Morgan Stanley points out that Meta's current stock price does not fully reflect the potential value of its AI transformation, suggesting its growth logic is undergoing a qualitative leap. If the previous stock surge was based on cost-cutting during the "Year of Efficiency," the next phase of growth will be entirely driven by the core return on invested capital (ROIC) delivered by AI.
According to the analysis, on January 29, Morgan Stanley's analyst team led by Brian Nowak significantly raised their price target for Meta from $750 to $825 in a report, implying approximately 15% upside from current levels.
The Morgan Stanley report not only increased the 2027 earnings per share (EPS) estimate by 10% but also made a bold prediction: Meta's advertising revenue is expected to surpass that of Google Search for the first time in the second quarter of 2026, ending the latter's long-standing dominance in the digital advertising arena. The confidence behind this stems from the clear signals Mark Zuckerberg conveyed to investors during the earnings call: Meta is using large language models to completely rewrite its recommendation system, which in his view currently appears "quite primitive." Morgan Stanley believes that if the current system is indeed primitive, then the long-term opportunities and improvements are likely to unlock significant further growth. The firm anticipates Meta will make further strides in personalization (enabled by more efficient analysis of vast data), agent services (like Meta AI), new creative/messaging services for users/advertisers, and wearables. This represents not just a technological upgrade, but a generational leap from "monetizing traffic" to building an infrastructure for "personal super-intelligence."
A Historic Moment: Meta's Ad Revenue to Surpass Google Search in Q2
The most impactful prediction in the research report lies in the comparison of the advertising businesses of the two tech giants. Morgan Stanley notes that while Alphabet is also accelerating its execution, Meta's gains in user engagement and monetization driven by AI are more astonishing. For perspective, the report mentions that when Google Search's quarterly revenue was around $57 billion, its year-over-year growth rate was 15%; in contrast, when Meta reaches a similar scale, its growth rate is projected to be a robust 34%-35%. Based on Morgan Stanley's current model projections, Meta's quarterly advertising revenue will officially surpass that of Google Search in Q2 2026, and the gap between the two is expected to widen from that point onward. This prediction not only signals Meta's increasing dominance in the ad market but also validates the absolute advantage of its AI-driven recommendation algorithms in capturing user time and advertiser budgets.
Zuckerberg's "Humblebrag": Recommendation System Being Rewritten by LLMs, Current System Still "Primitive"
During the earnings call, Mark Zuckerberg's comments provided the underlying technical foundation for Morgan Stanley's bullish outlook. He stated that while Meta's world-class recommendation system has already driven a more than 30% year-over-year increase in Instagram Reels watch time in the US, the current system appears very "primitive" compared to the impending technological transformation. Meta is undertaking a massive engineering project: rebuilding the entire recommendation system into a scalable engineering architecture akin to large language models (LLMs). The current system, while effective, is poised to be replaced by models capable of understanding users' unique goals, context, and interests. This architectural overhaul not only determines user dwell time on the platforms but also directly influences advertising pricing and conversion rates. Data shows that merely by simplifying the ranking architecture, Facebook increased views of organic News Feed and video posts by 7% in Q4, which management described as the single most significant product optimization for revenue growth in the past two years.
The Tangible Impact of AI Monetization: Comprehensive Efficiency Gains from Reels to Conversion Rates
The Morgan Stanley report details how Meta is converting computational power into tangible revenue through AI technology. This is not an abstract concept but is reflected in every key operational metric:
On the engagement side: Through more efficient model scaling, Instagram Reels watch time surged over 30% year-over-year; by simplifying the ranking architecture, Facebook boosted views of organic News Feed and video posts by 7%, marking the most significant product-driven revenue boost in two years. On the advertising side: Meta expanded the coverage of its ad ranking model, GEM, to all Reels, supported by training on double the GPU capacity. The results showed a 3.5% increase in Facebook ad clicks and an over 1% improvement in Instagram conversion rates. On the efficiency side: Introducing new runtime models boosted Instagram conversion rates by 3%; meanwhile, the adoption of AI coding tools for internal employees increased engineer output by 30% in 2025, with output for some power users surging by up to 80%.
Capital Expenditure and Infrastructure: Paying for the Future
Addressing market concerns about capital expenditure, Morgan Stanley's report provides clear expectations. Meta's capital expenditure guidance for 2026 is in the range of $115 billion to $135 billion, primarily allocated to infrastructure for its super-intelligent labs and core business operations. Morgan Stanley's model anticipates infrastructure spending (including cloud spend, depreciation, etc.) will increase by approximately $36 billion, driving about 75% of the projected operating expense growth in 2026.
While the investment is substantial, Morgan Stanley views it as a necessary cost for sustaining "durable growth." Meta explicitly stated that capacity constraints persist, with demand growing faster than supply. To address this, Meta is aggressively procuring chips (including from NVIDIA, AMD, and its own MTIA) and building the Andromeda architecture capable of running across different chips. For investors, as long as revenue growth outpaces the cost of investment—as Morgan Stanley predicts—this high-intensity capital expenditure is justified, as it is building a formidable AI moat for Meta.
Comments