Разбор исследования: сможет ли Intel вернуть былые позиции благодаря Apple? По расчетам Bernstein направление верное, но цена уже завышена.

Автор: TideResearch

Автор: Rita

Введение от TideResearch

Аналитик Bernstein Стейси А. Расган 18 июня представил исследовательский отчет по Intel, оценивающий недавнее заявление Трампа в поддержку сотрудничества Apple и Intel по разработке и производству ПК-чипов на территории США. В отчете указано, что это сигнализирует о переломе в ландшафте производства чипов, однако первоначальные масштабы будут ограничены: проект носит скорее характер концептуальной проверки (Proof of Concept), нежели возможности для немедленной монетизации. Bernstein сохраняет рейтинг «Держать» для Intel с целевой ценой $100. Послание подразумевает позитивный тренд, которого недостаточно для обеспечения дальнейшего роста акций. Данный отчет рекомендуется инвесторам, отслеживающим политику США в области производства полупроводников, планы расширения мощностей Intel и меры государственной поддержки.

Три ключевых вывода

① Apple 进英特尔代工厂是"小试牛刀"，短期收入贡献微乎其微

According to Bernstein's data, Apple shipped approximately 23.68 million laptops over the past 12 months, with 22.15 million being premium models (priced above $700). If Intel wins 40% of these orders as an initial target, the annual foundry volume for PC chips would be around 5 million units. Based on Bernstein's assumptions, at an average foundry capacity price of $25,000, this business segment would generate roughly $500 million in annual revenue, contributing approximately $0.03 to earnings per share.

Compared to Intel's approximate $55 billion annual revenue and $1.5 annual EPS, this potential contract is virtually negligible from a financial perspective.

The report emphasizes that the value of this order lies not in its immediate revenue scale, but in it representing a trust endorsement from a US client regarding Intel's manufacturing capabilities.

② 特朗普推动的政策温度需打个问号

Despite Trump's recent public statements encouraging Apple and Intel to design and manufacture chips domestically in the US, Bernstein points out that this encouragement is not "mandatory." The report analyzes that clients won't be "forced" to adopt a supplier unless that supplier can prove three conditions: ability to produce to specifications, competitive cost structure, and stable, reliable supply. Intel has currently entered the risk production phase for its 18A process, demonstrating credible technological progress, but mature capacity and cost competitiveness still require observation.

In other words, policy encouragement is a plus, but insufficient to replace market competitiveness.

③ 从概念验证到规模量产的“过渡困境”仍然存在，伯恩斯坦未因此上调评级

Bernstein's report repeatedly emphasizes “there is still a lot of wood to chop here”, meaning that moving from small-scale conceptual verification to mass production requires substantial work, considerable time, and significant capital investment.

In other words, for Intel to truly cross this stage, it must simultaneously address several things: invest tens of billions of dollars to expand capacity, navigate complex and strict customer certification processes, and prove its cost and yield advantages amid fierce foundry price competition.

It is precisely due to this uncertainty that, while acknowledging the positive significance of this partnership, Bernstein maintains a Market-Perform ("Hold") rating rather than upgrading to Outperform ("Beat the Market"). Its stated target price is approximately $100, implying room for correction compared to the current stock price of about $121.10 (as of the report's baseline date).

芯片供应多元化背后的地缘政治逻辑

Apple has long pursued a supplier diversification strategy to reduce reliance on any single foundry. Historically, this diversification mainly manifested in choices between TSMC, Samsung, and different process nodes under Intel.

However, in recent years, geopolitical factors have emerged as a new core variable. The US government is investing massive subsidies through the CHIPS and Science Act to drive the reshoring of critical chip manufacturing. In this context, Apple entrusting some orders for high-end PC chips to Intel's US factory aligns with policy directives and helps enhance supply chain resilience, reducing the risk of over-reliance on a single region (particularly Taiwan).

For Intel, this partnership resembles a crucial validation of market confidence. Previously, due to lagging process technology, Intel had lost Apple's foundry trust. Being able to re-enter Apple's supply chain and participate in collaborative design and manufacturing already implies that its 18A process possesses a certain level of readiness. This signal may also generate spillover effects on other potential customers (such as data center CPU or AI accelerator manufacturers).

短期验证与长期想象空间之间的张力

The core of Bernstein's analysis lies in exposing a structural contradiction: the massive gap between short-term scale and long-term narrative.

In model assumptions, even if Intel secures about 40% of Apple's PC chip orders, the annual shipment volume would be roughly in the range of 100,000 to 150,000 units. Within Intel's overall foundry business plan, this still falls under the "prototype verification" or "conceptual validation" stage, offering very limited revenue contribution—only amounting to several million dollars in revenue, with an impact of mere cents on EPS, making it difficult to substantially alter the company's overall growth trajectory.

However, in the long term, this "small-scale entry" carries clear path-dependency value. If Intel can leverage Apple's orders to demonstrate its process stability and delivery capability, it will have the opportunity to subsequently secure much larger foundry orders for cloud computing chips, AI accelerators, and communications chips. The total addressable market for these segments far exceeds PC chips, possessing significant long-term expansion space.

Bernstein did not quantify this "option-style long-term value" in the report, merely noting its existence, while emphasizing that the realization path remains highly dependent on subsequent customer acquisition and continuous process ramp-up.

投资逻辑：押什么？不押什么？盯什么？

押什么：

• 英特尔 18A 制程的推进不会显著低于市场预期
• 美国在半导体制造本土化方面的政策支持保持延续性

不押什么：

• 短期内苹果合作会显著改善英特尔财务表现
• 政府补贴会直接、快速提升英特尔整体利润率

重点观察信号：

• 英特尔下一季度财报中代工业务的收入确认与毛利率变化
• 18A 制程的良率提升速度与成本下降曲线
• 除苹果外是否有其他大型客户正式导入英特尔代工
• 《芯片法案》相关补贴的实际兑现节奏与规模

本文系潮向研究对第三方券商研究报告的整理与解读。文中引述的评级、目标价、盈利预测及相关判断，均为该券商分析师的观点，仅代表其所属机构立场，不代表潮向研究的观点，也不构成任何投资建议。

阅读时请注意三点：一、目标价是分析师对未来约 12 个月的预期，是预测而非承诺，会随业绩和市场环境反复调整。二、卖方研报天然偏多，且部分覆盖公司与该券商存在业务关系。三、研报的价值在于主线逻辑及其前提假设，而非某一个目标价。看逻辑，不要只看价格。

市场有风险，决策需独立。本文不应作为买卖任何证券的依据。

数据来源：伯恩斯坦研报 (Stacy Rasgon, 2026 年 6 月 18 日)· 英特尔历史财务数据 (SEC)

潮向研究 · TideResearch · 2026 年 6 月