MONEY MASTER

Meta reports  4Q25 and FY2025 results after the close on Jan 28 , but the quarter itself is unlikely to decide the stock’s direction. Management has already guided  4Q revenue at US$56–59bn , and consensus sits modestly above the upper end. Instead, this earnings event is best viewed as a  gateway to 2026  — where  AI spending discipline, monetisation credibility, and execution timelines  will matter far more than headline beats. Quick Setup: Q4 Is Fine — The Debate Is 2026 Consensus snapshot Revenue:  ~US$58.3bn (+20% YoY) EPS:  ~US$8.19 (implies ~US$20.6bn net income) A “not-bad” Q4 is largely priced in. The real swing factor will be: FY26 OpEx & CapEx guidance Clarity on AI monetisation milestones Evidence that spend is converging toward earnings, not drifting away from it 3 Things That Will Move the Stock 1. FY26 CapEx & OpEx: Spend Is Assumed — Accountability Is Not Street expectations are aggressive: FY26 OpEx:  ~US$147–152bn (+...

This week is not about chasing upside — it’s about  managing exposure through a policy and earnings convergence . With the  FOMC decision  and a heavy slate of  mega-cap earnings , markets are testing whether valuations can hold  without immediate rate relief . The base case is stability, but positioning remains vulnerable to  tone shocks , not data surprises. Think of this week as  risk calibration , not trend confirmation. Macro Playbook: FOMC Is the Risk Switch Base expectation Rates held unchanged No new forward guidance Powell reiterates “data-dependent” stance What actually matters How Powell frames  inflation persistence vs growth cooling Whether he  pushes back  on market-priced mid-year cuts Any signal on  financial conditions  becoming too loose Market interpretation Flexible tone  → risk assets stabilize, volatility fades Rigid tone  → duration reprices, equities de-rate quickly Playbook stance: Stay neutra...

Meta Platforms周二透露，今年围绕生成式人工智能可能扰乱全球主要选举的担忧大多未成现实。根据Meta的报告，该技术在其平台Facebook和Instagram上的影响相对有限。 Meta全球事务总裁尼克·克莱格（Nick Clegg）在新闻发布会上表示，试图利用人工智能传播宣传或虚假信息的协作网络未能获得显著影响力。他补充道，AI生成的虚假信息总量较低，Meta迅速对相关内容进行标记或删除。 AI对公众舆论的有限影响 假消息专家也指出，AI生成的深度伪造（deepfake）和虚假叙述在影响公众舆论方面效果不佳。一些引起广泛关注的例子，包括模仿总统乔·拜登的深度伪造视频和音频，均被迅速揭穿，限制了其传播范围。 克莱格将这一现象部分归因于Meta改进的检测系统。今年，Meta共拆除了约20个隐秘影响行动。然而，他也警告称，不良行为者正转向安全措施较少的平台，或创建自己的网站以规避内容审核。 优化内容审核策略 回顾以往的实践，克莱格承认，在新冠疫情期间，Meta可能在内容审核方面过于严格，导致某些合法内容被误删。Meta已调整了其审核策略，更好地平衡了言论自由和规则执行之间的关系。 “我们觉得自己可能有点过火了，”克莱格坦言。“尽管减少有害内容仍然是优先事项，我们正专注于提升规则执行的精准性和准确性。” 与新一届美国政府的互动 克莱格还提到，Meta首席执行官马克·扎克伯格计划与即将上任的特朗普政府就技术政策进行互动。扎克伯格特别希望参与关于人工智能及美国全球技术领导地位的讨论。 “马克非常愿意在任何政府需要讨论的AI和技术进步议题中发挥积极作用，”克莱格表示。 社交媒体监管的更广泛辩论 Meta的声明发布之际，共和党议员持续批评社交媒体涉嫌审查保守派观点。扎克伯格此前曾在8月致信众议院司法委员会，承认在拜登政府压力下删除内容时存在错误，并对此表示遗憾。 随着人工智能的不断发展，Meta应对这些挑战的方式可能决定未来在数字平台上虚假信息管理与言论自由的平衡走向。