Shares of big data platform Palantir (PLTR -2.38%) stunned investors on Monday, blowing away analyst revenue and profit targets on its fourth-quarter earnings report.
In response, Palantir’s stock rallied another 24% on Tuesday, surpassing $100 per share and reaching a market cap around $236.5 billion. The stock is now up over 500% over the past year, cementing its standing as one of the biggest winners of the AI revolution.
The stock also looks wildly expensive at this moment. Still, Palantir’s stock has always looked expensive, yet that hasn’t stopped it from defying skeptics again and again.
Could the stock quadruple again to reach a $1 trillion valuation, putting it in league with the Magnificent Seven? Here’s the case to be made.
Palantir’s valuation looks insane
Palantir no doubt delivered some impressive results in the fourth quarter and throughout 2024. In Q4, overall revenue grew 36%, with U.S. revenue up 52%, including U.S. commercial growth of 64% and U.S. government growth of 45%.
The company also closed a record-high new quarterly contract value, at $803 million. While GAAP operating income came in at just $11 million or 1%, that was heavily affected by “one-time” stock appreciation rights, or SAR, for employees. On an adjusted non-GAAP basis, operating earnings were $373 million, good for a 45% operating margin. For the full year, Palantir grew 29%, with an adjusted operating income margin of 39%.
Those results marked a big acceleration. Last year’s fourth quarter saw only 20% growth, with full-year 2023 revenue growth of 17% and a 28% adjusted operating margin.
While accelerating revenue and profitability is awesome, investors may get nosebleeds when looking at Palantir’s valuation. Shares currently go for 505 times 2024 GAAP earnings. And even when using “adjusted” net income of $1 billion, which adds back stock-based comp that is a real expense to shareholders, shares still go for about 236 times earnings.
When one looks ahead to 2025 guidance, Palantir still trades at 63 times 2025 sales guidance, even after 31% projected revenue growth, and 148 times adjusted free cash flow. Notably, management didn’t guide for a GAAP net income figure, which would have been much lower than free cash flow due to high stock-based comp.
There is still a bull case to be made, though
While Palantir’s valuation is currently at nosebleed levels, there is a case to be made for the stock, believe it or not. That case rests on Palantir’s becoming a near-monopolistic platform for the enterprise in the age of artificial intelligence. Wedbush sell-side analyst Dan Ives is a bull, claiming Palantir’s dominance of this new area could be akin to Oracle‘s dominance in enterprise databases or Salesforce‘s dominance of customer relationship management platforms.
While Palantir isn’t a builder of large language models, its software platform, which has been many years in the making, allows large language models (LLMs) to be harnessed in a useful and secure way for enterprises today.
That appears to be a rare capability. Palantir unveiled its AIP platform in the second quarter of 2023, which brings LLMs into Palantir’s core platforms. Since the introduction of AIP, growth has reaccelerated:
PLTR Operating Revenue (Quarterly YoY Growth) data by YCharts
Why Palantir’s platform gets such a lift from AI
Palantir’s platform isn’t traditional enterprise software. Rather, it composes what management calls an enterprise “ontology.” As Palantir defines it, an ontology is essentially a “digital twin” of an enterprise’s architecture, which all elements within an organization are reduced to a digital object with known properties, specified links between them, and which undergo certain defined actions.
Thus, if one wants to “automate” one’s enterprise through the use of AI, that AI will have to interface digitally with every database, factory, HR profile, and other objects in the enterprise’s universe. And it also has to do so in a secure and compliant way.
Palantir executive Shyam Sankar said on the conference call with analysts said if you only wanted to automate a single business, you wouldn’t need an ontology platform. Moreover, if one were to build traditional software for all kinds of businesses, it wouldn’t work in this AI context, as businesses aren’t one-size-fits-all.
However, Palantir’s ontology framework, some 20 years in the making, enables a sort of mass-customization across players in various industries that makes enterprises easier to “automate” with AI. And because Palantir began with the Department of Defense for anti-terror activities, businesses also trust Palantir to manage a company’s data in a highly secure way. Sankar elaborated:
If you have the ambition to build software that works across government, across 50 different industries, you have to build Ontology. It is absolutely the longest path between Point A and Point B. But it gives you superpowers when you show up to your customers.
Dominant enterprise platforms tend to do well
It still remains to be seen if Palantir will become a monopolistic de facto platform for enterprise AI, but its accelerating revenue growth, even with a minimal sales force, certainly paints an optimistic picture. Whether one is talking about, say, an Oracle database, a Salesforce CRM platform, or a Microsoft Office suite, once a software platform achieves dominant status, it tends to get beneficial network effects.
This happens because once a platform gets a big lead and becomes standardized, everyone tends to learn how to operate that platform, ingraining it even further. You don’t want to have to learn three, four, or five different enterprise software platforms in school or at any one business, after all.
That’s why enterprise software platforms tend to be wide-moat, cash-generative businesses, and it’s the reason Palantir may eventually reach $1 trillion over time — even though it looks very far away from achieving that milestone today.
