$PLTR | Palantir deep dive
Enabling data-driven decision making
If you are a current subscriber, welcome back to another deep dive! If you are not currently subscribed, feel free to do so! It’s great to have you here!
0. OBJECTIVE AND DISCLAIMER
The objective of this document is to provide a deep dive on Palantir Technologies ($PLTR) but it should never substitute your own due diligence. I would never recommend investing or trading solely on third party information.
The following analysis was conducted in the weeks before the 3rd of June 2021 so it does not take into account any information released about the company after that date.
I’m not long $PLTR as of the moment of this writing.
I post these analyses on my Twitter account so you can follow me if you want to stay updated @invesquotes. You can also subscribe to my newsletter, this way you will receive further deep dives directly in your email.
During this deep-dive we will talk about the following:
2. Key Highlights
4. What Palantir does and how it makes money
5.The business model
7. Key performance indicators (KPIs)
8. Management Quality and Culture
9. Financial Statement Analysis
10. Common Stock
12. Palantir’s opportunity
14. Why Palantir?
16. Performance and valuation
17. Final word
2. KEY HIGHLIGHTS
Palantir builds software platforms that enable institutions to analyze big data sets by transforming them into an integrated data set that reflects their operations.
The company was founded in 2003 with the vision of reducing terrorism while preserving civil liberties.
The company came public via a direct listing in 2020 and is currently valued at $46 billion with a stock price of $24. The company trades in the NYSE under the ticker $PLTR.
The company has two main products (Gotham and Foundry) that are powered by a third product (Apollo). Gotham targets governments while Foundry targets commercial institutions.
Palantir follows a land and expand model that has three phases: acquire, expand and scale. The company also uses partnerships to sell its products in other industries and countries.
All of the company’s KPIs (number of customers, contribution margin, ARPC…) improved in the most recent quarter.
Three of the co-founders still remain in the company after 18 years.
Palantir shows strong revenue growth but it is not expected to be profitable in the coming years. Big cash position gives the company optionality.
The business was destroying cash but turned CFO and Adjusted FCF positive in the most recent quarter.
Multiclass structure to concentrate voting power around founders. Significant stock dilution in the past years.
Big growth potential, specially in the commercial segment.
On the downside, the company currently trades at 38 P/S and has significant execution risk with a fairly new sales team.
Palantir was founded in 2003 (according to SEC fillings) by Peter Thiel, Nathan Gettings, Joe Lonsdale, Stephen Cohen and Alex Karp with the vision of reducing terrorism while preserving civil liberties.
The company initially struggled to find investors and they only had $32 million of funding, $30 million of which came from Thiel’s venture capital firm. The other $2 million came from In-Q-Tel (the CIA’s venture capital arm). The company thought that to defeat an adaptive adversary they would need human analysts to explore data from many sources.
During 2009 and 2010, Palantir’s software uncovered the Ghost Net and the Shadow Network, two Chinese-based espionage networks. In 2010, the company partnered with Thomson Reuters to sell Palantir Metropolis product as a quantitative analysis tool. That same year, vice president Joe Biden credited Palantir for helping the federal government to fight fraud, announcing that it would be deployed on other government agencies.
In September 2013 the company disclosed $196 million in funding and Alex Karp (CEO at that moment) announced that the company would not pursue an IPO as being a public company would make “running a company like Palantir very difficult”. That same year the company raised an additional $450 million from private investors which valued the company around $9 billion. During the following years the company received additional funding which valued the company at $20 billion in 2015.
During 2016, the company acquired Kimono Labs (a startup that made easy to collect information from public facing websites) and Silk (a data visualization startup). During 2020 the company helped the NHS to manage the pandemic and developed a vaccine allocation software used in the United States. This same year, in September, the company became public through a direct listing under the ticker “PLTR”.
4. WHAT PALANTIR DOES AND HOW IT MAKES MONEY
4.1. What the company does
Palantir builds software platforms that enable institutions to analyze big data sets by transforming large amounts of information into an integrated data set that reflects their operations. It basically helps these institutions make data-based decisions. It’s important to note that Palantir does not create or does not sell data, it helps interpret data that already exists. The company currently has the following products:
Gotham: it’s the company’s first software platform (launched in 2008) and it was built for analysts at defense and intelligence agencies. It enables users to identify patterns hidden deep within data sets.
It is used broadly across government agencies although it is also offered to commercial customers, specially those in the financial services industry to fight fraud.
Below you can see more or less how Gotham works. If you want to take a look directly at how this software works, you can do that here but be ready to think you are playing Call of Duty.
Source: Palantir website
Foundry: this software solution was built in 2016 to target commercial institutions with the objective of transforming the ways in which organizations interact with information by creating a central operating system for their data.
It is basically a data integration and management platform, a suite of analytical tools and an operational platform for business users. The core objective of Foundry is to translate data and models into outputs that human operators can use to make operational decisions. This is what I can Imagine Foundry doing:
The two silos can be two departments of a company where data is not connected by any specific format. What Palantir aims to do is to create “A” with the information of both departments so they can act on the information of the whole company in an integrated data set.
One key characteristic about Foundry is its modularity. Institutions don’t need to deploy the whole product, they can’t take what they want so it is a very flexible offering.
The advantage of these products is that they are all vertically integrated which allows users to collaborate effectively in the platforms without needing any technical background. These platforms can be used independently or bundled together in a single ecosystem.
Although these products are top-tier by themselves, the company has built a system that gives these products an edge over those of its competitors (specially over those that are competing for government contracts): Palantir Apollo.
Apollo: it’s not a product per se but it was built by the company to power Gotham and Foundry. It’s a continuous delivery system that allows Palantir’s products to run in places where no other SaaS platform can reach such as disconnected environments (drones and submarines) or purpose-built government or classified clouds. Apollo gives Palantir an important edge over other traditional SaaS products that typically run in a single public cloud.
Another important feature of Apollo is that it enables the company to continuously update the different services included in the platforms (Gotham and Foundry) as it is configured to decide what to upgrade, when to do it and how. Apollo allows Palantir to deploy updates more than 41,000 times a week which enables its users to always have cutting-edge technology while reducing the workload of the company’s engineers.
In the diagram below you can see how it works:
Source: Palantir website
Recently, the company announced the launch of Apollo for Edge AI which enables customers “to train, manage and deploy multiple independently versioned chained models to the Edge with ease”. This is a proprietary technology that will most likely allow the company to act independently from edge providers such as Cloudflare and Fastly. Here you have a good thread to see why this offering is very important for Palantir.
Another interesting thing that is worth pointing out is that even though the company was founded in 2003, it didn’t sell to its first customer until 2008. So basically this tells you that they had been working on the product for 5 years before even being able to sell it. Alex Karp in an interview said that, during the early life of the company, going public was not an option as they didn’t want to be publicly exposed while developing their platforms.
There is an ongoing debate between bulls that argue that Palantir is a SaaS company and bears that argue that the company is a services company which will suffer significant margin compression as it grows.
4.2. How does Palantir make money?
The company generates revenue from the sale of subscriptions which enable institutions to use their software. These subscriptions come from three main sources:
Palantir Cloud: these subscriptions grant customers the right to access software functionality in a hosted environment controlled by Palantir.
On-premises software: these subscriptions grant customers the right to use Palantir’s software in the customer’s internal hardware infrastructure or on their own cloud instance.
Both of these services are sold jointly with O&M services which include critical updates and support and maintenance services required to operate the software.
Professional services: support the customer’s use of the software with on-demand user support, user-interface configuration, training…
5. THE BUSINESS MODEL
The company’s business model with which they plan to acquire and grow their customer accounts has three phases: acquire, expand and scale. It’s basically a land and expand model.
The company sends Forward Deployed Engineers (FDEs) to their potential customers. These are engineers that directly work hand in hand with customers to understand their data and the problems that they are trying to solve. They are basically like consultants that observe the customers’ challenges and tailor Palantir’s products to the specific needs of each one. This is one of the reasons why many people see Palantir more as a services firm than a SaaS firm, because the customizable platforms make the business model quite capital intensive. As we will see later on, modularity should help the company end with this stigma as FDEs spend less and less time per customer.
The business is currently focused on big institutions but in their most recent earnings call the company said that they are starting to move down market to medium-sized enterprises. Moving down the market is possible thanks to the significant investments the company made in modularity. As you can imagine, Palantir’s products are more likely to add value to big enterprises with huge amounts of dispersed and unstructured data.
You can find a brief description of the three phases below but before going over these phases I want you to see the definition of contribution margin as it plays an important role in determining in what phase customers are located:
“We define contribution margin as revenue less our cost of revenue and sales and marketing expenses, excluding stock based compensation, divided by revenue.”
Contribution margin is a non-GAAP metric that measures the efficiency of the business.
During the acquire phase, Palantir provides short-term pilot deployments of their software at no or minimal cost to the customer with the objective of making the customer see the value that their platforms can provide. The company includes customers in the acquire phase if they have not recognized more than $100,000 in revenue during that calendar year.
The company usually operates accounts in this phase at a loss although in 2020 customers in this phase generated $77.1 million in revenue which yielded a contribution profit of $13.2 million (17% contribution margin).
During the last quarter (Q12021), customers in this phase generated $3.6 million in revenue which yielded a contribution loss of $4.3 million.
During this phase the company makes significant investments with the objective of understanding the challenges faced by their customers. The company includes in this phase customers from which they have not recognized more than $100,000 in revenue in the calendar year and whose account had a negative contribution margin during the year at issue.
The company normally operates at a contribution margin loss during this phase although during 2020 customers included in this account generated $360.4 million in revenue with a contribution margin of +47%.
During the last quarter (Q12021), customers in this phase generated $12.4 million in revenue which yielded a positive contribution margin of 4%.
During this phase, platform usage increases and investment costs generally decrease. This occurs because the core platform is already set up and building modules on top is much faster. Customers start becoming more self-sufficient although they can continue benefiting from Operations and Maintenance (O&M) services. The company includes in this account customers from which they have recognized more than $100,000 in revenue during the calendar year.
In 2020 the company generated $613 million in revenue from this account with a contribution margin of 70%. During the last quarter (Q12021), customers in this phase generated $324.8 million in revenue which yielded a contribution margin of 66%.
As you can already imagine, the objective of Palantir is to make customers flow through the different phases until they end up in the scale phase. I think that this business model clearly helps the company stay focused on the long term because they have to make efforts to make their initial heavy investments payoff over the long term.
As it is obvious, contribution margin is expect to increase throughout the different phases as the company scales its product:
Sales cycles can take from 6-9 months to over a year depending on the customer. Palantir is incurring significant time and money in acquiring customers which makes sense because they are trying to offer a customized product to every customer to help them solve their specific problems. This approach differs from the more traditional SaaS companies that offer a more generic product / service. This approach brings higher CAC (Customer Acquisition Cost) but at the same time it offers a competitive advantage for Palantir as customers stay with them for very long periods of time (high contract duration).
This said, the modularity of Foundry is helping the company shorten the sales cycles as pilots can be deployed much faster now and it should also help the company target smaller companies.
Palantir acquires customers through different channels. The company plans to invest heavily in its direct sales and marketing team to target commercial growth but to succeed in its growth efforts in this segment, the company is also building partnerships and joint ventures. The idea of the company is to use the relationships of these partners with their customers to cross-sell their product.
One of the most successful partnerships the company has built up to date is their partnership with Airbus to develop the platform Skywise which aims to target the whole aviation industry. It’s an open data platform that allows all of the players in the aviation industry to obtain valuable insights from the data they produce. Although this partnership provides strategic advantages for Palantir it also limits its ability to independently provide their services to Airbus’ competitors. Up to this date, the platform connects more than 9,000 aircraft across more than 100 airlines.
In February this year, the company also announced a partnership with IBM to build a new product that will leverage Palantir Foundry and will integrate with IBM Cloud Pak for Data services. The idea of the product is to help businesses accelerate their digital transformation using Artificial Intelligence. IBM booked the first customer within 16 days of the product’s general availability.
The company also uses partnerships/joint ventures to expand internationally as it has been the case in Japan. Palantir Japan was jointly founded by Sompo Holdings and Palantir in 2019 with the idea of expanding the company’s products to that country with the help of a domestic company (Sompo Holdings is one of the largest Japanese insurance companies). Under this arrangement Palantir enjoys 50% profits of the combined company.
The company also initiated a joint venture in 2020 with Merck under the brand name Syntropy with the objective of empowering scientists and research centers with a collaborative technology platform.
My thoughts about partnerships are that they are a great way to grow in the commercial segment but they also limit the profitability of the company because margins through partnerships are lower. Now that the company is expected to invest significantly into their direct sales force I would like to see how direct selling improves relative to partnerships. Going forward, the company expects to grow using both direct sales and sales via partnerships.
7. KEY PERFORMANCE INDICATORS
When looking at Palantir I think that the following KPIs should be analyzed: number of customers, revenue per customer, average contract duration and total commercial deal value. I would also consider financial measures such as revenue or contribution margin to be KPIs but we will go through these measures more in detail in the Financial Statement Analysis section.
7.1. Number of customers
As discussed above, Palantir has two distinguished customer segments: government and commercial. The company defines a customer as:
“Organization from which we have recognized revenue during the trailing twelve month period.”
Palantir currently has 149 customers. The number of customers has evolved in the following way during the last quarters:
Source: 10Qs of the company
Although this might not seem many customers you have to remember that the company’s target is big institutions and as such you would expect to see a low number of customers with moderate growth sequentially but with high revenue per customer.
This said, the company plans to expand to medium-sized institutions in the coming years. The company said the following during the earnings call:
“Since the beginning of February, qualified commercial opportunities in the US and the UK are up 2.5x. Active commercial pilots across the business have more than doubled.”
This metric should be analyzed jointly with average revenue per customer to arrive at worthwhile conclusions.
7.2. Revenue per customer (ARPC)
This is another important metric to look at as if this metric is increasing it means that the company is being able to begin new contracts with existing customers or expand/renew existing contracts. The company calculates the global ARPC and this same metric for their top 20 customers. For the last quarter (Q12021) both of these metrics showed strong growth:
Source: Palantir Q12021 Earnings Presentation
Going forward, the company expects revenue per customer to decrease as they expand their user base to middle market and SMBs. So basically you have two components of revenue: total number of customers and average revenue per customer. The company is telling investors that they expect to grow the number of customers going forward while diminishing the average revenue per customer as they move down the market.
7.3. Average contract duration
This metric can be used to analyze the stickiness of Palantir’s products. High contract durations mean that institutions build long lasting relationships with the company. As Palantir uses AI for data mining purposes, the longer any customer remains with Palantir, the more difficult it is for this customer to change to another provider as switching costs increase.
The company reports this metric for all customers and for contracts of commercial customers. Average contract duration for all customers has evolved in the following way:
Source: Palantir quarterly earnings presentations
It’s also important to note that most of the contract can be cancelled at will by the customers so these numbers don’t assure that the full contract will be fulfilled. In the last earnings call, the company said the following about contract duration:
“Average contract duration was 3.7 years, up from 3.6 years at the end of 2020. These results are particularly strong when considering government customers entered into shorter-than-usual contracts in 2020 to accelerate procurement as they responded to the COVID crisis.”
This explanation makes sense if we look at this same measure for commercial customers which is higher at 4.6 years (remember that Foundry has been in the market for 5 years). Going forward the company expects tailwinds for contract duration as government contracts renew and expand.
7.4. Total Commercial Deal Value
The company signs long term contracts but recognizes revenue over the contract term. This means that revenue is not capturing the whole potential of the company as it does not take into account money that is already “signed” but not yet recognized. This revenue recognition policy makes total sense especially if you take into account that many contracts can be terminated beforehand.
The company finished Q12021 with $5.4 billion in total remaining deal value + IDIQ contracts. IDIQ contracts refer to contracts with the government where the funding of such contracts has not been yet determined and is not guaranteed. If we only take into account the total remaining deal value, the number is $2.8 billion.
These numbers mean that if none of the customers terminate these contracts beforehand the company has $2.8 billion waiting to be recognized in revenue in the coming years.
8. MANAGEMENT QUALITY & CULTURE
In this section I will go over management, the vision of the company and corporate culture.
8.1. Management Quality
In my opinion, Palantir has great potential in its management team. In this section I will talk specifically about Alex Karp, Peter Thiel and Stephen Cohen. These three co-founders still remain in the company and are three great assets of the team in my opinion.
Alex Karp is one of the founders of Palantir and has been CEO since the company was founded. He is not a usual CEO, he pursued a PhD in Philosophy before founding Palantir and teaches meditation classes in the company from time to time. He converted Palantir into a multibillion company so I think that his track record can’t be questioned. He is quite eccentric, he once said that he is always thinking about Palantir except "when I'm swimming, practicing Qigong or during sexual activity."
This said, there has been plenty of controversy around his figure, especially regarding some of the company’s work with the government and his compensation. His work with the government has brought protests even from employees who didn’t want to sell the platform to the government for illegal immigration purposes. Protesters have even gone to the extent of protesting in front of his house.
He has been very critical of other tech companies in the US who don’t want to support the government by saying that their “CEOs confuse having a high IQ with being sensible.” He is well known for being critical of Silicon Valley as a whole, especially of those who sell data “without creating real value to customers.”
He received stock options that placed his compensation in more than $1 billion dollars in 2020 which is kind of crazy. IMHO, I don’t think that such a high compensation is necessary because it is coming at a cost: shareholder dilution. I think that you can like or dislike Alex Karp as a person, but as a CEO I think he has more than a proven track record! Regardless if you agree or disagree with him, he has shown his ideas in a transparent manner in many of the interviews he has done.
I could also talk a lot about Peter Thiel but I think that there is not much that I could say that could add value to you. Peter Thiel currently serves as Chairman of the Board and has a more than proven track record of successes managing many companies as well as investing in others such as Facebook and SpaceX.
Last but not least I would like to talk briefly about Stephen Cohen. He is the brain behind Palantir’s software and currently serves as President and Secretary of the company. He stands out to be a very clever guy who had a vision of changing the world since he started his studies at Stanford. He has shown great consistency and determination until coming up with what Palantir would become today.
He shares an anecdote in this interview where he says that he had to develop a prototype for a meeting with In-Q-Tel’s president and he only had eight weeks to do it. This meeting was in fact what opened the doors of the government to Palantir so we can say that he was one of the building blocks of the company.
In conclusion, I think that the company’s management team is high quality and has a proven track record of success. You might like their ideas more or less but you can’t deny that they are being transparent about what they want to do with the company. The only thing I don’t like is the somewhat excessive compensation of Alex Karp. Obviously, as you would imagine after seeing the high SBC expenses, management has significant skin in the game.
The company has always argued that they know that computers and technology have some limitations and they will never be able to fully replace humans. What the company is focused on achieving is facilitating the interpretation of data for human analysis. Their end goal is to augment human intelligence, they are not aiming to replace it.
In an interview in the early days of the company, Alex Karp mentioned that the aim of Palantir was to build a product that could be sold to many companies but this was difficult at first because big institutions have large amounts of unstructured data that differ from institution to institution. Even though the company still deploys a big amount of FDEs to sell their products, they seem to be working on the scalability of the product.
8.3. Company Culture
The company is focused on the well being of its employees and this can be seen by the policies they have in place to allow them to balance work and life. The truth is that Palantir is very focused on hiring the most talented people and this can be seen by the high SBC expenses of the company.
Taking a look at Palantir on GlassDoor we can see the following:
Overall rating: 4.2/5
Values and culture: 4/5
Diversity and inclusion: 4/5
Personal/Work balance: 3.3/5
Career opportunities: 3.8/5
Overall this is a good rating for the company. Take into account that the controversies of the company probably have affected these ratings. 82% approve of Alex Karp and 83% would recommend the company to a friend.
9. FINANCIAL STATEMENT ANALYSIS
In this section I will go over the main financial statements of the company: the income statement, the balance sheet and the statement of cash flows. I will try to go into detail in the income statement and the statement of cash flows while I’ll simply go over the most important accounts of the balance sheet.
9.1. Income statement
Firstly, I will look at the income statement focusing especially on revenue and operating costs, their evolution and what can be expected going forward. Making a distinction in this section between adjusted measures (non-GAAP) and GAAP measures is key because the company has significant share-based compensation expenses.
As I mentioned before, the company generates revenue by the sale of subscriptions that enable institutions to use their software and the sale of professional services. Revenue is generally recognized during the contractual period so declines in contracts or renewals don’t tend to affect revenue until later quarters. On the flipside, contractual increases or renewals don’t tend to positively affect revenue until later quarters.
Sales experience seasonality and they tend to be lower in Q1, showing sequential increases in subsequent quarters towards year end which is consistent with the fiscal year end of many of the company’s customers.
Yearly revenue for the company has evolved in the following way (in millions):
Yearly revenue has exhibited the following YoY growth rates:
As you can see, revenue growth has accelerated lately despite the pandemic and this was mainly caused by high growth in the government segment. In their latest earnings release, the company said that it expects to see +30% growth from here up until 2025. Assuming the low end of this estimate (30%) we could expect to see revenues of about $4.05 billion in 2025. I do think that this estimate is quite conservative and I would expect to see a minimum of 40% growth in the coming years.
Taking a look at the recent quarterly results, the company reported revenues of $341 million which was a 49% YoY growth. In the following graph you can clearly see that revenue for any year grows sequentially as quarters go by (in thousands):
Revenue distribution across segments
As I have mentioned before in this report, Palantir has two main customer segments: government and commercial. Prior to the release of Foundry, the company mainly worked with the government but since releasing Foundry the company also added commercial institutions as customers. This is the current distribution of revenue across segments:
Even though commercial came later and you would expect it to grow at a faster pace, the government segment is growing faster and increasing in weight:
Taking a look at the revenue YoY growth rates for both segments during the last quarters we can explain the above:
As much as the weights for these segments and their growth rates might seem a bit confusing at first sight, there might be a logical explanation for the numbers. I say that they might be confusing because before digging in I would’ve expected the commercial segment to grow at a higher rate (or maybe similar) than the government segment as it is in its early stages of growth.
The truth is that in 2020, the pandemic affected both segments in opposite directions:
Government segment: as the pandemic unveiled, the company worked closely with the US and UK governments to control it, opening their relationship with the healthcare departments in these countries.
On top of this, the company has signed 9 new contracts with the government since the end of Q1 and is expected to enjoy tailwinds from President Biden’s Executive Order around cybersecurity.
Commercial segment: to face the pandemic, many companies aggressively cut their expenses and this affected Palantir because companies were less willing to spend on Analytics. The good thing for the company is that the pandemic generated strong tailwinds going forward as companies have realized that they will be generating increasing amounts of data.
Another thing that is important to note is that Palantir did not even have a sales department in 2018 and most of their contract negotiations were reliant on management, which clearly limited their ability to expand across commercial customers. The company has already said that they plan to invest aggressively in the expansion of their sales department (hired 50 sales people just in Q1 and guiding for triple digit hires in the full year) and in expanding their partnerships.
At first, I was surprised when I saw the distribution of revenue across segments, especially its trend. Going forward, I would like to see the commercial segment gaining weight relative to the government segment which will give an indication of the true potential of Foundry and will also indicate that the company’s investment in the sales team and partnerships is paying off.
Revenue distribution across geographies
Although the company does not directly disclose this separation it did report “Annualized revenue run rate” from US customers which is basically domestic revenue multiplied by 4 (number of Qs). From this we can calculate US revenue for 2021Q1 at $200 million which would give the following distribution across geographies in the most recent quarter: 59% US and 41% international. We can say that the company is well diversified geographically speaking.
In the last quarter the company reported that the US commercial segment grew 72% YoY and that the US government segment grew 83% YoY. Seeing that both of these growth rates are higher than the overall growth rates of the quarter, it must mean that the US is driving global revenue growth, especially in the commercial segment.
It’s also important to note that, even though the company generates recurrent revenue in the form of subscriptions, most of their contracts can be terminated by the customer at any time for convenience. The company does not disclose how many contracts are terminated beforehand every period which would serve as some sort of NPS proxy but I don’t think that this number would be significant because before signing any contract, every customer goes through the whole sales cycle and is able to evaluate the value add.
Many investors argue that customer concentration is an issue for the company. In the S-1 filing the company discloses how much revenue the top 3 customers account for:
Obviously, looking at the numbers you can see why this might be a concern for some people but the trend is positive and the company is increasingly diversifying revenue across customers. Also note that the top three customers have been with the company for an average of 8 years so they have built meaningful relationships with them. Once the company continues to consolidate their relationship with newer customers, revenue per customer should diversify further.
Another important factor to take into account is that revenue can be volatile because the company is focused on big institutions. For this reason, two or three big contracts can dramatically change revenue growth.
9.1.2. Cost of revenue
Once we have looked at revenue, it’s time to take a look at cost of revenue to see if the company is being able to generate more revenue with less direct costs (i.e. improving gross profit margins).
Cost of revenue includes the following (source 10K):
“Salaries, stock-based compensation expense, and benefits for personnel involved in performing O&M and professional services, as well as third-party cloud hosting services, allocated overhead, and other direct costs.”
As Stock-based compensation is included in these costs I will also show “Adjusted Cost of Revenue”, the Non-GAAP measure.
Cost of revenue for the past years has evolved in the following way (in millions):
While the CAGR of revenue for these years has been 22.4%, the CAGR for cost of revenue has been 28.7% which must mean that gross margin (non-adjusted) has been worsening. This can also be seen if we calculate cost of revenue as a percentage of total revenue:
The company has a great component of stock-based compensation included in cost of revenue which I normally take out of the calculation because I expect it to fade away as the company matures. Palantir also takes out the effect of stock-based compensation in the calculation of “Adjusted Gross Profit”. The company allocated the following stock-based compensation expenses to cost of revenue (in thousands):
As you can clearly see in the table above, stock-based compensation was especially significant in 2020 as the company came public.
The increase in cost of revenue from 2019 to 2020 was mainly caused by share based compensation and an increase in headcount (O&M) to support the greater volume of customers which were partially offset by a decrease in travelling expenses due to COVID. Third party cloud hosting costs also decreased due to volume based discounts.
Quarterly cost of revenue varies widely between periods due to the timing of the share based compensation expenses. From 2019 up to today quarterly cost of revenue as a percentage of total revenue has been in the range of 52% and 22%.
9.1.3. Operating costs
After looking at cost of revenue I would like to go through operating expenses which are key when it comes to analyzing how the company is being able to scale as it grows.
Sales and marketing costs
As I have already mentioned, these costs are expected to increase significantly going forward as the company expects to invest significantly in their sales and marketing team. These costs are incurred in all of the stages of the sales cycle (Acquire, Expand and Scale) and they include the following (source 10K):
“Salaries, stock-based compensation expense, and benefits for personnel involved in executing on pilots and customer growth activities, as well as third-party cloud hosting services for our pilots, marketing and sales event-related costs, and allocated overhead.”
These costs, including stock-based compensation expenses, have evolved in the following way during the past years (in thousands):
If we take out the effect of stock-based compensation expense we arrive at the following numbers (in thousands):
As you can see from the numbers above, Sales & Marketing expenses include a huge amount of stock-based compensation expenses which makes sense because the company is basically investing in expanding the team and they are choosing this compensation method. The company indicates that these increases in sales and marketing costs come mainly from increased personnel costs which were offset by reduced travel expenses and temporary office closures.
As it is the case with cost of revenue, these costs vary widely between quarters due to the timing of share based compensation expenses.
Research and development costs
These costs are also significant for the company and this is what you would expect from a SaaS company that is constantly identifying new use cases for their software for many industries. The company indicates that R&D includes the following:
“Salaries, stock-based compensation expense, and benefits for personnel involved in performing the activities to develop and refine our platforms, internal use third-party cloud hosting services and other IT-related costs, and allocated overhead.”
These costs, including stock-based compensation expenses, have evolved in the following way during the past years (in thousands):
If we take out the effect of stock-based compensation expense we arrive at the following numbers (in thousands):
As you can see, these costs follow a similar trend than sales and marketing costs as they have mainly been increasing due to share based compensation costs. When we take these costs out of the equation, R&D costs have actually been decreasing as a percentage of revenue.
General and Administrative expenses
These costs include the following:
“Salaries, stock-based compensation expense, and benefits for personnel involved in our executive, finance, legal, human resources, and administrative functions, as well as third-party professional services and fees, and allocated overhead.”
These costs, including stock-based compensation expenses, have evolved in the following way during the past years (in thousands):
If we take out the effect of stock-based compensation expense we arrive at the following numbers (in thousands):
These costs follow a similar trend than S&M and R&D as they have mainly been increasing due to share based compensation costs.
Operating costs summary
In the following table you can see a summary of the operating costs of the company including stock-based compensation expense (in thousands):
As you can see from the table above and we have seen throughout this section, operating costs have increased this year as a percentage of revenue due to stock-based compensation expenses primarily related to the direct listing in 2020. Don’t get me wrong, even though these costs are non-cash they should be taken into account because they are basically paying employees with stock and it’s diluting your ownership.
The same way as I can understand that S&M and R&D costs can increase or to stay flat as a percentage of revenue going forward due to aggressive investments in these areas, I would like to see how G&A costs decrease as a percentage of revenue. You have to take into account that Palantir is still in its early stages of growth as they expand the commercial segment of the business so I’m more focused on the top line rather than the bottom line. As the business grows, I think that the company will achieve profitability sooner than later due to the inherent scalability of the business. If we take a look at the evolution of operating costs as a percentage of revenue (without including SBC) we see that they were decreasing steadily:
Source: Seeking Alpha
9.1.4. Other relevant costs
Non-operating costs don’t seem relevant when looking at the income statement. The only non-operating cost that I see that is increasing lately is interest expense which includes primarily interest expense and commitment fees incurred under the company’s credit facilities.
Interest expense for 2020 was $14.1 million, up from $3.1 million in 2019. This increase was caused primarily because there was practically no debt outstanding up until Dec, 2019.
When it comes to profitability, companies that have such a significant portion of expenses tied to stock-based compensation tend to calculate both an Adjusted measure (non-GAAP) and a GAAP measure.
GAAP measures (Net Income and Operating Income)
The company is currently unprofitable at the operating and net levels and losses have increased during the last year (in thousands):
From the table above you can see that profitability slightly improved from 2018 to 2019 but worsened from 2019 to 2020 caused primarily by higher stock-based compensation expenses as part of the direct listing. The company is not expected to achieve profitability soon on a GAAP basis as they invest in Sales and Marketing and R&D. I would however expect profitability in 2021 to improve significantly as share based compensation for the year is expected to decrease if compared to 2020.
Non-GAAP measure (Adjusted Operating income)
Adjusted Operating income for the company excludes stock-based compensation expense and related employer payroll taxes. Adjusted Operating income has evolved as follows in the last quarters (in millions):
The company was not profitable in an adjusted operating income basis in 2019 but it turn profitable in 2020 (in millions):
Adjusted EBITDA margins are improving which is really positive for the company. Even though the company is not profitable on a GAAP basis I would expect the company to turn profitable in the coming years as stock-based compensation falls to normal levels.
In this section I will analyze the margins of the company and their evolution. Here we have to take into account two non-GAAP metrics that the company calculates: adjusted gross profit margin and contribution margin.
Gross profit margin
The company’s GAAP gross profit margin is high at 68% (2020). They calculate Adjusted Gross Profit Margin to eliminate the effect of share based compensation in cost of revenue, which was significant in 2020 as part of the direct listing. Adjusted gross profit margins were 80.5% (2020) which is significantly higher than GAAP Gross margins obviously.
This is how both these metrics have evolved in the past three years:
As you can clearly see in the chart above, adjusted gross profit margins have significantly improved during the last year while GAAP gross margins have shown a moderate improvement. As you can imagine, the difference between both lies in a high stock-based compensation expense for the year 2020.
EBITDA margin and Net Margin
The company currently has a negative EBITDA margin of (107%) and a negative net income margin of (107%). These margins have evolved in the following way:
As I mentioned before there are no significant non-operating costs and this is the reason why EBITDA and Net Margins are almost identical. These margins have worsened during the last year due to increased stock-based compensation expenses as part of the direct listing.
The company considers contribution margin to be its key performance indicator. I already gave in the “Business model” section a definition of this metric but I will bring it down here again:
“We define contribution margin as revenue less our cost of revenue and sales and marketing expenses, excluding stock based compensation, divided by revenue.”
This metric basically measures the effectiveness of the company to generate sales with costs that are directly related to these sales.
Yearly contribution margin has evolved in the following way:
Quarterly contribution margin has also been improving in the last quarters:
As the company improves its ability to flow customers from the acquire phase to the scale phase, contribution margin should improve even further and it should approach 70% which is the contribution margin of the scale phase.
9.1.7. Final thoughts on the income statement
Just to wrap up the analysis of the income statement I would like to write a small summary.
Strong top line (revenue) growth which is expected to be maintained at +30% from here until 2025. As I said, I find this quite conservative
High gross profit margins
Stock based compensation makes up a big part of operating expenses
No significant non-operating costs
Not profitable on a GAAP basis but profitable on an adjusted basis
Improving gross profit and contribution margins
9.2. Balance sheet
In this section I will briefly go over Palantir’s balance sheet looking mainly at cash, liabilities and contractual obligations for the company.
Assets in the balance sheet were divided in the following way as of end of Q1,2021:
The main current asset is cash and cash equivalents which stood at $2.34 billion at end of quarter. This gives the company great optionality to invest in growth ventures.
Liabilities were divided in the following way between current and noncurrent as of end of quarter:
Palantir just has $198 million of non-current debt.
The company is going to face the following payments (contractual obligations) in the foreseeable future (in thousands):
Most of the contractual obligations come from noncancelable purchase commitments which are related to purchase commitments for third-party cloud hosting services which means that they won't probably be postponed to posterior periods.
9.2.3. Some thoughts on the Balance Sheet
Although I do think that the balance sheet is important I don’t tend to go much in detail into its analysis because it shows a static picture of the company. I prefer to go much more in detail in the income statement and the statement of cash flows because they will paint a clearer picture of the performance of the company.
When it comes to the balance sheet I usually look for a high cash position (optionality) and low debt or debt that the company can easily repay without the need to refinance themselves through a share or debt offering.
9.3. Cash flows
Cash Flow from Operations (CFO), Cash Flow from Investing Activities (CFI) and Cash Flow from Financing activities (CFF)
CFO, CFI and CFF have evolved in the following way for the past years (in thousands):
In the most recent quarter the company has seen a change in trend in both CFO and Adjusted Free Cash Flow:
Operating cash flow was positive at $116.8 million in Q12021 (34% margin)
Adjusted free cash flow was positive at $151 million in Q12021 (44% margin)
Adjusted free cash flow is defined as CFO plus cash paid for employer taxes related to SBC, less cash used to purchase property and equipment.
The fact that these measures turned positive is very good news for the company but I would like to see if the trend continues in the coming quarters.
10. COMMON STOCK
Palantir’s stock is a topic that is widely discussed due to the significant share dilution that shareholders have been experiencing lately. Before going into this topic, let's take a look at the common stock structure of the company.
10.1. Multiclass Structure of Common Stock
The company has a Multiclass Structure with the objective of concentrating voting power among the Founders and their affiliates. The company has three types of common stock:
Class A shares
Shares that are publicly traded in the NYSE, hold 1 vote per share. As of May 7th there were 1,805,225,985 class A shares outstanding.
The lockup period for these shares ended February 18, 2021 and the stock dropped almost 8% that day.
Class B shares
Shares that are not publicly traded, hold 10 votes per share. As of May 7th there were 70,515,320 class B shares outstanding. These shares are convertible into Class A shares at any time.
Class F shares
Shares that are not publicly traded, hold a variable number of votes per share which gives the Founders an aggregate of 49.9% of voting power thanks to the Founder Voting Trust Agreement and the Founder Voting Agreement. This share class is convertible at any time to class B shares at the option of the holder.
As of May 7th there were 1,005,000 class F shares outstanding.
So basically this share structure makes the company a truly founder led company. As a shareholder you will probably have no say in Palantir’s decisions. I’m usually okay with this because a great component of my bull thesis for many of the companies I hold is their management.
10.2. Stock dilution
Any shareholder should be aware that their ownership is being significantly diluted due to the company’s preference to pay employees with stock rather than cash. This is something quite usual in companies in their early stages of growth that prefer this type of compensation as it is non-cash and will allow them to hire quality employees to grow without incurring in significant debt
Palantir has incurred significant higher stock-based compensation costs in 2020 if compared to the year 2019. Stock-based compensation expenses have evolved as follows (in thousands):
If we take a look at how the weighted-average class A shares outstanding is evolving we can see the following:
If we take a look at the data for the most recent quarter we can see that the company had 1,821,158,000 weighted-average shares of common stock outstanding as of the end of March 2021 while they only had 591,850,000 shares outstanding during the comparable quarter in 2020 (307% increase Y/Y).
Once we already know that SBC expenses are significant we have to study if it is expected to go forward. The company presented the following balances and weighted-average service periods for the SBC outstanding as of March , 2021:
In the table above we can see that the company may need to issue 652 million of shares during the coming years which is more or less an additional 35% of end of March values. If we look at the average prices of the table above you could conclude that the most likely outcome is that all of the SBC will end up being recognized in the coming years and so investors have to expect significant SBC expenses going forward.
This said, management has said that as the company matures, stock-based compensation should fall into normal levels.
Analyzing who holds the stock is also important as it may give us some important clues. In this section I will take a look at institutional ownership and inside ownership.
11.1. Institutional Ownership
I normally check which institutions hold the stock to see if I can find some of my preferred funds holding it. This should never substitute due diligence but it may be another positive sign of the quality of the company. You should also check if the stock is overheld by institutional investors as this will only create significant downside if they decide to sell.
Institutional ownership of Palantir is at 22% if we take into account the ownership of Sompo Holdings (Japanese partner) and the 5 institutions that held the most shares at 31/12/2020 were the following:
Note: Palantir is not currently trading as part of any index so these holdings are not held for index funds. ARK has been buying heavily the “dip” on Palantir lately.
11.2. Inside Ownership
Individual insiders own approximately 10% of the company. Two of the co-founders that still work in the company have a significant stake in the company, specially Peter Thiel:
Insider interests are likely to be aligned with those of shareholders. If management has no skin in the game I would probably consider it as a red flag.
Remember that, as we saw before, the founders combined hold 49.9% of the voting power of the company through their holdings of Class F shares.
Insider selling has been significant and is one of the main reasons why management has been criticized for. As I always say, I don’t consider insider selling to be a red flag in itself because insiders may be selling for a wide variety of reasons but I may consider it as a warning sign if management is reducing their stake significantly. I think it would be stupid to think that a manager that founded the company 18 years ago wants to cash out now to be honest.
12. PALANTIR’S OPPORTUNITY
The company estimated their TAM for the government and commercial segments combined to be $119 billion. This estimate already excludes those countries where the company will not sell their services such as China. The company also provides in the S-1 how this TAM is separated between both segments:
12.1. Commercial Segment
Estimated TAM for the commercial segment is $56 billion. To arrive at this number the company calculated the number of potential customers around the world (those companies with more than $500 million in annual revenue) and assumed an annual contract value based on existing customer spending. Even though the company is currently focused on big institutions, you have to take into account that TAM is expected to expand going forward as the company moves down to middle-size businesses.
12.2. Government Segment
Estimated TAM for the government segment is $63 billion which is divided as $26 billion domestically and $37 billion internationally. The method used was assuming that 5% of government spend goes to software and consulting services and that Palantir’s product can ultimately achieve more than 5% of this spending going forward.
The company argues that the market opportunity is growing as they develop new functionalities that address a larger spectrum of use cases of their customers.
Although the company is currently focused on big institutions, they are starting to let investors see that they will probably move down to smaller sized businesses with the objective of capturing more market share. One might think, as we mentioned in the Business Model section, that the company’s software is specially suitable for large institutions with vast amounts of dispersed data but the key to being able to offer its products to smaller sized businesses is: modularity.
This said, the value add is greater if their customer is big because, as the data sets increase in size, the gap between data and insights widens. This is exactly where Palantir adds value. A small new company with a small data set won’t probably need Palantir to draw insights yet, but as the company grows and generates more data it may be a potential customer in the future.
The company has invested significant amounts of money to make Foundry a modular platform. This means that customers don’t need to purchase Foundry as a whole, they can just take what they need. This is not only helping with the acquisition of smaller companies, it is also allowing the company to do many more pilots because modular pilots take less time to deploy (something that would have taken 3-4 months can now be done in one week or less) which increases their chances of acquiring customers.
I worked a couple of years ago on a very big project for a big institution here in Spain and I can tell you that Palantir would have solved many of the problems we were trying to solve. The company was founded by the merger of many other smaller companies that had varying data formats so drawing insights from the data was almost impossible. Everything was dispersed and didn’t follow a particular structure so in the project we were trying to give a common shape to this data so management could act on it. Now that I know what Palantir does I think that this problem would’ve been solved with its platforms. Luckily for me, the company didn’t know what Palantir offered so we got paid instead.
You may argue that estimated TAM by the company is overstated because they are basically implying that they will capture almost every customer worldwide. Regardless of any interpretation, the opportunity for the company is huge as they stand at $1.1 billion in revenue.
13.1. Quantitative guidance
In the most recent earnings call, the company gave guidance for the full year and the next quarter although they didn’t give guidance for many metrics.
For the next quarter the company expects revenue of $360 million which represents a YoY growth of 43% and adjusted operating margin of 23% (this adjusted operating margin does not take into account share based compensation expense). For the full year 2021, the company expects Adjusted Free Cash Flow in excess of $150 million.
Palantir also gave a long term guidance of +30% growth until FY 2025 which would mean that the company generates +$4 billion in revenues in 2025. This is a good revenue growth but it seems conservative as it would mean a slowdown if compared with the actual revenue growth rates of the company. I would expect quite higher revenue growth for the coming years.
This is what the company expects but what do analysts expect? In the following table you can see analysts' estimates for the company:
Looking at these numbers I think that they are too conservative for several reasons:
The company’s revenue grew 49% in Q1 and the company guides for revenue growth of 43% for Q2. Seeing this strong growth in the first half of the year and with increased business spend as the pandemic stabilizes I don’t see any reason to think that the company can’t beat this estimate.
The company has directly told investors that they expect revenue growth of +30% up until 2025. In the quarters as a public company they have always provided conservative guidance so I don’t see why this time should be different. The company is expanding contracts with the government and is targeting medium-sized enterprises thanks to the modules of Foundry. Additionally, they have just signed the agreement with IBM which should also help them acquire customers.
13.2. Qualitative guidance
Going forward the company is going to do the following to achieve their desired growth:
Building a direct sales force that helps expand their commercial segment. The company said that it expects to add triple-digit sales headcount this year.
Expand their partnerships across complementary software companies, system integrators and cloud providers in both the commercial and the government segments.
14. WHY PALANTIR
In this section I will outline some reasons why I think that the company may be a good long term investment:
Sticky product - High switching costs
Palantir’s software platforms are sticky. The company makes significant investments during the acquire and expand phases and usually operates these accounts at a loss with the objective of showing potential customers the value that their products can add. Once a customer goes into the scale phase they normally stay a very long time with Palantir as can be seen in the contract duration metric that the company calculates. Contract duration currently stands at 3.7 years for all customers and at 4.6 years for commercial customers.
The majority of the company’s customers can terminate the contracts beforehand with no penalty which speaks even better about the stickiness of the products, customers don’t choose to leave even if they can do so for free. This stickiness will probably give the company pricing power when they have to renew contracts with existing customers going forward. You can also think about it logically: Palantir spends many months trying to figure out the customer’s problems so they can customize the platform so once the platform is deployed it would be very time consuming and costly for the customer to switch to another provider or to develop a proprietary solution.
The bigger the data set, the higher the value add by Palantir
The company offers its customers a data integration platform so as the customer uses these platforms in more verticals, they get more value from them. Once a company has tried the platform in any department they can add value by expanding the platform to other departments or business lines. This facilitates up selling of products to customers and this is in fact what is happening because customers are opening new contracts with Palantir. So basically, a larger and more unstructured data set clearly benefits Palantir.
This can also be seen in the government segment. As Palantir widens its relationships with the government they are signing more and more contracts and already existing contracts are being renewed. Not many companies can say that they have such a profitable relationship with the government. Peter Thiel was asked in an interview about unicorns that worked in national security with the government and he said that he could only think of two: SpaceX and Palantir.
Commercial segment opportunity
The company is still growing fast in the government segment but most of the bull thesis for the company stands in their ability to achieve higher growth in the commercial segment. I think that the company is well positioned to achieve this as they are investing heavily in their direct sales force team as well as pursuing strategic partnerships specially with cloud providers.
On top of this, the modularity of Foundry allows the company to deploy pilots in a much shorter period of time which will reduce the sales cycle and will allow the company to pursue many more commercial opportunities in medium sized companies without needing to deploy a significant number of employees.
Security and privacy focus
As the company has been working for many years with sensible governmental information it has a clear focus on security and privacy which is very important in the current context with many companies investing in cybersecurity. The company said the following:
“From the start we have repeatedly turned down opportunities to sell, collect or mine data.”
I think that trust is something that many potential customers are looking for in a company such as Palantir because many companies gather sensitive information that should not be accessed by any external party. This, believe or not, gives the company a competitive advantage with respect to potential incumbents.
Focused on industry wide solutions
Another thing I like about the company is its industry wide focus. By partnering with key players on key industries the company has the ability to “cross-sell” its platforms to other key players. This works particularly well when the company partners with a key provider in the industry such as Airbus.
As we saw in the “Partnerships” section this can also bring headwinds because the company may limit its selling ability depending on the company with which it signs the partnership.
Apollo gives the company an edge with respect to competition
Apollo is the platform that distinguishes Palantir from other SaaS providers, especially when it comes to working with the government. It allows Palantir’s software to run almost anywhere and this is key for the defense arm of the government.
On top of this, as we discussed earlier, Apollo for Edge AI is widening the company’s TAM and makes it less reliant on other edge providers such as Cloudflare and Fastly.
Companies are increasingly reliant on data analytics
As I mentioned earlier in this writing, even though COVID provided headwinds for the commercial segment, the pandemic is expected to provide significant tailwinds going forward. With the pandemic, many companies understood that digital transformation is not an option, it’s a requirement. Companies that are starting or investing heavily in this transformation process need to have the ability to understand and analyze their data and Palantir aims to be at the core of this process. With almost all of the developed world going “online”, data generation is expected to increase substantially so companies will be increasingly reliant on services offered by firms such as Palantir. It’s almost as if in the future any company will probably need products as the ones that Palantir offers.
Also to note that with a strengthening economy, business spend should increase and the commercial segment of the company should start to gain traction after going through the COVID bump.
Revenue growth estimates by management might be conservative
Management has guided +30% revenue growth from FY2021 to FY2025 which will leave the company with +$4 billion in revenues in 2025. I think that these estimates are somewhat conservative if we take a look at growth rates for the last two years which were in the high 40s. Now that business spending is picking up and the company is focusing on the commercial segment and experiencing high growth in the government segment I don’t think that growth will slow down.
I think that the company can deliver growth in the 40s for the coming years and maintain adequate growth in the following years. Also remember that Palantir’s management has been conservative in their guidance for the past quarters.
This said, Palantir has a “low” number of customers so revenue can be quite volatile as I mentioned earlier. A couple of deals can skyrocket revenue and a couple of terminations can make revenue growth slow down so this is definitely something to take into account.
The company is founder-led and management has a significant stake in the company. I like that management is transparent in what they want the business to become and clearly stating what side they are on which can help the company going forward.
Here I will outline some risks that I see that the company may face in the foreseeable future:
Even though there are some companies that you can consider competitors of Palantir such as Salesforce, C3.AI and Alteryx, the company argues that their own customers are their main competition. They argue that if competitors develop their own internal solutions then they might not need Palantir’s products..
As much as I think this might be true, working as a consultant I can tell you that, even though this might sound as a threat, it is not a real one in my opinion. When a company acquires Palantir you are not only acquiring a product, you are acquiring all the expertise the company has from working with other customers and you are also buying the flexibility to terminate the contract at any given time. These things are not available if you are directly building your platform from scratch. With Palantir you get constant updates and top-tier technology at all times.
The fact Palantir is offering “customized” software gives it a competitive technical advantage over other SaaS companies that are offering generic software products. Going forward, as data generation increases in many companies, generic software products will probably lack the specific requirements for any specific company’s data set so they will have to invest significant amounts of money to customize this software. For Palantir it is much easier to go from customizable software to an intermediate platform that is half customizable and half generic that does not need so much time and human resources to deploy.
The company’s technology is kind of a black box
Not much is known about Palantir’s technology. We know what their products are used for but we don’t know how they were built or what specific technology powers them. Many investors think that this is a problem because the company lacks the necessary transparency that you would expect from a public company. As well as I might agree with this I think that it is understandable when coming from a company that works with highly sensitive governmental data that is used to protect the US from potential threats.
This said, I also think that with the information available, you, as an investor, should be able to assess the product without the need to go into the specifics of the technology. If big institutions are using it and they keep renewing their contracts, they must be finding added value from these products. If you are worried about the moat of the product you can think that Palantir has a head start with respect to other competitors.
Controversy - Reputational risks
Some of the bear thesis goes around the fact that the company may suffer reputational risks going forward. This is IMHO something that you can expect from a company that is working with the government as people will always be against some positions you are taking with the government. Some years ago the media leaked that Palantir may have worked with the government to identify illegal immigrants with the objective of deporting them. Alex Karp addressed these concerns in an interview with CNBC and argued that that company is positioned to help the government enforce the law and even though he respects those that disagree, the company is not going to change. In fact he says that part of the Government’s trust in the company relies on the company being able to stay firm in its beliefs while being unpopular sometimes.
This said, the company has made clear from the start that one of its objectives is helping the government, the same way they provided their products for free to help the UK and the US fight COVID.
The company is expected to enjoy strong growth in the coming years but valuation seems a bit extended at 38x sales. Needless to say that if the company executes well in 5-10 years this valuation may even seem ridiculous compared to the future valuation of the company but the truth is that there is still high execution risk.
This execution risk mainly comes from the fact that management is going to step aside and delegate some of the commercial segment to the new sales team so I would like to see how the new team performs in the next quarters.
Stock dilution - Stock based compensation
As I mentioned in the “Financial Statement Analysis” section the company is incurring significant stock based compensation expenses. Even though this is a non-cash expense it significantly dilutes your ownerships as a shareholder and can affect cash flows in the future if compensation is changed to cash. During the last year the company has significantly increased share count and you can expect at least 35% dilution going forward just with the exercise of stock options.
This is something that every investor should monitor as high SBC are “okay” when the company is growing (reduces the need for debt) but should fall to normal levels as the company matures and turns profitable (both on an accounting and cash basis).
Profitability still not in sight
As we saw in the financial statement analysis section, the company is not profitable at any level on a GAAP basis although it turned profitable on a non-GAAP basis in 2020. In the most recent quarter the company also reported positive CFO and Adjusted FCF but I would like to see if this trend continues in the foreseeable future. I will obviously focus on cash flow generation before “accounting” profitability.
Customer concentration is also an issue as losing one of the big customers could severely affect the company’s results. Although this will remain a risk until the company can diversify more across customers I think that the risk is starting to fade away because the top 3 customers account for less of total revenue as years go by. Customer diversification should increase as the company makes efforts to expand its commercial segment.
Customer concentration can also be seen as positive as it means that existing customers are expanding their relationships with Palantir. Remember that many of the big customers have been with Palantir for more than 5 years.
Long sales cycles with high CAC
As the company has a land and expand model with three phases, the sales cycle tends to be long, from 6 to 9 months up until over a year for some customers. These long sales cycles also bring high costs for the company but the company argues that heavily investing in showing customers what they can get from Palantir’s platforms pays off in the long term.
Even though these sales cycles are currently long, the investments that the company has made in Foundry’s modularity will help shorten these periods, specially pilot deployments, which will lead to shorter sales cycles at a lower cost.
16. PAST PERFORMANCE & VALUATION
Before starting this section I would like to clarify that valuation is not one of the first things I look at when analyzing any company, it’s probably the last. I think that great quality growth companies have the ability to grow very quickly into their valuations. This said, if you like the fundamentals of the company it’s important to study the valuation to see if you can wait for a better entry point. I am increasingly focusing on a balance between fundamentals (90%) and technicals (10%).
Before looking at some valuation ratios for Palantir let’s take a look at their performance since becoming a publicly listed company.
Palantir came public via a direct listing on the 30th of September 2020 at an opening price of $10 per share. The shares closed that day at $9.50, valuing the company at $20 billion. The stock has been very volatile, reaching a high of $45.00 and a low of $8.90 since coming public:
This said let's take a look at some of most relevant ratios for Palantir:
EV / Sales Ratio
The company currently trades at a TTMEV/Sales ratio of 34.2x which is considerably less than the peak a couple of months ago:
The company currently trades at a TTM Price to sales ratio of 35.8x which seems high specially even after the most recent correction that has affected growth stocks:
I will assume that the company is able to grow revenues at 40% in the following two years (2021 and 2022) and I will then use the estimate given by the company which I still think is conservative.
The company is currently trading at 9.1x 2025 expected sales with conservative assumptions which might appear to be high still but I think that the company will still have a considerable runway ahead after 2025.
Simple moving averages
The company is currently trading below the 50 SMA.
The stock is not cheap by any means but I wouldn’t consider it extremely expensive either if you take into account the growth that the company might enjoy in the following years. I do think that execution risk is still high because the commercial segment is just getting started so I would need a lower valuation to consider the stock although I will obviously watch it closely in case it goes down to desired levels.
17. FINAL WORD
I don’t currently hold Palantir but I will watch it closely from now on because I think that the company has a sticky product and a big runway ahead. To capture this potential growth the company has to demonstrate that its new sales team can execute appropriately. Right now I don’t think that the company at these prices can’t substitute any of my holdings so I won't start a position but it will definitely be included on my watchlist in case it lowers to more attractive levels.
During the next few quarters I will be focusing on analyzing how the commercial segment performs and how much of this growth can be attributed to direct sales and how much to sales through partnerships. I would also like to see improvement in profitability, less stock dilution going forward and a continuation in the cash flow generation trend.
If you liked this analysis feel free to share with whoever you think will enjoy it.