来自 Maksym Misichenko · Nasdaq ·
AI智能体对这条新闻的看法
The panelists agree that both Oracle and Palantir face significant risks, with Oracle's heavy capex and Palantir's high valuation being the primary concerns. They differ in their assessment of which company is better positioned to navigate these challenges.
风险: Oracle's heavy capex and Palantir's high valuation
机会: Palantir's software-first business model and high operating margins
本分析由 StockScreener 管道生成——四个领先的 LLM（Claude、GPT、Gemini、Grok）接收相同的提示，并内置反幻觉防护。 阅读方法论 →
Palantir 增长更快，但其估值几乎没有犯错的余地。
Oracle 庞大的云端积压订单为投资者提供了对未来人工智能需求的强有力洞察力。
Oracle 存在执行风险，但其较低的估值可能提供更好的风险回报率。
Palantir Technologies (纳斯达克：PLTR) 和 Oracle (纽约证券交易所：ORCL) 均受益于人工智能 (AI) 热潮。
Palantir 增长速度更快，并且正在迅速采用其人工智能平台 (AIP)。 Oracle 的增长速度较慢，但其巨大的云端积压订单和积极的人工智能基础设施建设为投资者提供了另一种类型的人工智能机会。
人工智能会创造世界上第一个万亿富翁吗？ 我们的团队刚刚发布了一份报告，内容是关于一家鲜为人知但提供英伟达和英特尔都需要的关键技术公司，被称为“不可或缺的垄断”。 继续 »
Palantir 的收入在 2026 年第一季度增长了 85% 年比率，达到 16 亿美元，而美国的商业收入增长了 133%，达到 5.95 亿美元。 该公司的调整后运营利润率为 60%，调整后自由现金流利润率为 57%。 Palantir 还记录了 150% 的净美元留存率，突出了其在扩大现有客户支出方面的成功。 AIP 越来越多地被用于核心业务运营，这使得 Palantir 的增长看起来比短期人工智能炒作周期更持久。
但该公司近 153.9 倍的市盈率和 62.7 倍的市销率几乎没有犯任何执行错误的余地。
Oracle 的增长速度较慢，但其估值也更合理，约为 34.7 倍的市盈率和 8.7 倍的市销率。 分析师预计该公司的收入将同比增长约 20%，在 2026 财年的第四季度达到 191 亿美元。 在 2026 财年第三季度，收入同比增长 21.7%，而调整后运营利润率为 43%。 剩余的履约义务也增长了 325%，达到 5530 亿美元，为公司提供了对未来人工智能云端需求的异常强烈的洞察力。
风险在于 Oracle 的人工智能积压订单仍需要转换为有效的数据中心容量。 该公司估计 2026 财年的资本支出约为 500 亿美元。 Oracle 还宣布计划在 2026 年筹集 450 亿美元至 500 亿美元的资金，以资助数据中心扩张。 然而，这种人工智能数据中心支出正在对现金造成重大压力。 Oracle 的机会虽然巨大，但取决于资金、建设速度、电力供应、成本控制以及客户快速变化的人工智能计算需求。
Palantir 看起来像是一家更强大的公司，但 Oracle 似乎提供了更好的风险回报前景。
在您购买 Palantir Technologies 的股票之前，请考虑以下事项：
Motley Fool Stock Advisor 分析师团队刚刚确定他们认为投资者现在应该购买的 10 只最佳股票……而 Palantir Technologies 并非其中之一。 制作这份名单的 10 只股票在未来几年可能会产生巨大的回报。
考虑一下 Netflix 在 2004 年 12 月 17 日被列入这份名单时的情况……如果您当时投资了 1,000 美元，您将拥有 471,072 美元！ 或者考虑一下 Nvidia 在 2005 年 4 月 15 日被列入这份名单时的情况……如果您当时投资了 1,000 美元，您将拥有 1,303,352 美元！
现在，值得注意的是 Stock Advisor 的总平均回报率为 983%——与标准普尔 500 指数相比，这是一个市场领先的超额表现，标准普尔 500 指数增长了 210%。 不要错过最新的前 10 名名单，该名单可与 Stock Advisor 一起使用，并加入由个人投资者为个人投资者建立的投资社区。
**Stock Advisor 的回报率截至 2026 年 5 月 28 日。 *
Manali Pradhan, CFA 对上述任何股票都没有持仓。 Motley Fool 持有并推荐 Oracle 和 Palantir Technologies。 Motley Fool 有一份披露政策。
本文中的观点和意见是作者的观点和意见，不一定代表 Nasdaq, Inc. 的观点。
四大领先AI模型讨论这篇文章
"Oracle's massive capex and funding needs introduce dilution and execution risks that its 8.7x sales valuation fails to reflect."
The article correctly flags Oracle's $553B remaining performance obligations as a visibility edge, yet glosses over the $50B FY2026 capex commitment—roughly 2.6x its expected annual revenue—plus the planned $45-50B debt raise. This scale of spending risks margin compression, equity dilution, and power/construction bottlenecks that could stretch beyond 2026. Palantir's 150% net dollar retention and 57% FCF margins show durable commercial traction, but its 62.7x sales multiple assumes flawless AIP expansion. Investors should model scenarios where Oracle's backlog converts slower than the 20% growth forecast, eroding the touted risk-reward advantage.
Oracle's backlog could convert faster than modeled if hyperscale demand accelerates, allowing the stock to rerate despite near-term cash burn.
"Both stocks are priced for flawless execution in an environment where capex inflation, power scarcity, and AI adoption plateaus are increasingly likely."
The article's framing obscures a critical math problem: Palantir's 153.9x P/E requires 85% YoY growth to sustain indefinitely, but growth deceleration is inevitable. Oracle's 34.7x P/E looks cheap until you model the $50B capex burden—that's ~8% of market cap annually, diluting per-share value regardless of backlog conversion success. The real risk isn't execution; it's that neither company's valuation properly prices the probability of both growth normalization AND capex overruns. Oracle's backlog visibility is real but not destiny—data center buildouts routinely face power constraints, cost inflation, and customer demand shifts.
Palantir's 150% net dollar retention and 60% operating margins suggest pricing power and efficiency that could justify premium multiples if AIP becomes genuinely embedded in enterprise workflows—making the valuation less precarious than it appears.
"Palantir's superior operating margins and net dollar retention justify a premium valuation over Oracle's capital-intensive infrastructure pivot."
The article frames this as a growth-versus-value trade, but that ignores the structural divergence in their business models. Palantir is a software-first entity with 60% operating margins, effectively acting as the 'operating system' for enterprise AI. Its 153x P/E is egregious, but it reflects a land-grab phase where net dollar retention of 150% proves sticky, high-margin lock-in. Conversely, Oracle is pivoting to an infrastructure-heavy model. The $553 billion backlog is impressive, but it masks massive capital intensity and execution risk in power-constrained data centers. Oracle is essentially becoming a utility with lower margins, while Palantir is scaling as a high-margin SaaS platform. I am skeptical of Oracle's ability to maintain margins while funding $50 billion in annual capex.
If the AI market shifts from software-layer value capture to infrastructure-layer dominance, Oracle’s massive physical capacity will become the bottleneck that dictates pricing power, rendering Palantir’s software irrelevant without the underlying compute.
"Oracle's massive AI backlog and scalable cloud build-out offer a clearer, more durable risk-reward path for AI spend than Palantir's high-valuation growth story."
Palantir's growth is impressive but the takeaway is the valuation risk: 153.9x earnings and 62.7x sales leaves little cushion if AIP adoption slows or competition intensifies. Even with a 150% net dollar retention and 60% adj operating margins, sustainability hinges on multi-year expansion into non-government clients and price discipline. Oracle offers a different risk-reward: a vast AI cloud backlog (RPO around $553B) and a capital-light model once capacity is added, plus earnings visibility from recurring revenue. The challenge for Oracle is heavy capex (target ~$50B in 2026) and execution risk around data-center timing, cost, and financing conditions. If AI demand underdelivers, Oracle's cash flow could suffer.
Palantir could surprise with stronger enterprise penetration and government demand that keeps revenue growing at a credible pace, making the 'high multiple' argument less fragile. Oracle's backlog conversion depends on capex funding and project execution; if financing costs rise or projects delay, earnings visibility could erode.
"Both valuations embed the same unpriced assumption that AI spending keeps accelerating through 2027."
Claude's capex dilution math for Oracle is sound, but it misses the correlated timing risk across both names. Palantir's 150% NDR and 57% FCF margins assume continued enterprise budget expansion through 2026-27; if ROI questions slow that spending, Oracle's backlog conversion and Palantir's retention could both disappoint in the same window. The $45-50B debt raise only amplifies Oracle's exposure to that single macro trigger.
"Palantir's embedded customer base buffers it better than Oracle's forward-looking backlog if AI spending sentiment deteriorates."
Grok's macro trigger insight is sharp, but it assumes both companies face identical budget-cycle risk. That's incomplete. Palantir's 150% NDR suggests existing customers are *deepening* spend—less vulnerable to budget freezes than new logo acquisition. Oracle's backlog conversion depends on *new* capex commitments from hyperscalers. If enterprise AI ROI questions hit, Palantir's retention holds; Oracle's backlog stalls. They're correlated but asymmetrically exposed.
"Palantir's net dollar retention is more vulnerable to macroeconomic budget cuts than Oracle's contractual backlog."
Claude, you're missing the 'lock-in' fallacy. Palantir’s 150% NDR is a trailing indicator of past pilot success, not a guarantee of future budget resilience. If enterprise AI ROI fails to materialize, even 'deepened' spend is the first item cut. Meanwhile, Oracle’s backlog is contractual, not discretionary. By focusing on 'new logo' risk for Oracle while assuming Palantir's retention is immune to macro-budgetary shifts, you’re miscalculating where the real revenue cliff lies when the AI hype cycle inevitably cools.
"Debt-financed capex timing and financing mix, not a fixed capex percentage, will dominate Oracle's and Palantir's cash flow and margins."
Claude's capex dilution math assumes the burden lands as a fixed ~8% of market cap annually and ignores how debt financing and tax shields alter cash flow per share. The real risk is timing and financing mix of the $50B, not just headline capex. If hyperscaler capex stalls or financing costs rise, Oracle’s margin and backlog conversion stall; Palantir's retention could still hold, but higher levered cash burn drags both names into a riskier cycle.
The panelists agree that both Oracle and Palantir face significant risks, with Oracle's heavy capex and Palantir's high valuation being the primary concerns. They differ in their assessment of which company is better positioned to navigate these challenges.
Palantir's software-first business model and high operating margins
Oracle's heavy capex and Palantir's high valuation