Why is Palantir, with a P/E ratio of 173 times, so expensive?

NEXT PICK · Market Insights

Why is Palantir, with a P/E ratio of 173 times, so expensive?

Wednesday, October 14, 2026

On October 14, Palantir closed at $200.97, down 3.69% in a single day. Just a few days ago, it hit an all-time high of $213, and in early August it was just over $125.

A software company with annual revenue guidance of $8.15 billion has a market capitalization of $482.9 billion and a price-to-earnings ratio (P/E) of 173 times. Last quarter's revenue soared 93% year-on-year, with operating margin soaring to 47%, and its fundamentals are almost flawless.

But precisely because there are no flaws, the stock price has already written a perfect script for many years to come. On the day the 10-year U.S. Treasury yield reached 5.21%, there was a $20.9 million short intraday bet. The market is asking: Is this the infrastructure of the AI era, or is it a presale ticket that has been overdrawn for too long?

Latest price $200.97 ▼ -3.69%

Price-to-earnings ratio 173.25 times ▲ This is 20-30 times higher than the technology sector

Q2 revenue growth rate 93% ▲ Revenue was $1.94 billion

From the basement of an intelligence agency, it has grown into an operating system for the AI era

Palantir was founded in 2003 and is now headquartered in Aventura, Florida. It holds four cards: Gotham, which serves the intelligence and defense systems of the United States and its allies; Foundry, the enterprise data operating system; Apollo, the underlying deployment architecture; and AIP, which unifies access to various large models.

In terms of revenue structure, it has long been rooted in the U.S. government, with clients covering the Intelligence Community, Department of Defense, Department of Homeland Security, Department of the Treasury, and Department of Health and Human Services.

But the latest season's changes are especially striking

U.S. government revenue was $809 million, up 90% year-on-year, and U.S. commercial revenue was $764 million, surging 149% year-on-year. For the first time, the commercial side is outpacing the government side**.

Its irreplaceability comes from twenty years of deep coupling with intelligence analysts, Apollo's scarce ability to securely update software in air-isolated networks and sovereign clouds, and AIP's confidence in not having to delegate data to overseas hosting models.

Collaborating with PwC to transform enterprise operations and deploying Nemotron open models in sovereign environments with NVIDIA are both steps toward "replicating government capabilities for business."

If Snowflake is a large warehouse storing data, then Palantir is more like the central nervous system inside the warehouse that directs forklifts, dispatches goods, and makes decisions for people. Research gives it an 8 out of 10 for irreplaceability: The government side is close to a perfect score, the enterprise side is only moderate. The question is, which side is the market paying the price?

A market value of 480 billion—are you buying today or twenty years from now?

Let's look at the contrast first

FY2025 revenue is $4.48 billion, but its market value is $482.9 billion.

Even using the company's revised FY2026 revenue guidance of $8.15 billion, the forward-looking price-to-sales ratio (P/S) is close to 60 times. Based on the past twelve months, the research report estimates it as high as 83.95 times, and the enterprise value multiple (EV/EBITDA) is above 180 times.

The control group was more glaring

The report estimates Snowflake's P/S ratio to be between 15 and 20 times, the average P/E for the tech sector is only 20 to 30 times, and the enterprise value multiplier for the enterprise software sector is 15 to 25 times.

Palantir's pricing logic almost bought out the growth for the next twenty years in one go.

Cash flow itself is textbook-level. FY2025 free cash flow (FCF) reached $2.13 billion, up 85% year-over-year, with an FCF margin as high as 47.5%, surpassing the net profit of $1.63 billion that year.

In the first quarter of 2026, single-quarter FCF was $899 million, and the company has raised its full-year adjusted FCF guidance to $4.5 billion to $4.7 billion.

Yet even so, the free cash flow yield based on FY2025 standards is only 0.44%, and even with full-year guidance fulfilled, it is less than 1%, far below the 5.21% yield of the 10-year Treasury Treasury.

Return on equity (ROE) of 27.67% and PEG of 0.48 may seem cheap, but it's built on a historical net profit growth rate of 396%. Can such growth be extrapolated linearly? **

The report cross-validates using the price-earnings ratio, price-to-sales ratio, and simplified DCF method, and the conclusion is surprisingly consistent: even with an optimistic 60x P/S, the corresponding market value is only at par with the current level.

DCF's results are even lower than current prices. No doubt about self-sustaining ability, but doubtful valuation absorption ability This is exactly what the research report uses as a judgment of "self-sustaining but valuation stretching."

Moving from a period of high investment to a period of large-scale release

The report marks Palantir as in the "breakout period," with a clear logic: revenue growth jumped from about 29% in FY2024 to 93% in Q2 2026, and operating margin expanded from 10.8% in FY2024 to 47.0%.

FY2026 guidance of $8.15 billion represents an 82% year-over-year increase, with next quarter's guidance being 1.23 times the current quarter's actual figure.

Contracts from the Department of Homeland Security, Department of the Treasury, and Department of Health all mark the stage of large-scale expansion, indicating that demand has shifted from single-point trial to organization-wide penetration. The reason it cannot yet be called the "maturity stage" is that valuations are still priced at pure growth stocks, and the income structure has not yet broken free from dependence on the government.

Technically, the stock price is about 11% above the 50-day moving average of $180.36 and 31.8% above the 200-day moving average of $152.48, with the moving averages in a bullish alignment. The market still accepts this breakout narrative. However, the farther from the long-term moving average, the more aggressive any slowdown signal triggers the rebound.

The government faces deep defense, while enterprises attack from multiple angles

In the defense and intelligence battlefield, Lockheed Martin leverages departments like Skunk Works to continuously invest in AI battlefield situational awareness, holding the Department of Defense's budget channels and trust relationships, but its software DNA and iteration speed lag behind Palantir.

Raytheon's strength lies in sub-sectors such as signal intelligence, rather than cross-domain data fusion platforms.

In the federal IT modernization market, IBM leverages its Watson and Red Hat hybrid cloud to hold a significant share in compliance analysis for departments such as the Ministry of Health and Energy. It has deep government channels and compliance experience, but its AI-native capabilities are relatively weak.

The real battles are in the commercial market. Snowflake is a leader in cloud data warehousing, offering flexible pay-as-you-go billing; Databricks' Lakehouse architecture is highly regarded in the data science community.

Both are backed by AWS, Azure, and GCP ecosystems, where customer acquisition efficiency and cost both put real pressure on Foundry.

The report ranks Palantir's moats in order of: strongest switching costs, followed by regulatory barriers, moderately strong economies of scale, moderate network effects, and moderately weak patent technologies.

If intelligence agencies want to change platforms, it means data migration, process restructuring, personnel retraining, and even a temporary vacuum in combat capability. Which national security agency dares to bear such a cost? **

A trillion-yuan track, yet it can't fit an 84-fold price-to-sales ratio

The report estimates that annual US federal IT spending ranges from $1 trillion to $1.1 trillion, with data analytics and AI-related spending of about $120 billion to $165 billion.

Compared to member states, the global government AI analytics platform market (TAM) is conservatively valued at $50 billion to $80 billion, while Palantir's truly serviceable government-side market (SAM) is between $15 billion and $25 billion.

On the enterprise side, the global enterprise AI software market is about $300 billion to $400 billion, with Foundry and AIP targeting medium and large enterprises serving a market of $30 billion to $60 billion.

If the average annual growth rate over the next three years remains at 40% to 60%, and FY2029 revenue is between $22 billion and $35 billion, the P/S ratio for the current market cap is still 13 to 22 times.

If commercial penetration falls short of expectations and growth slows to 20% to 30%, revenue will be about $14 billion, and the P/S ratio will still be as high as 35 times. In other words, even if dreams come true three years from now, today's price will already be ahead of the curve.

The four pillars of the bullish head

Those who are bullish on Palantir hold not just stories, but report cards that exceed expectations:

▲ Bull Case

The bears' five layers of doubt

Bears do not deny that Palantir is a good company; they question: does a good company deserve any price?

▼ Bear Case

The November earnings report is the next touchstone

In the coming months, Palantir's stock price is likely to be dominated by several types of events: whoever pays first will have the say in the next move:

•	Early November Q3 2026 earnings report: The market consensus expects revenue to be around $2 billion and earnings per share of $0.36, with the threshold for exceeding expectations raised, indicating a strong impact
•	In mid-October, a contract dispute settlement with the London police was reached, removing a positive barrier to overseas expansion
•	Goldman Sachs upgraded its rating to buy, Barclays gave its first coverage and overweight, indicating a warm institutional sentiment, but insiders continued to reduce holdings according to the 10b5-1 plan
•	The federal budget for fiscal year 2027 is being implemented, with additional purchases from the Department of Defense, Department of Homeland Security, and Treasury Department, expected to have a high impact in the first quarter of 2027
•	Nemotron's sovereign AI model, developed in partnership with NVIDIA, is being deployed with more government customers, and is expected to have a moderate impact in the fourth quarter
•	Sovereign AI demand in Europe and Asia-Pacific is unleashed, with overseas revenue accounting for over 20%. FY2027 is expected to have a moderate impact

From $7 to $213, valuation expansion outpaces profits

After going public directly in September 2020, Palantir surged from around $10 to above $45, then declined during the tech stock pullback in 2021, dropping to $7–8 by the end of 2022.

At that time, FY2021 and FY2022 were still operating at a loss, and the market was skeptical about the government contract cycle and commercialization prospects.

After ChatGPT ignited the AI boom, it rebounded to the $20 to $25 range between 2023 and mid-2024, with FY2024 revenue of $2.87 billion and net profit of $462 million, marking the start of profitability.

FY2025 revenue grew by 56% and net profit by 253%, with a significant acceleration in value revaluation.

The 2026 scenario is more like a roller coaster: it experienced a sharp sell-off in the first half of the year, closing at $125.65 on August 3; Q2 earnings surged in after-hours hours that night, then soared over 8% in a single day, with a cumulative Q3 gain of about 60%. In October, it hit a record high of $213, and remains up 16% in the past month.

Looking back, the part that was right was due to continued a surge in volume government demand, effective commercialization, and significant improvement in cash flow quality; The part where the expectation was wrong was that the stock price rose far more than the fundamentals, with valuation expansion contributing most of the returns. Can the next round of gains continue to be supported by multiples?

⚠️ Risk Notice • Valuations have been pushed extremely high, and a single "expected" financial report could trigger a double blow in both earnings and valuation • Government revenue is highly concentrated, and federal budget cuts and political cycle changes directly impact revenue • Snowflake and Databricks are competing head-on on the commercial side, which may squeeze pricing power and gross margins • Foundation models accelerate commercialization, and AIP's technological differentiation may be weakened • Export controls and overseas regulatory uncertainties are limiting the pace of international business expansion

🟡 Neutral The fundamentals are impeccable, but valuations have already overdrawn years of growth

💬 Discussion A good company does not equal a good price; be cautious when chasing highs. Post PLTR for real-time buy and sell positions on the official account

Data source

•	Source: NextPick real-time snapshot + company quarterly earnings and guidance + SEC announcement + mainstream financial media and public information from sell-side research reports