Palantir Vs. Snowflake: Which Is The Better Buy For Long-Term Investors?
Summary
Big data analytics and decisioning is one of the cornerstones of the digital transformation revolution sweeping through much of the IT firmament.
The two leading companies in providing the building blocks for large data warehouses and for decisioning applications based on big data analytics are Palantir and Snowflake.
Both of these companies have already compiled enviable growth records; indeed, the growth Snowflake is enjoying is as rapid as any ramp I have ever seen.
The issue for investors is really not the likely success of these companies, or the fact that they are down a great deal from recent post IPO valuations. The issue is their current valuation and future growth.
My conclusion, based solely on valuation, is that Palantir will provide a somewhat greater return than Snowflake over the coming years. But in neither case is it likely that long term returns can rise beyond the teens.
A sharp sector rotation affords long-term investors with opportunities not often encountered!
It will probably come as no surprise to most readers that the last 3 months have been marked by a very strong sector rotation away from high-growth, high value names to cyclical/reopening names. Even some of the strongest companies have seen their valuations eviscerated over that period. The culprit for the rotation is investor concern with both prior valuations and investor concern about spikes in inflation leading to interest rate increases. The valuation compression is even sharper than it may appear on the surface. Not only are share prices down, but revenue estimates and operational performance are showing very strong positive trends. In most cases, EV/S ratios have fallen by 25% or more and in a few cases the fall has been 50% and even greater. Much like a lake which is plagued by drought, this valuation compression has made visible several islands of opportunity.
This is not an article about sector rotation per se, or about the valuation of tech names per se. Have the last several days of trading been a harbinger that sector rotation is ending, or was it simply a long awaited response to several days of extreme valuation compression in the tech space? At some point, the combination of falling share prices and rising growth rates for revenues and free cash flow will obviously lead to a reversal-but lacking 2nd sight, I have to leave that vexed question to others.
One of my favorite components of Lincoln’s 2nd inaugural address is the ringing phrase, “fondly do we hope, fervently do we pray.” Now the President was talking about the passing of the scourge of war, and I am writing about the end of a sector rotation. But it has been a difficult period that has scourged the portfolios of many investors and I do hope it will pass away soon.
One of the several artifacts of what some might call “The Great Sector Rotation” has been the emergence of several growth names that had been un-investible due to valuation for months or since they became public. This article is focused on two such names, Palantir (NYSE:PLTR) and Snowflake (NYSE:SNOW). These are names that are prominent in the IT space. are amongst the leaders in the Big Data/Analytics space and which have generated remarkable interest amongst many investors and readers of SA. Both of them are quite exceptional companies-but I don’t think either is a particularly remarkable investment. That said, at current valuations (I am using share prices as of 5/24 to calculate my valuation relationships), both of these names can potentially provide investors modest-but double digit long term returns. That is usually less than many tech investors find acceptable-but the visibility and relative stability of both of these companies is an unquantifiable intangible that some may find particularly attractive. Again, entirely based on valuation, and no other factor, Palantir is a more attractive investment-although the calculated return difference compared to Snowflake isn’t huge. Obviously, it is the CAGR differential between businesses that make it so difficult for an observer to make an unqualified choice.
I want to make clear than when asked for a choice between these two names, my first response would be neither. There are simply better choices in the IT world as potential investments. Investors have a wide choice of names in the space in the wake of the valuation compression and there are different attributes associated with these names.
For those investors looking for my choices in the high growth/high valuation segment, I would recommend CrowdStrike (CRWD), Datadog (DDOG). Zscaler (ZI) and ZoomInfo (ZI). Could Snowflake make this list-possibly because it certainly has the highest CAGR of any of the names that I follow. But even using what is essentially a 4 year CAGR of 70% to reach a terminal sales level of more than $6 billion, and a terminal growth rate in the mid-thirty percent range still does not produce super returns for Snowflake because it is hard to do so when the starting point is a valuation of 53X EV/S (as of the close on Monday, May 24th). And forecasting that a company with a $6 billion run rate to continue growth in the mid-30% range is certainly a bit of a reach-although the market that SNOW addresses is enormous and growing.
For investors looking for a combination of growth coupled with less than average valuation in terms of their EV/S ratio, choices include names such as Upstart (UPST). Affirm (AFRM), Asana (ASAN), Elastic (ESTC), Wix (WIX) and Jamf (JAMF). Some investors are looking for names with very high free cash flow margins. Here choices include Atlassian (TEAM), Dynatrace (DT), Veeva (VEEV), Microsoft (MSFT) and Adobe (ADBE). Many investors would choose to use some kind of combination of free cash flow margin + growth rate in compiling a buy list. Here, the regnant champion is ZoomInfo (ZI) but perhaps Palantir makes the list-depending on whether the free cash flow metric achieved last quarter was an outlier or represented the start of a trend. Trade Desk (TTD), in the wake of its recent share price compression of 40% since mid-February would also be on this list.
Some investors look for bargains amongst fairly IPO’s. There are some bargains now to be seen in that sector such as Affirm and also nCino (NCNO) and perhaps Jamf. These are companies whose price action after the IPO took their valuations to unsustainable levels and which have now compressed to what appear to be bargains. I try to approach the recent IPOs with the same methodology as the rest of the names I follow. Inevitably, given the small number of shares in most IPOs, stock prices can subsequently reach dizzying levels as institutions, in particular, look to establish full positions that they were unable to create with an IPO allocation. But once that demand ceases, these names fall, and can do so to very compressed valuations.
Finally, there are some investors who look for proverbial bargains by screening for percentage declines over some fixed period. That is not a methodology I favor. Many companies can be mis-priced in terms of valuation and so the fact that a name is down a particular percentage is of less importance to me, than how the companies score on EV/S valuation and free cash flow margins. I have seen many articles that focus on either Snowflake or Palantir because of their steep declines during and even before the sector rotation. While it is not uncommon for commentators to use a percentage decline as a screening point, my approach doesn’t value that technique. Specifically, there was no conceivable logical methodology that could have supported the peak valuations of either Snowflake or Palantir, so the fact that they have fallen substantially is not proof-at least to this writer-that they are now reasonably valued.
I try not to be obsessively tied to formulas in creating a buy list for my own portfolio or for my recommendations. I am, admittedly, biased toward growth, but when I see a combination of growth and free cash flow margin, I can fall in love-at least figuratively. (Sort of like Daumier’s famous lithograph, “Fusion des Compagnies. Effusion des actionnaires. (Les Beaux Jours de la Vie), from Le Charivari, Honoré Daumier ^ Minneapolis Institute of Art. I use some combination of all approaches in compiling my own list of names that I think are buy rated. I don’t include any megacap names, more because I think their investment merits are well recognized and there isn’t too much in the way of value add that I might provide.
To repeat: both Palantir and Snowflake are and will remain remarkable businesses that are revolutionizing the way software is used in both government and commercial applications. Bringing storage to the cloud the way it is done by Snowflake is hastening digital transformation and making it easier to migrate workloads to the cloud-a significant priority for many-probably most enterprise users of IT. The combination of data integration, AI and search which enable users to find patterns and develop useful insights is achieving some of the more “Buck Rogerish” dreams of software engineers and ultimately all classes of users.
But having said that, it is still necessary to look at growth and cash flow to arrive at a valuation and while the valuation compression has led to opportunities, the opportunities are not of the once in a generation scale. Investors who want to own the best companies frequently are going to be asked to sacrifice some percentage upside to buy the best of the breed, and that is what I see here, regardless of my admiration of the offerings of both companies.
One thing to note: I would be greatly surprised if either Palantir or Snowflake will precisely look the way they do at this writing. They are going to wind up making acquisitions-I doubt that valuations and chemistry would allow either company to be acquired. Speculating about acquisitions is a fun parlor trick, but not something that can really be forecast with any degree of specificity. But one reason as to why I have suggested that these companies will continue to grow at elevated rates relates to my belief that there will be some element of inorganic revenue in the results of both companies 4 years from now.
The background of Snowflake and Palantir
Both of these businesses have been and remain high growth companies with strong technology moats and a host of the most prominent IT users in the world. Palantir reported recent results that swung strongly to free cash generation; SNOW is perhaps the fastest growing name I have seen at scale. And the shares of both companies have seen rather substantial compression. Since the rotation began in early/mid-February, shares of Snowflake have fallen as much as 39% before bouncing 11.5% on Friday, May 14th partially due to a new recommendation from the analyst at Goldman Sachs. Shares of Palantir had fallen as much as 53% before its bounce 9.3% on Friday May 14th.
Both of these companies seem destined to be major factors in the software over the course of the coming years. Over time, as they mature, I expect that both of these company’s will evolve highly profitable business models. Based strictly on the way I value companies, I find the shares of Palantir to be more attractive to investors than the shares of Snowflake-even though self-evidently, Snowflake is growing faster than Palantir. I believe that long-term investors will achieve a somewhat greater return investing in Palantir than in Snowflake-but the difference isn’t huge and speculating about an end-result 4 years from now is inevitably a fraught undertaking.
I have been frequently asked by subscribers to my Ticker Target service. by investment advisory clients and by readers of SA articles to provide some opinion on both of these companies with a plurality inquiring about Palantir. For months, until now, I haven’t chosen to make much of a response given the valuations have been of a magnitude that made any kind of positive recommendation more faith based than logical.
Trying to find a formula for relative valuations
Let’s face it-trying to decide between the investment merits of two companies with a great base of IP, addressing hot spaces within the enterprise software space is a bit like handicapping the results of sporting events before the start of a season. There is loads of pure guesswork and less substance than most would like. Many analysts demur doing something like this-or if they are like me, they suggest to clients that there is no reason just to own a single name of this kind in a portfolio. I am going to attempt to present some qualitative as well as quantitative analysis-but in the nature of things it will be subjective.
One consideration is always management. Does either Palantir or Snowflake have better management? Frank Slootman of Snowflake probably has better bona fides than the Alex Karp, the CEO of Palantir - he has, after all , sold one company he founded, Data Domain to EMC for an incredible valuation and he guided Service Now to huge success, and in the process more or less ran over BMC Software which was one of the stalwarts of the enterprise software space for many years.
The CEO of Palantir is Alex Karp and he co founded the company 17 years ago. He has a host of beliefs that many might consider outside of the mainstream, especially for a company that makes a living selling to the US government and particularly the US military. He has been described as eccentric and I imagine that any self-professed socialist who has built a net worth of almost $2 billion is likely to have an unusual set of values. The fact that he keeps Tai Chi swords in his office-which at times can be in a barn in New Hampshire-while not entirely abnormal amongst tech entrepreneurs, is more than a bit different than the background of Mr. Slootman.
But company’s such as Palantir are run by teams and there are some extraordinary players on Palantir’s team. Peter Thiel, well known as a co-founder of PayPal (PYPL) and one of the first outside investors in Facebook, also co-founded Palantir and remains on the board. So too, is famed tech investor, Joe Lonsdale who recently said that:
“As a director of a public co there are regulations about what you can say – you’re discouraged from speaking up & nobody does in our risk-averse society. But #’s came out yesterday and were misunderstood… we are going to crush the shorts / I am extremely bullish."
I am not quite sure why Mr. Lonsdale felt that the quarterly results were misinterpreted. Palantir was caught up in a tech rout on the day after its earnings were released. It was just reflecting the sector rotation that has marked much trading in these names .
Finally, there is Stephen Cohen, another co-founder of Palantir and currently an Executive VP of the company. Famously, at the ripe age of 23, he has been credited with writing the initial prototype of the Palantir platform in all of 8 weeks.
As must people who know me would agree, I am a somewhat desiccated older curmudgeon, whose acceptance of eccentricities is perhaps less than it should be. But overall, I believe that in evaluating these two companies, there is not all that much to choose. It is a bit easier to conclude that Snowflake has a management structure that will lead to better investor returns simply because of the track record of the CEO. But Peter Thiel and Joe Lonsdale have made themselves billions and those along for their various rides have done quite well. I can wish that there was less mystery surrounding some parts of the Palantir business, but at the end of the day, I don’t think an investment decision between these two names can be made based on differences in management capabilities.
Like most other analysts, I try to look at relative valuations in making a recommendation. The fact is that Snowflake still has the highest 12 month forward EV/S ratio of any name I follow, and that valuation is based on a revenue estimate of $1.2 billion. That is a forecast for growth of just over 100% for the next 12 months-and about 10% above the current published consensus for the same period. Despite the forecast for triple digit growth, I wouldn’t find it terribly surprising for Snowflake to continue to exceed estimates-the momentum in its space is just that strong.
Essentially the problem I have with recommending Snowflake shares is just how much the company will have to grow in order to justify the current valuation-even after the huge haircut of recent months. I have used a 3 year forward CAGR estimate of 70%-I think that is reasonable, if growth this year is over 100% as seems likely.
Snowflake in its latest reported quarter started to generate free cash flow. To do so, however, it needed to have an enormous growth in its deferred revenue balance-a result that is partially seasonal as its clients renew their agreements. I expect SNOW to generate a modest level of free cash flow in its current year, but that is not going to be a sufficient reason to recommend the shares. In order to generate 70% revenue growth over several years, I anticipate that the company’s ability to generate a substantial free cash flow margin will be challenging. The company will simply have to continue to make outsize investments in sales and marketing and research and development. I imagine that some of that will continue to be part of the company culture even looking out several years; I would not anticipate anything more than average cash flow margins by that time-and those margins could easily be less than average.
In any event, using a multi-year CAGR of 70%-and then starting the compounding from the base of $1.2 billion of revenues that I anticipate for the current year produces a terminal revenue estimate of about $5.9 billion. I assume that this company will still be enjoying hyper growth at the end of the period-just not at the current elevated levels. I think using a terminal growth rate of 35% is reasonable and even after valuation compression, the average EV/S for that growth rate is 16X. So, that might suggest that the enterprise value for Snowflake 4 years from now ought to be about $94 billion-the enterprise value as of the close on Friday was $56 billion, or thereabouts. That works out to a 14% annualized return. That is certainly far better than such a calculation might recently have been, and far greater than any assumed inflation rate I have seen…but I wonder if it is enough for most investors who are usually looking for something more to be properly compensated for risk in investing in a name such as this.
A comparable calculation for Palantir starts with estimated revenues of $1.8 billion for the next 12 months. This estimate was revised based on the results that the company reported on 5/11/21. That is considerably greater than the currently published consensus-for 2021 of $1.47 billion-but my estimate goes out an additional quarter and is not burdened by the adherence to the company’s rather mechanical guidance of “greater than 30% for the foreseeable future” which has been the mantra of the company CEO, and which is used by many analysts as a substitute for preparing their own set of expectations.
Last year, the company reported a 47% growth in revenues and had forecast a 45% growth in revenues for Q1. Q1 revenue growth came to 49%; the company is forecasting 43% revenue growth year on year in the current quarter but given the rather muted sequential growth implied in that forecast (5.6%), I believe it will be exceeded by some noticeable amount. The company reported a free cash flow margin of 34%, a very dramatic change from the negative free cash flow margin reported in 2020.
While the company saw a decline in its deferred revenue balance in Q1, the more inclusive metric of remaining performance obligation rose by 4.7% sequentially, which is a strong performance given the typical seasonal decline usually seen in that metric in Q1. Overall, calculated billings were up 248% year on year and the year on year increase in the RPO balance came to 129%. These are, in my opinion, strong indicators for future growth.
Since the time that Palantir became a public company, it has been criticized for the slow growth of its commercial business compared to its government business. But in the last quarter, the company’s US commercial business finally showed some decent growth of 72%. I will cover this subject more fully later on in this article.
In any event, I have chosen to use a 3 year forward CAGR of 42% in evaluating Palantir, based more on its historical growth than some special knowledge about how fast it might grow. Because of the multiplicity of products and solutions that are enabled by Palantir’s platforms, it can be a bit more difficult to estimate a longer term growth rate than would be the case when dealing with a company whose revenues are coming from a more targeted focus. In any event, using a 42% CAGR, and my current estimate for 12 month forward revenues yields a run -rate estimate 4years out of greater than $5.2 billion. My guess, and I make no representation that it is more than that, is that the company will be still growing in the low 30% range at that point, with a free cash flow margin of greater than 20%.Just to be clear, the cash flow results seen last quarter, while perhaps not enough to suggest a trend, are certainly suggestive of a business model that is potentially very profitable. Taking the estimated cash flow generation into account, the CAGR that I am estimating for Palantir is currently worth an EV/S of about 16X-17X looking at the average EV/S metric for a low 40% growth estimate. In turn, this leads to an enterprise value forecast of about $86 billion compared to last Friday’s enterprise value of $$44.5 billion. This suggests a 4 year return of about 18% compounded, somewhat better than the rate of return I calculate for Snowflake. Snowflake’s elevated valuation simply makes it very difficult to realistically project exceptional long-term returns-even though in many ways Snowflake is an exceptional company operating in an exceptional market.
Where the analysis could be off and what are the risks?
This article is basically about which of the two names I would rather hold or invest in for the long-term. It isn’t a terribly obvious choice-although the numbers, as I see them, suggest that Palantir will relatively outperform Snowflake-mainly because even after a substantial valuation compression, Snowflake shares are still the most expensive name in the IT space in terms of EV/S by a fairly substantial margin. Just to make that point abundantly clear, Snowflake shares, as of the close on Friday, May 21th had an EV/S ratio based on forward revenues of 53.5X; the next two highest ratios in my coverage universe were those of Bill.com (BILL) at 36.5X and Cloudflare (NET) at 35.9X. Meanwhile, Palantir shares currently sell for an EV/S of 25.5X.
There are certainly flaws in the investment merits of both companies. I have presented a quantitative model that attempts to deal with the difference in growth rates for the two companies at the present time. But I would be the last analyst on the planet to suggest that I have some crystal ball. I really have no specific way of addressing the potential growth of Snowflake over the next 4 years. I feel reasonably comfortable in suggesting that my use of $1.2 billion for SNOW revenue over the coming 4 quarters is supported both by qualitative comments made by company management and by using sequential quarterly growth estimates that are consistent with recent history. Further, the company’s RPO balance grew to $1.3 billion, up 213% for the year and its DBE ratio was 168%. The RPO balance actually rose by 44% sequentially the latest reported quarter, after rising by 35% sequentially the prior quarter, and the sequential growth in revenues was 19% for the quarter compared to 20% the prior quarter. Given all of those statistics, I felt that forecasting $1.2 billion for the next 4 quarters, compared to the company’s forecast of about $1075 million for the current (2021) year made sense.
But when it comes to supporting a CAGR of 70% for the 3 years after this one, I would acknowledge that it is somewhat of a guess-and a CAGR of that rate would be breaking new ground in terms of growth at scale. I will be reviewing some of the reasons for the company’s exceptional growth opportunities below-but those specifics are simply not going to allow me, or anyone else, to determine if the most reasonable CAGR is 50% or 70% or some other number. I have yet to see a 3 year CAGR of 70% for a company of this scale. But given that I anticipate that the first year in the forecast period will be nearly 100%, then 70% growth is quite likely and allows for slowing growth as the company’s scale approaches and exceeds a $6 billion revenue run rate.
Not terribly surprisingly, many analysts rate SNOW shares as a hold although some percentage do rate it as a buy. The issue is almost entirely one of valuation-with a current EV/S of 54X based on the share price of May 28th, at least a plurality of analysts are forecasting some level of multiple compression; it makes price target setting a fraught undertaking.
There is, perhaps a bit more murkiness, when it comes to evaluating Palantir’s multi-year CAGR and that is a function of the long standing comment of by the CEO, “Per long-term guidance policy, as provided by our Chief Executive Officer, Alex Karp, we continue to expect:
Annual revenue growth of 30% or greater for 2021 through 2025.”
This comment appears regularly as part of the guidance section in the quarterly earnings release-just my opinion-but I think the company ought to drop the statement or revise it to take some account of what appears to be happening in the market.
In turn, this has led to consensus forecast that have 2022 revenues rising by just 30%. It should be reasonably obvious that no one owning the shares can believe such a forecast and analysts who recommend the shares can’t really do so with a straight face using a 30% revenue growth estimate.
That said, Palantir shares certainly don’t have a particularly strong consensus rating compared to many other enterprise software names. First Call suggests that on average the rating is between a hold and an underperform. At the moment, however, only 7 ratings and 8 estimates are being reported to First Call. Most estimates were raised in the wake of the latest earnings report.