Palantir Technologies U.S. Deep Research Report: The Clash Between High-Quality Growth and Extreme Valuation

Data as of the U.S. market close on 2026-08-04. This article is compiled based on the company's SEC disclosures, company platform documentation, and public market data; market price, enterprise value, diluted share count, and trailing earnings conventions in the valuation may differ due to different data vendors' update times.

Original research report: Download PDF version

Executive Summary

As of the 2026-08-04 close, Palantir Technologies（NASDAQ: PLTR）traded at 162.66 U.S. dollars, up about 29.45% on the day, on volume of about 175 million shares. The direct trigger for the re-rating was second-quarter results: revenue grew 93% year over year to 1.935 billion U.S. dollars, U.S. commercial revenue grew 149% year over year, U.S. government revenue grew 90% year over year, and the company also raised its 2026 revenue guidance to 8.15 billion to 8.158 billion U.S. dollars.

Palantir's operating quality is at an extremely high level for a large software company: growth, GAAP profitability, cash flow, and contract backlog are all accelerating at the same time. The company's core platform stack consists of Gotham, Foundry, AIP, Apollo, and Ontology, which together cover data integration, model access, permission governance, business processes, and cross-environment deployment.

Palantir's core contradiction is that operating performance and valuation are both in extreme territory. Using about 2.4 billion basic shares as an estimate, current equity value is around 390 billion U.S. dollars; different data vendors put market capitalization at roughly 375 billion to 418 billion U.S. dollars. Using trailing-twelve-month revenue of about 6.1 billion to 6.2 billion U.S. dollars, the price-to-sales ratio is about 63x; public data vendors show historical price-to-earnings ratios of about 153x to 183x. This price requires the company to sustain growth and margins far above traditional software peers for many years, and any slowdown in growth could trigger a meaningful valuation compression.

This report uses the end of 2027 as the target date. The base case expects 2027 revenue of about 12.07 billion U.S. dollars, GAAP net margin of 43%, diluted earnings per share of about 1.98 U.S. dollars, and assigns an 80x P/E ratio, implying about 158 U.S. dollars; the bearish case is about 87 U.S. dollars, and the bullish case is about 253 U.S. dollars. Under an illustrative 25%/50%/25% probability weighting, the result is about 164 U.S. dollars, close to the current price, indicating that the market has already priced in fairly strong execution and that the margin of safety is limited.

Company, Products, and Industry Positioning

Palantir was founded in 2003 and initially provided counterterrorism and data analytics software to U.S. intelligence agencies before expanding into commercial fields such as defense, public health, manufacturing, energy, finance, automotive, aerospace, and supply chains. The company went public via direct listing on the New York Stock Exchange in September 2020 and moved its Class A common stock to trading on the Nasdaq Global Select Market in November 2024.

Palantir describes Ontology as the center of its architecture. Ontology represents objects, relationships, decisions, and actions within an enterprise, allowing AI agents to call real data, execute workflows, and feed results back into operating systems under the constraints of permissions and business rules. AIP is responsible for connecting models with business processes, Foundry provides the data, applications, and governance environment, and Apollo manages deployment and updates across cloud, edge, and classified environments.

This business model has relatively high initial deployment complexity. Palantir typically requires engineers and customers to work together to map out data, permissions, and workflows, resulting in higher upfront costs and longer sales cycles. But once customers complete Ontology and key application deployments, switching costs rise as data models, process flows, permission rules, and user dependencies increase. Deployment and operations teams also continue to identify new use cases, driving expansion within the same account.

Industry Environment and Growth Drivers

Palantir operates simultaneously in enterprise AI platforms, data integration, decision intelligence, defense software, cloud software, and systems integration markets. Because market boundaries overlap, no authoritative institution publishes a directly verifiable global market share for Palantir. A ranking for a single data science or machine learning platform cannot fully capture its position in field operations, classified deployments, supply chains, and defense command systems.

AI demand remains strong. Stanford AI Index 2026 shows that AI adoption among surveyed organizations reached 88%, and about 70% of organizations had adopted generative AI in at least one business function; actual deployment rates of AI agents across most functions remain in the single digits, indicating that enterprise AI is still at an early stage in the transition from experimentation to large-scale operations. Palantir's opportunity lies not only in model access, but also in governance, permissions, data sovereignty, model orchestration, and production environment execution.

Government demand also has structural support. SIPRI estimates that global military spending in 2025 reached 2.887 trillion U.S. dollars, with European military spending up about 14% year over year and Asia and Oceania up about 8.1%. Global military spending cannot be directly treated as Palantir's addressable market, but defense digitization, unmanned systems, intelligence fusion, missile defense, and supply chain resilience can expand software's share of defense budgets.

The most important current growth driver comes from the U.S. commercial market. AIP Bootcamp, forward-deployed engineers, prebuilt industry solutions, and Ontology shorten the cycle from trial to production. In Q2 2026, U.S. commercial revenue reached 764 million U.S. dollars, up 149% year over year; U.S. commercial total contract value reached 2.132 billion U.S. dollars, up 153% year over year; and U.S. commercial remaining deal value reached 6.238 billion U.S. dollars, up 124% year over year. Together, these indicators show that new bookings and future revenue backlog are expanding.

U.S. government business is also clearly accelerating. Q2 U.S. government revenue grew 90% year over year to 809 million U.S. dollars. Geopolitical tensions, missile defense, battlefield AI, and government efficiency projects support contract demand, but government contracts remain exposed to budget delays, administrative transitions, bid protests, audits, and termination-for-convenience clauses.

Another long-term driver is "sovereign AI." Enterprises and governments want control over training data, inference logs, access permissions, and deployment locations while retaining the ability to switch among multiple open-source and commercial models. Palantir allows customers to connect models to local or controlled cloud environments, which is attractive in highly regulated industries, but this advantage also depends on whether customers are willing to accept deep system embedding and relatively high contract pricing.

Financial Performance, Revenue Mix, and Balance Sheet

The table below is in billions of U.S. dollars, except per-share data. Free cash flow uses the simplified convention of "operating cash flow minus capital expenditures," which differs from the company's adjusted free cash flow definition.

Revenue CAGR from 2021 to 2025 was about 30.5%, and the growth curve re-accelerated after bottoming in 2023. Profitability improvements came from revenue growth outpacing headcount, cloud infrastructure, and sales expense growth. In Q2 2026, GAAP operating profit was about 912 million U.S. dollars, with an operating margin of about 47%; adjusted operating margin was about 62%; GAAP net profit was about 1.062 billion U.S. dollars, with net margin of about 55%; operating cash flow and adjusted free cash flow were both about 1.22 billion U.S. dollars.

Quarterly net margin is higher than operating margin partly because of interest income generated by the large cash and Treasury balance and certain tax items. Long-term valuation should focus more on the sustainability of operating margin and earnings per share from operations rather than simply extrapolating quarterly net margin.

The revenue mix is tilting toward the commercial business, while the geographic mix is becoming more concentrated in the U.S. In Q2 2026, U.S. commercial and U.S. government revenue together totaled 1.573 billion U.S. dollars, accounting for about 81.3% of total revenue, while international revenue accounted for about 18.7%. Faster growth in the U.S. market improves near-term visibility, but also increases the company's sensitivity to U.S. government policy, budget cycles, and domestic AI capital spending shifts.

The company does not separately disclose revenue for AIP, Foundry, Gotham, or Apollo, so it is not possible to quantify AIP's standalone revenue, renewal rate, or gross margin. Investors can only judge AIP commercialization speed through U.S. commercial revenue, contract value, customer expansion, remaining deal value, and management case studies.

At the end of June 2026, the company held about 2.03 billion U.S. dollars in cash and cash equivalents and 7.38 billion U.S. dollars in marketable securities, for total cash and short-term U.S. Treasuries of about 9.2 billion U.S. dollars; current assets were about 11.1 billion U.S. dollars, current liabilities about 1.536 billion U.S. dollars, shareholders' equity about 9.774 billion U.S. dollars, and total liabilities about 1.794 billion U.S. dollars.

Stock-based compensation expense fell from 778 million U.S. dollars in 2021 to 684 million U.S. dollars in 2025, with the absolute amount stabilizing and its share of revenue declining steadily. However, total common shares at the end of 2025 were about 2.391 billion, up from about 2.339 billion at the end of 2024, an annual increase of about 2.2%. Investors should monitor SBC expense, diluted share count, and revenue per employee together; looking only at adjusted profit can overstate the economic return attributable to common shareholders.

Share Price History and Valuation

Palantir's month-end price rose from about 9.50 U.S. dollars in September 2020 to 35.18 U.S. dollars in January 2021, then fell to 6.42 U.S. dollars in December 2022 during rising rates, software valuation compression, and growth slowdown, a peak-to-trough decline of about 82%. After achieving GAAP profitability in 2023, the stock recovered, rising to 75.63 U.S. dollars at the end of 2024 and reaching 200.47 U.S. dollars at the end of October 2025.

From the end of September 2020 to 2026-08-04, the price increased by about 16.1x, equivalent to cumulative gains of about 1,612% and an annualized growth rate of about 62%. After October 2025, the share price fell to a low of about 106.37 U.S. dollars in June 2026 before rebounding to 162.66 U.S. dollars following second-quarter results. The past 52-week range has been about 106.37 to 207.52 U.S. dollars; the current price is still about 22% below the prior high, but well above the 2022 low.

August 4 volume was about 175 million shares, several times the previously common daily volume of 30 million to 50 million shares. Heavy volume may reflect institutional rebalancing, short covering, options hedging, and momentum funds at the same time; volume alone cannot distinguish buy from sell direction. Technically, this formed a high-volume breakout. If volume quickly fades and the price cannot hold the earnings gap area, the market may retest the 125 to 145 U.S. dollar range; if the company keeps raising guidance, the 207.52 U.S. dollar prior high will become the main price reference. This is only a technical scenario analysis and cannot replace fundamental valuation.

Palantir's premium is partly supported by fundamentals: revenue growth, adjusted operating margin, free cash flow margin, and the balance sheet are clearly better than those of most high-growth software peers. But the current valuation still requires that advantage to be sustained for many years. Even if 2026 revenue reaches $8.154 billion, based on a market cap of about $390 billion, the 2026 forward price-to-sales ratio would still be about 48x.

Forecast Scenarios and Sensitivity Analysis

The company's latest 2026 guidance is as follows:

The 2026 revenue midpoint of $8.154 billion represents about 82.2% growth from $4.475 billion in 2025. First-half revenue has already reached $3.568 billion, and the Q3 guidance midpoint is $2.162 billion, so the company would only need about $2.424 billion in Q4 to hit the full-year midpoint, implying quarter-over-quarter growth of about 12%. Under current contract momentum, the guidance appears somewhat achievable, but the company will face an increasingly difficult year-over-year comp.

The bear case assumes AIP growth slows, international markets remain constrained, and U.S. commercial growth falls back to around 40%, with the market de-rating the stock to a 55x P/E that is still above mature software companies. The base case assumes U.S. commercial continues to expand rapidly, government business keeps growing at 20% to 40%+, and international business gradually recovers; revenue growth slows from about 82% in 2026 to 48% in 2027, while net margin rises to 43%. The bull case assumes Palantir becomes the mainstream operational AI platform for governments and large enterprises, the pace of AIP agents entering production is meaningfully faster than the industry, and international sovereign AI projects offset political headwinds.

Using 2.62 billion diluted shares in 2027 and an 80x P/E:

At $162.66, the current price roughly corresponds to: about 108x P/E if 2027 EPS is $1.50; about 81x if EPS is $2.00; about 65x if EPS is $2.50; and about 54x if EPS is $3.00. This means the company needs to gradually "work off" the valuation through EPS growth. If 2027 EPS is close to $2, the current price is already equivalent to about 81x 2027 earnings.

Competitive Landscape and Market Position

Palantir does not have a single direct competitor; instead, it competes across multiple layers:

• Data and AI platforms: Databricks, Snowflake, Google Cloud, Microsoft Azure, AWS, IBM watsonx.
• Systems integrators and government contractors: Accenture, Booz Allen, Leidos, SAIC.
• Defense technology companies: Anduril, Shield AI, and others.
• Customer in-house development: in-house data stacks, AI agents, and operating systems.

Palantir's differentiation centers on Ontology, access control, complex operational workflows, and deployment in edge and classified environments. In the commercial market, Palantir both competes and cooperates with Snowflake, Databricks, and cloud platforms: customers may use Snowflake or Databricks to store and process data, while Palantir builds the Ontology, applications, and decision workflows. If data platforms move upward into AI agents, governance, and the application layer, they could shrink Palantir's differentiation space; if Palantir controls the customer's operational semantic layer, it can commoditize the underlying data warehouse and model vendors.

There is currently no verifiable global market share data, for several reasons: Palantir spans multiple markets, government contracts are often multi-year framework agreements or task orders, some key competitors are private companies, cloud vendors embed AI capabilities into broader cloud revenue, and Palantir does not disclose product-level revenue. More useful proxy indicators are 2025 revenue of $4.475 billion, 2026 revenue guidance of about $8.154 billion, U.S. commercial remaining deal value of $6.238 billion, 220 deals of at least $1 million in Q2, and U.S. revenue accounting for more than 80%. These numbers show the company has strong growth momentum in U.S. large-institution operational AI, but they do not prove it already dominates the entire enterprise AI platform market.

Regulatory, Privacy, and Governance Risks

The EU AI Act took effect in August 2024 and enters broad applicability and enforcement from August 2, 2026. Corporate fines for certain transparency rules can be as high as 15 million euros or 3% of global annual revenue. Depending on the specific project, Palantir may act as a platform provider, AI system provider, deployer, processor, or contractor, which determines the scope of its specific responsibilities.

In its 10-K, the company explicitly lists risks related to GDPR, the EU AI Act, NIS2, data localization, cybersecurity, and cross-border data transfers. Violations could lead to fines, lawsuits, regulatory investigations, contractual indemnities, product modifications, and the suspension of certain business practices.

Palantir holds the NHS England federated data platform contract, with a maximum value of about £330 million. U.K. parliamentary members and some healthcare organizations have raised concerns about the vendor's role, access to health data, military business, and public trust. Such disputes could affect contract renewals, international healthcare business, and government procurement.

Defense, immigration, and conflict-related business also bring reputational risk. The software is used in military, intelligence, and immigration enforcement settings, placing the company under continued scrutiny from human rights groups, employees, customers, and governments. These contracts can boost revenue and technological barriers, but they may also limit acceptance in European public-sector institutions, universities, healthcare, and values-sensitive enterprises.

Palantir uses an A, B, and F multi-class share structure, and the F shares allow the founders to retain significant voting control under certain conditions. Ordinary A shareholders have limited influence over the board, M&A, compensation, and corporate strategy, and this governance discount is likely to persist over the long term.

Catalysts and Downside Risks

Investment Conclusion, Position Sizing, and Holding Framework

Palantir has a rare combination: rapid revenue growth, gross margins above 80%, GAAP operating margins near or above 40%, low capital expenditure, strong free cash flow, and a net cash balance sheet. Ontology integrates data, models, permissions, and business actions, while Apollo supports deployment across cloud, edge, and classified environments, giving the company strong switching costs in complex and regulated settings.

The U.S. commercial business has already evolved from a secondary growth source into the main engine, and second-quarter growth of 149% plus $6.238 billion in remaining deal value support strong growth over the next several quarters. U.S. government business also grew 90%, showing that commercial expansion has not come at the expense of its traditional strengths.

But the current share price already embeds many years of high execution assumptions. About 60x trailing revenue, about 150 to 180x historical earnings, and about 80x base-case 2027 earnings make returns highly sensitive to the terminal multiple. The company can continue to outperform, and the stock can still experience large pullbacks even while operations keep improving.

For investors who do not already own the stock, waiting for valuation to come down, for growth to be further validated, or entering in small tranches offers a more favorable risk/reward profile.

A more disciplined approach is to split the target position into three to four tranches and add based on earnings delivery and valuation, rather than on price appreciation alone. Later additions can require U.S. commercial growth, GAAP operating margin, SBC ratio, and international revenue to at least meet the base case. If a single stock rises to more than 5% of the portfolio, consider rebalancing to avoid valuation swings dominating overall portfolio risk.

Position Monitoring Checklist

The final judgment is that Palantir is likely to remain an important winner in operational enterprise AI and defense software, but the certainty implied by the current price is lower than the obvious appeal of the company's operating quality. For a typical diversified portfolio, an initial position of 1% to 2%, a normal cap of about 3%, and a holding period of 3 to 5 years are appropriate; any increase in position size should be based on both earnings delivery and valuation compression.

Main Sources

• Palantir 2026 年第二季度 8-K 财务附件
• Palantir 2025 Form 10-K
• Palantir 平台架构文档
• Palantir 2026 年第一季度 Form 10-Q
• Stanford AI Index 2026
• SIPRI Trends in World Military Expenditure 2025
• Reuters: Palantir raises annual revenue forecast
• EU AI Act regulatory framework
• Palantir price history
• Google Finance: PLTR

Risk reminder: This article is for information organization and research discussion only and does not constitute any investment, tax, or legal advice. Stocks may fluctuate significantly, and historical performance and scenario valuation do not represent future results. Investors should make independent decisions based on their own risk tolerance.