代理型AI会推动苹果估值的下一个拐点吗？目标价上调至380美元

Agentic AI shiftscontrol from model to trusted endpoint 26 May 2026 In an agentic world, value accrues to the platform that controls user intent, personalcontext, app access, permissions, identity, authentication, payments, and trust.Smartphoneis the scaled consumer device where these factors already converge. If AI assistants become the new front door to search, apps, commerce, scheduling, payments,and workflow completion, we think Apple should have meaningful leverage over modelproviders, app developers, merchants, advertisers, and payment networks. We raise ourPO to $380 on 37x our C27E EPS of $10.29 from $330 on 32x C27E of $10.29. We raisethe multiple to account for the significant opportunity in agentic AI. Reiterate Buy rating. Equity Wamsi MohanResearch AnalystBofAS+1 646 855 3854wamsi.mohan@bofa.com iPhone is well positioned because Apple controls the stack Apple’s agentic AI moat comes from Apple silicon & iOS. Silicon dictates how muchinference can happen locally (matters for latency, privacy, reliability, and cost). OSdetermines whether AI can access user context, call apps, route requests, authenticateidentity, request confirmation, & complete tasks in a trusted way. Our analysis aroundtoken usage (Fig 16, 17, 19) suggests Apple will need to rely on distributed compute (ondevice), Private Cloud Compute (PCC) and third party compute to deliver a robust AIexperience. Removing the net cash neutral target, in our opinion, is a reflection of a newperiod of AI driven investments to enable this agentic era. Ruplu BhattacharyaResearch AnalystBofAS+1 646 855 0315ruplu.bhattacharya@bofa.com Ryan Seungin ChoiResearch AnalystBofAS+1 646 743 0587ryan.choi2@bofa.com Aisling GrueningerResearch AnalystBofAS+1 646 855 4273aisling.grueninger@bofa.com What AAPLneeds to do: make Siri the execution layer To win, Apple needs to evolve Siri into the orchestration layer of the iPhone.Thisrequires five priorities: 1) make newer iPhones materially better AI devices throughstronger NPUs, memory, bandwidth, etc; 2) scale hybrid inference across on-devicemodels & PCC; 3) evolve Siri into an agent that understands intent, retrieves context,calls apps, and completes workflows; 4) make App Intents a broad framework formachine-callable app actions; and 5) preserve trust through granular iOS permissions,and clear user confirmation. The near-term risk is execution which will determine theability to capture more of the agentic value pool (Figure 5). We estimate agentic Siricould represent a ~$15-30bn incremental F30E revenue opportunity in the base caseand ~$40-65bn in the bull case or potentially adding up to $2 in incremental EPS (Fig23). See Abbreviations at the end of thereport BofA Securities does and seeks to do business with issuers covered in its researchreports. As a result, investorsshould be aware that the firm may have a conflict ofinterest that could affect the objectivity of this report. Investors should consider thisreport as only a single factor in making their investment decision.Refer to important disclosures on page 35 to 37. Analyst Certification on page 33. PriceObjective Basis/Risk on page 33. iQprofileSMApple Inc. Company SectorIT Hardware Company Description Apple Inc. (AAPL)designs, manufactures, and marketsconsumer electronics and computers, and has developed itsown proprietary iOS, Mac OS, TvOS and Watch OS operatingsystems and related software platform/ecosystem.Revenues are principally derived from the iPhone line ofsmartphones, Services, hardware sales of the Macintoshfamily of notebook and desktop computers, iPad tablets,and wearables. Investment Rationale Our Buy rating on Apple is based on 1) expected strongiPhone upgrade cycle in F25, F26 driven by theneed forlatest hardware to enable Gen AI features, 2) higher growthin Services revenue, 3) higher margins from more internallydeveloped silicon, 4) continuing capital returns, 5) AIfeatures that can drive higher institutional ownership, and6) risk around legal issues being manageable. Stock Data Average Daily Volume46,962,460 Contents AAPL’s AI narrative today Apple may look like an AI laggard today, but in an agentic AI world thekey control point shifts from owning the best model to owning thetrusted endpoint where user intent, identity, payments, apps, andprivacy converge So far, the AI trade has benefited a wide range of the infrastructure layer that comprisedof semis, semi cap, hardware, networking, memory, power, and cloud. When compared tothis mix, Apple has remained a relative bystander with a much more muted stock growthover the last two years. We think some of that skepticism is warranted given that Sirihas lagged expectations, Apple has not been a visible participant in the compute buildoutrace, and frontier model building is not where the company has historically differentiatedvs. other Mag 7 peers or frontier labs. This has fed the perception that Apple is an AIlaggard in a race that is already well underway. We