证券研究报告·宏观报告·宏观深度报告 ◼ 市场表现回顾： ◼ 事件驱动盘点： ◼ 指数后市展望： ◼ 风险提示： 1)行业政策或监管环境突变；2)宏观经济不及预期；3)发生重大预期外的事 件。 金融产品深度报告 20260202 纳斯达克 100ETF，1 月复盘与 2 月展望 2026 年 02 月 02 日 [Table_Tag] [投资要点 Table_Summary] 证券分析师 芦哲 执业证书：S0600524110003 luzhe@dwzq.com.cn 证券分析师 唐遥衎 执业证书：S0600524120016 tangyk@dwzq.com.cn 相关研究 2026-02-01 东吴证券研究所 1 / 19 请务必阅读正文之后的免责声明部分 ◼ 走势复盘：2026 年 1 月，纳斯达克 100 指数月度上涨 1.20%，整体震荡 向上。上旬弱经济数据强化宽松预期，推动指数上行；中旬因 PPI 超预 期、鲍威尔遭调查及特朗普地缘言论等风险集中爆发，引发指数调整； 下旬政治危机缓和，财报季验证 AI 与半导体需求强劲，但高位 PPI 及 特朗普提名鹰派美联储主席候选人重燃政策疑虑，导致指数回调。 ...

Core Insights - Ruoming Pang, head of Apple's AI models, is leaving to join Meta's "superintelligence" team, marking another significant talent acquisition for Meta following similar moves from OpenAI, Google, and Anthropic [1] - Pang's career trajectory includes a 15-year tenure at Google, where he was a key contributor to the Tacotron 2 project in the fields of speech recognition and human-computer interaction [4][6] - At Apple, Pang led a team of approximately 100 people focused on developing models for Apple Intelligence, which supported various AI features announced at WWDC [9] Company Developments - Meta is actively building a strong AI team, having recently recruited several high-profile talents, including Yuanzhi Li from OpenAI and Anton Bakhtin from Anthropic, with Pang's compensation package reportedly worth tens of millions annually [13] - In contrast, Apple's AI team is experiencing turmoil, with reports of multiple engineers planning to leave for Meta and other companies, indicating potential instability within Apple's AI division [11] - Following Pang's departure, the Apple Foundation Models (AFM) team is now led by Zhifeng Chen, with a shift towards a more distributed management structure [11] Industry Context - The competition in the AI sector is intensifying, with major players like Meta, OpenAI, and Apple not only competing on technology and products but also on talent acquisition [15] - Apple's approach to AI has been described as conservative, focusing on a "device-first" philosophy that emphasizes privacy and local computation, which may hinder its competitiveness in the era of large models [15]