Summary
Most skeptics don't get it. I believe Palantir has the strongest narrative in the US equities market today, with spectacular growth results. I believe this is just the beginning.
I expect the AIP platform to increasingly sign more enterprise customers as reliability improves with more efficient LLM models (adopting MoE architectures) and more powerful GPUs, like Nvidia's B300.
The latest shareholder letter shows the CEO’s bold conviction, declaring all the recent events as just the beginning, the first act, of the AI revolution.
I maintain my strong buy rating, considering a longer investment timeframe and the possibility of seeing a pullback in the next weeks, following the recent price action.
This is a follow-up coverage of my December article on Palantir Technologies Inc.
Prior to initiating my coverage on this stock, between November 19 and December 10, 2024, pretty much every article on SA was rating this stock as a strong sell, a sell, or a hold (just look at the first image in my previous article). I guess that's why one should take these ratings with a pinch of salt, including mine, as I've had a few call options (not on Palantir) blow up in my face recently.
In that period of time (November 19 to December 10), no one on SA had given this stock a buy rating, let alone a strong buy. So, when I did and used a bold title for my article, readers had some pretty interesting reactions. This was the title:
Be Greedy When Others Are Fearful.
Needless to say, after reading the comments, very few readers truly understood the meaning of the title. I'll clarify it here.
(Note: if you are not interested in reading my rant, feel free to skip ahead to the earnings section.)
Well, let me put it this way. Look at the share price below as of February 6, 2024. Just by looking at the chart, would you buy at this point?
TradingView
If you said no, that's great! You are a logical agent operating in an efficient market. Ask yourself why not, and you may realize that the word fear may come up among the reasons (i.e. fear of a pullback, fear of a panic selloff, fear of a recession, fear of the AI bubble busting, fear of speculators taking profits, fear of an upcoming short seller report, etc.)
Now, take a look at the chart for December 10, which is when I released my article.
TradingView
Let's assume we go back in time, and you have no information on the future of this stock. Now, I ask you again: would you buy the stock?
Are the same reasons popping up (i.e. fear of ____ you name it)?
Well, this is exactly the fear that I mentioned in the title of my article. To me (a sub-blue collar high-risk speculator) buying when others are fearful means jumping in when the stock price skyrockets, shooting up aggressively. That's when I sweat to open a position in a stock, regardless of how compelling it may be.
I'll be honest. I never understood Buffett's quote. Why? Because people, particularly value and deep value investors, love to buy shares after a selloff, looking for a discount on a compelling stock. Let me put it this way: how many undervalued companies have been growing in the triple digits over the past year? How about in the past 3 months? If the stock is still compelling, but there is no discount on the share price, that's when it takes courage to buy.
From a psychological perspective, people enjoy feeling unique. People like to be contrarians, as this makes them stand out from the masses. People like to be heroes, that's why buying the dip feels better than joining the train of hype. By buying the dip, you go against the masses, separating yourself from the crowd. You become meaningful.
So, massive tangent aside (I have a midlife crisis, baby!... even more reasons not to follow my trades), let's dive deep into the Q4 results.
Q4 2024 Results
You've probably already heard about the Q4 and FY 2025 results. You may have come across a summary somewhere else. I won't summarize here all the results, but only mention what I believe to be the most important developments for the company.
In the past year, Palantir's customer base has increased by 43% yoy. In Q4 alone, 129 deals were closed worth over $1 million, 58 deals exceeded $5 million, and 32 deals were valued at more than $10 million. Here is the exciting part. The TCV (Total Contract Value) at the end of last quarter for US commercial deals was at a record $803 million. To put things into perspective, that's a 134% YoY increase and an impressive 170% QoQ jump.
This is the segment that the market wants to see growing. Look, I have absolutely no doubts that the government segment represents a better use of resources, with use cases that overshadow the importance of commercial applications. However, I believe the market is more excited about commercial deals, even though these deals involve AI agents to automatize the follow-up process during disruptions in the supply chain of a shoe-making company (to put a random example).
Another interesting highlight from the press release was the $5.2 billion in cash & ST investments. This is a 15% QoQ and a 42 YoY increase. The company is accumulating capital from the cash generated from operations. Speaking of which, cash from operations was $460 million for the quarter with a margin of 56%. Let me say it again: for every $1 of revenue, Palantir generated $0.56 in cash from operations.
Finally, Q1 2025 guidance, baby! This is where investors were biting their nails prior to Q4 earnings. The Street was expecting a Q1 revenue guidance of $776 million for Q1 2025. The company has eviscerated (a word they actually used in their investor relations page, more on this in a section below) their expectations, guiding $828 million (midpoint) for Q1 2025. Here are their full year 2025 projections:
Revenue: $3.74B - $3.76B (~31% YoY Growth)
US commercial revenue: > $1.08B (+54% YoY)
Adjusted income from operations: $1.55B - $1.57B
Adjusted FCF: $1.5B - $1.7B
GAAP operating income & net income: positive every quarter
Spectacular growth, to say the least.
What's So Special About Palantir Anyway?
I'll be blunt: the AIP platform is their moat. To my eyes, all the other platforms are glorified data analytics tools (I admit I can be wrong). I believe the true value of Palantir lies in the AIP platform and its programmable AI agents. Why?
Among the entire suite of platforms, I believe the AIP platform will transform how businesses operate, especially in white-collar jobs, where decision-making is a major part of the work.
Look, the Foundry platform is nice for aggregating data from different sources and running a data analytics report (the classical multicolor background glossy report you can find in PowerPoint presentations). However, you still need humans to interpret the data and make the best decisions. With AIP and the use of AI agents, I believe enterprises will be able to cut down a significant amount (not all, though) of white-collar jobs. Even Salesforce is pivoting towards AI agents.
Side tangent: What does the author think about the outlook for blue-collar jobs? I’d say there’s a strong chance that, in a decade or two, we’ll see a hybrid workforce where a significant number of jobs are taken over by humanoid robots (like Tesla’s Optimus) or autonomous vehicles (like Tesla’s Robotaxis). Exciting times for humanity, right? I’m already browsing Zillow listings for a prime spot under a bridge.
On a more serious note, my day job is in engineering consulting within the aerospace sector. While I can't provide here the details of what I do, I can tell you that over 80% of the daily tasks my team and I handle could easily be automated using AI agents. In fact, this week alone, I'd say +90% of my work could’ve been done by a half decent LLM model connected to a large database of documents, drawings, 3D CAD models, spreadsheets, and files. The other 10% involves cracking some jokes with my colleagues and sipping coffee. I encourage you to think about your day job and ask yourself if AI agents connected to absolutely all the data points within (and even outside) your organization could replace you.
I know what some skeptics may be thinking.
Hallucinations, baby! This author forgets that the AI nonsense (as skeptics usually call it) is not reliable. It can't steal my office job! I am still meaningful!
I'll address this argument in the risk section.
Pure Boldness
To show how bold the company is with its execution strategy, they actually used the word eviscerating in the title of the Q4 press release. Previously, in Q3, the company only used this term in the introduction section of their earnings press releases.
Another bold statement can be found in the latest shareholder letter, where CEO Alex Karp begins with the following statement:
We are still in the earliest stages, the beginning of the first act, of a revolution that will play out over years and decades.
Furthermore, this is how the CEO concludes the shareholder letter:
As Samuel Huntington has written, the rise of the West was not made possible “by the superiority of its ideas or values or religion . . . but rather by its superiority in applying organized violence.” He continued: “Westerners often forget this fact; non-Westerners never do.”
I’m not making this up, those were the CEO's exact words. If that's not a bold letter, then I don’t know what is!
Risks
Look, I agree with the skeptics—the issue of AI hallucinations is keeping many enterprises on the sidelines when it comes to integrating AI into their operations. That's why AI agents are not in the daily news today. That's why we are not seeing unions and governments fighting with corporations laying off a significant amount of their office employees. While I agree that today's LLMs are not 100% reliable (that's why I believe authors on SA still exist, by the way), just wait until the main LLMs in the West implement the Chinese MoE (model of experts) architecture and enterprises train these highly efficient models in ultra-high performance GPUs, like the upcoming Blackwell B300. The speed of innovation in the AI space will catch many people (mainly the skeptics) off guard.
Additionally, Palantir is already addressing hallucinations in AI models through their ontology-based approach. This is a structured representation of an organization's data, workflows, and decision-making processes. Without this structure, AI can hallucinate as these LLM models lack inherent knowledge about an organization’s specific processes, rules, and constraints. Additionally, ontology allows organizations to trace AI decisions back to structured data. This is key in order to fine tune the model during the setup process.
To make it nice and simple, imagine a ChatGPT trained specifically with all the available data from your own organization that is capable of performing tasks like filling spreadsheets, writing reports, writing code, preparing presentations, finding inefficiencies, following up with customers, reaching out to new customers, do a root cause analysis, you name it.
So, to conclude, yes, the fear or AI hallucinations may be the main reason we are not seeing big corporations adopting Palantir's AIP platform. However, I believe this will change in the near term, as these models become more powerful.
Conclusion
To wrap up, I believe my bull case still holds up. That doesn't mean one should jump into this stock now expecting short-term gains. Pullbacks are possible, and the investment timeframe I aim for is well beyond one year.
The company is aggressively expanding its commercial customer base, with US commercial revenue expected to exceed $1.08 billion in 2025, a 54% YoY increase. I believe the AIP platform will soon become the company's main source of revenue (by far), offering programmable AI agents that could revolutionize business operations. Even though the company says their AI agents are meant to support, not to replace humans, I argue this is PR jargon due to ethical implications. In my view, if an enterprise (in the US) identifies a strong business case to replace (expensive) human workforce with an AI agent that can work 24/7/365, I have a high conviction that they will, regardless of how ethical or not this may be.
Looking at the latest letter to the shareholders, the CEO's confidence in leading the way in AI for enterprise solutions is evident.
From a risk perspective, I do believe there are a large number of enterprises that remain hesitant to fully integrate AI agents due to reliability issues, mainly hallucinations. Additionally, despite strong performance and, what I believe to be the strongest narrative on the US equities market today, many investors still perceive Palantir as an overhyped AI play. As this perception changes in the next years, I believe the share price will appreciate in the long run.
I maintain my strong buy rating with a long-term investment timeframe. Will there be a pullback in the next weeks? Probably, but over the next 24 months, I believe this will be just noise.