Palantir

More Sizzle than Steak

First – disclaimer

TLDR – This feels like a hyped meme stock, so I’m staying away. Insider selling is massive.

Palantir’s stock has been on a tear recently. It’s up from $17 a year ago to $80 today. The last $40 of the move has come since their November earnings release, which happened to coincide with Trump’s reelection. While the earnings were certainly strong, they were not materially better than the previous quarter, so why has the stock doubled since then?

My sense is that the market now thinks revenue growth can be sustained at 25-30% for the next five to ten years and that margins will continue to get better. I think there’s also a view that the new administration will result in more government contracts for Palantir.

To justify today’s valuation of $180B, you need to believe revenue will grow at 25% for the next 5 years and that operating margins will improve from about 20% to 45%. This would put Palantir’s operating margins at the same level as Microsoft’s. In addition, you’d have to assign a 50x multiple to those earnings five years out. While revenue and margins could certainly get to these levels, a lot must go right.

I’ve built out a simple model here. Feel free to use your own assumptions for revenue growth, margins, and a PE multiple five years out.

Palantir is not a simple business to understand. Here is what they do (from their latest annual report) -

We build software that empowers organizations to effectively integrate their data, decisions, and operations at scale.

We were founded in 2003 and started building software for the intelligence community in the United States to assist in counterterrorism investigations and operations. We later began working with commercial enterprises, who often faced fundamentally similar challenges in working with data.

We have built four principal software platforms, Palantir Gotham (“Gotham”), Palantir Foundry (“Foundry”), Palantir Apollo (“Apollo”), and Palantir Artificial Intelligence Platform (“AIP”).

Gotham and Foundry enable institutions to transform massive amounts of information into an integrated data asset that reflects their operations, and AIP leverages the power of our existing machine learning technologies alongside large language models (“LLMs”) directly within Gotham and/or Foundry to help connect AI to enterprise data. For over a decade, Gotham has surfaced insights for global defense agencies, the intelligence community, disaster relief organizations and beyond. And Foundry is becoming a central operating system not only for individual institutions but also for entire industries.

Apollo, which we began offering as a commercial solution in 2021, is a cloud-agnostic, single control layer that coordinates ongoing delivery of new features, security updates, and platform configurations, helping to ensure the continuous operation of critical systems. Apollo allows our customers to run their software in virtually any environment.

In 2023, we began deploying our newest offering, AIP, which is designed for customers across the commercial and government sectors, enabling them to derive value from recent breakthroughs in artificial intelligence via the combination of our existing software platforms with LLMs. We believe AIP uniquely allows users to connect LLMs and other AI with their data and operations to facilitate decision-making within the legal, ethical, and security constraints that they require.

There’s a narrative that Palantir has no competitors. I can’t profess to be an expert in this space, but based on the digging I’ve done over the last week it seems that both Snowflake and Databricks are real competitors. Databricks has the same revenue as Palantir, but is growing at 60% YoY. See recent press release. They also recently raised $10B at a $62B valuation. It’s hard to see why Databricks is worth just one third of Palantir given its growing revenue twice as fast.

Databricks is private, so it could be argued this comparison is irrelevant. Snowflake, however, is public and their revenue is growing at the same rate as Palantir from a bigger base ($4B vs $3B). Snowflake’s NRR is 127% vs Palantir’s 118%. In spite of this, Snowflake’s market cap is about 1/3^rd of Palantir’s at $58B.

I don’t believe in valuing companies relative to each other. My goal is to buy good businesses at a fair price, but it stands out when a company in a sector is trading at 3x the price of similar firms. Palantir is without doubt a great company, but the stock is way too rich for my tastes. If it drops to a valuation closer to $60B, I may be a buyer.

I’ll be looking at Snowflake in detail next.

Some other thoughts –

1. Palantir spends $400m on R&D each year vs Snowflake’s $1.6B. Maybe Palantir’s engineers are just that much better, but I struggle to believe that

2. In the last year, Palantir insiders have sold $4.2B in stock. Yes, the market cap expansion has been massive, but this number really stands out. Below is a list of insider sales at other companies over the last year along with rough market caps.

a. SNOW - $270M ($60B market cap)

b. NOW - $105M (200B market cap)

c. TSLA - $286M ($1.3T market cap)

d. GOOG - $49M (2T market cap)

3. Snowflake and Databricks were founded almost ten years after Palantir, but do about the same revenue. This would indicate that Palantir’s products and management team are likely not exceptional enough to justify a 3x premium to similar businesses. If anything, Snowflake and Databricks should trade at a premium to Palantir.