, and other cyclicals. As Jim Cramer puts it, into “companies that make real things” (because software, of course, is imaginary). Regardless of the real reason for this rotation (cough, algorithmic momentum trading, cough), the common justification has been that inflation and the auspices of higher interest rates have caused a “great reset” of multiples. Okay, fine. So, perhaps now that hypergrowth has fallen anywhere from 25-75%, are we nearing a bottom, have we overshot the bottom, or is there no bottom at all and we’re going to slowly watch DataDog drift to zero whilst my marriage falls to pieces?
First, let’s see what all the fuss is about, here is a list TTM P/S multiples of favored hypergrowth companies (bounded by “**” for easy reference) compared to a subset of “blue chip” companies that are pushing all-time-highs amidst this carnage.
Ticker Cap($B) 2021Rev($M) P/S (TTM)
**Snowflake** $89 $1,025 87.2
**SentinelOne** $12 $164 75.4
**Cloudflare** $34 $589 57.9
**DataDog** $45 $880 51.6
**Zscaler** $37 $761 48.3
**Monday** $10 $255 40.5
**Crowdstrike** $42 $1,286 32.7
**Amplitude** $5 $147 32.7
**Zoom Info** $22 $665 32.5
**Upstart** $10 $634 15.2
Middlesex Water $2 $144 14.4
McDonalds $200 $22,520 8.9
Apple $2,870 $365,810 7.8
Coca Cola $263 $37,800 6.9
Qualcomm $209 $33,570 6.2
Proctor & Gamble $397 $77,140 5.2
Deere $114 $44,000 2.6
Raytheon $134 $63,760 2.1
Aflac $40 $22,580 1.8
Ford $95 $134,610 0.7
If we stop here, there’s only one conclusion to draw from the data—hypergrowth is poised to fall much further and I need to put my 401(k) into Ford. But, of course, it would be illogical to stop here, because TTM P/S is an overly simplistic ratio, and is not informing the true profitability of the company.
For the next table, I’m applying the TTM Gross Margin (GM%) of each of these companies, to arrive at a TTM P/GP. P/GP gives me a much better sense of the money coming into the business less the direct costs to produce the revenue.
Ticker Cap($B) Rev($M,TTM) P/S(TTM) GM% P/GP (TTM)
**Snowflake** $89 $1,025 87.2 60% 144.6
**SentinelOne** $12 $164 75.4 59% 128.6
**Cloudflare** $34 $589 57.9 81% 71.8
**DataDog** $45 $880 51.6 76% 67.5
**Zscaler** $37 $761 48.3 78% 62.1
**Amplitude** $5 $147 32.7 69% 47.1
**Monday** $10 $255 40.5 87% 46.6
**Crowdstrike** $42 $1,286 32.7 74% 44.4
**Zoom Info** $22 $665 32.5 86% 37.6
Middlesex Water $2 $144 14.4 56% 25.7
Apple $2,870 $365,810 7.8 42% 18.8
**Upstart** $10 $634 15.2 86% 17.7
McDonalds $200 $22,520 8.9 54% 16.5
Raytheon $134 $63,760 2.1 18% 11.8
Coca Cola $263 $37,800 6.9 61% 11.4
Qualcomm $209 $33,570 6.2 58% 10.8
Proctor & Gamble $397 $77,140 5.2 50% 10.2
Deere $114 $44,000 2.6 28% 9.4
Ford $95 $134,610 0.7 10% 7.0
Aflac $40 $22,580 1.8 41% 4.3
You’ll notice that every P/GP increased from the P/S, this is logical since the denominator “GP” is necessarily lower than the denominator “S” as “GP” backs out the cost of revenue. You’ll also notice that hypergrowth is still richly valued—albeit to a diminishing degree. While the dichotomy exists between hypergrowth and blue chips, the contrast is not as profound. There’s even some blurring of the lines, as Upstart now has lower TTM P/GP than Apple and an old-school water utility like Middlesex.
While the hypergrowth/blue chip contrast is diminished, hypergrowth on a TTM P/GP basis STILL looks overvalued, though. Bad news? Well, not really, because TTM P/GP is still imperfect because we’re looking into the past. Nobody buys a company for what they did, they buy it for what the company will do. So, the next step is to apply some growth expectations. Here’s where things get tricky, because now we’re making predictions about the future. A thorny exercise, for sure, but thankfully, when revenue is recurring, on a contracted subscription, and net retention rates are high, revenue is highly predictable. So, bear with me while I apply some reasonable top-line growth expectations for the next three years, to arrive at a P/GP for estimated 2024 Gross Profit:
Ticker P/GP Gr(22) Gr(23) Gr (24) P/GP (24)
Middlesex Water 25.7 4% 4% 4% 22.9
**Cloudflare** 71.8 51% 49% 47% 21.7
**Snowflake** 144.6 105% 90% 80% 20.6
**Zscaler** 62.1 60% 55% 50% 16.7
**SentinelOne** 128.6 120% 100% 80% 16.2
Apple 18.8 7% 6% 5% 15.8
**DataDog** 67.5 72% 67% 62% 14.5
McDonalds 16.5 6% 5% 4% 14.2
**Crowdstrike** 44.4 55% 45% 40% 14.1
**Zoom Info** 37.6 60% 50% 45% 10.8
**Amplitude** 47.1 70% 65% 60% 10.5
Coca Cola 11.4 6% 5% 4% 9.9
Raytheon 11.8 9% 8% 7% 9.3
Proctor & Gamble10.2 4% 4% 4% 9.1
Qualcomm 10.8 8% 7% 6% 8.8
**Monday** 46.6 85% 75% 65% 8.7
Deere 9.4 8% 6% 4% 7.9
Ford 7.0 9% 5% 4% 5.9
**Upstart** 17.7 120% 30% 25% 5.0
Aflac 4.3 -4% -4% -4% 4.9
Okay, now we’re cookin’ with grease. Upstart is darn near the cheapest company out there, and the comingling between hypergrowth and blue chips is like a man named Brady meeting a lovely lady. Before you deride my growth estimates, take note of the universal deceleration among the bunch. I factored in not a single year of acceleration, despite the high likelihood that such a thing occurs.
While I do feel better about the valuation of the hypergrowth companies now, I’m still not super excited. But, I’m still missing something… Oh wait, I see it. I’ve assumed in the above that the earth dissolves into the sun in year 2025. Assuming this does not happen, let me go out a couple more years, to 2026 assuming an 80% growth durability in 2025 and 2026 (e.g 50% growth in 2024 becomes 40% in 2025 and 32% in 2025). And I’ll tell you what, I’ll leave the blue-chip company growth steady. Here’s where we land with that reasonable assumption:
Ticker P/GP(2024) P/GP (2026)
Middlesex Water 22.9 21.6
Apple 15.8 14.7
McDonalds 14.2 13.4
**Cloudflare** 21.7 12.1
Coca Cola 9.9 9.3
**Zscaler** 16.7 9.0
Proctor & Gamble9.1 8.6
**Crowdstrike** 14.1 8.5
Raytheon 9.3 8.5
**Snowflake** 20.6 8.3
Qualcomm 8.8 8.1
Deere 7.9 7.5
**DataDog** 14.5 6.9
**SentinelOne** 16.2 6.5
**Zoom Info** 10.8 6.2
Ford 5.9 5.6
Aflac 4.9 5.2
**Amplitude** 10.5 5.1
**Monday** 8.7 4.1
**Upstart** 5.0 3.6
Wow, am I cheating here? I mean, Monday.com would have to rise 3.5x from here to just get to the same P/GP multiple as McDonalds! DataDog would have to triple to get to Middlesex Water—what the hell are they putting in that water?!
And oh, I can do you one better. Gross Margins are not static. In fact, these are little companies still that are (hopefully) becoming big companies. What happens when companies scale? Margins improve. So now I’m going to include some very meager margin expansion.
Ticker GM %(TTM) P/GP (2026) GM%(Est)P/GP (2026) w/ Est. GM%
Middlesex Water 56% 21.6 56% 21.6
Apple 42% 14.7 42% 14.7
McDonalds 54% 13.4 54% 13.4
**Cloudflare** 81% 12.1 85% 11.5
Coca Cola 61% 9.3 61% 9.3
Proctor & Gamble50% 8.6 50% 8.6
**Zscaler** 78% 9.0 82% 8.6
Raytheon 18% 8.5 18% 8.5
**Crowdstrike** 74% 8.5 75% 8.4
Qualcomm 58% 8.1 58% 8.1
Deere 28% 7.5 28% 7.5
**Snowflake** 60% 8.3 75% 6.7
**DataDog** 76% 6.9 80% 6.6
**Zoom Info** 86% 6.2 88% 6.1
Ford 10% 5.6 10% 5.6
**SentinelOne** 59% 6.5 73% 5.3
Aflac 41% 5.2 41% 5.2
**Amplitude** 69% 5.1 73% 4.9
**Monday** 87% 4.1 88% 4.0
**Upstart** 86% 3.6 88% 3.5
I’ll stop here, but I could (and rightfully should) go on. Oh heck, why not! Remember that 80% growth durability I used for 2025 and 2026? Let’s make that 60% growth durability, and go out another three years (e.g. 50% growth in 2026 becomes 30% growth in 2027, 18% growth in 2028, and 11% growth in 2029). Again, I’ll leave blue chip growth steady:
Ticker P/GP (2026) P/GP (2029)
Middlesex Water 21.6 20.0
Apple 14.7 13.3
McDonalds 13.4 12.4
Coca Cola 9.3 8.6
**Cloudflare** 11.5 8.3
Proctor & Gamble8.6 7.9
Raytheon 8.5 7.4
Qualcomm 8.1 7.2
Deere 7.5 6.9
**Crowdstrike** 8.4 6.3
**Zscaler** 8.6 6.0
Aflac 5.2 5.6
Ford 5.6 5.1
**Zoom Info** 6.1 4.4
**DataDog** 6.6 4.3
**Snowflake** 6.7 3.9
**Amplitude** 4.9 3.2
**SentinelOne** 5.3 3.1
**Upstart** 3.5 2.9
**Monday** 4.0 2.6
Price-to-Gross Profit Black Magic – Still Imperfect, but Actually to the Detriment of Software
To head off naysayers, I realize I’m using a gross profit multiple, which is different than “Operating Profit” or “Net Profit.” So maybe, you might be thinking, I am deliberately using “Gross Profit” instead of “Operating Profit” or “Net Profit” to skew my results and support my narrative? Negative—I’ll leave that bush league stuff to those that push P/S comparisons. In fact, using “Gross Profit” instead of “Net Profit” is to the detriment of software companies. Why? To understand, we must understand the difference between gross, operating, and net. Gross profit is reduced by Sales, General and Administrative (SG&A) expenses and Research and Development (R&D) expenses to arrive at Operating Profit (aka “EBIT”). “Operating Profit” is then subsequently reduced by interest and taxes to arrive at “Net Profit.”
So, let me ask you this—if Ford, Deere or Coke fired their marketing and sales staff—would revenues be impacted? Answer: omg yes, and drastically so. Conversely, if ZScaler or Snowflake stopped marketing, would their revenue be impacted? Answer: Yes, but to a much much lower degree. SaaS revenues are contracted under subscription, firing marketing staff would hit upsells, cross-sells, and renewals, but it’s likely that business would hum along for some time unaffected.
What’s more, which sort of company do you think has higher overhead (General and Administrative) costs? Deere, Ford and Raytheon—with their mountainous back-office functions, warehouses, factories and supply chains, or super-lean software firms like ZoomInfo and ZScaler? I won’t answer that.
And finally, these companies are almost all debt-free (ZI an exception), interest expense is a non-event.
There’s a reason why mature tech/software/cloud companies like MSFT, ADBE, NVDA, FB, and others can maintain net margins that are higher than many of the blue chip’s gross margins–they’re completely different businesses that can’t be reasonably compared at a simple sales multiple.
In Summary
I hope my analysis and forecasting exercise gives some of you all comfort that, indeed, there is a bottom, and we’ve actually overshot it. That is not immediately apparent when looking at illogical multiples, but it evidentially obvious when looking more than an inch deep and projecting a few, predictable years into the future, and looking at gross profit instead of sales.
May these stock prices soon realize this gravy train has not stopped because the 10-year yield is suddenly higher.
Eric Przybylski, CPA