From Metaverse to Muse: Bernstein Says Meta (META.US) Bull Case "Fully Returns," Muse Poised to Unlock "Endless" Growth Opportunities

Bernstein analysts led by Mark Shmulik published a new research report maintaining an "Outperform" rating on Meta Platforms (META.US) with a price target of $800. The firm said Meta is refocusing its business around Muse, positioning Muse as a new AI operating system (AI OS), with its various products becoming extensions of core AI capabilities, which could unlock a series of unknown growth opportunities; but in the near term, both valuation and expectation thresholds have risen, and the company faces a somewhat more challenging fundamental environment. Bernstein opened the report by quoting a line from Facebook's internal "Little Red Book": "If we don't build the thing that kills Facebook, someone else will." The analysts said the firm's earlier bullish view on metaverse investment was based on this perspective: if the world moves toward the metaverse, Meta will not only survive but lead the next era. That logic has now "fully returned" to Muse. The firm emphasized that the core of its bullish thesis on Meta is that Mark Zuckerberg will not let the company fail. Shmulik likened Meta to an "AI Thanos," saying it has been collecting the "Infinity Stones" needed to become an AI player: compute, talent, and frontier models. After the Connect developer conference, the firm's key takeaway was that Meta is repositioning its business around Muse, with Muse as the new AI OS and every product an extension of the AI core. From the core business and hardware to new products starting with the Muse app, the potential opportunities to serve creators, small and medium businesses, advertisers, and even enterprise customers are "endless." The firm noted that under the framework of Muse as the core, it is not overly concerned about near-term monetization of Muse Agent. Early download and retention trends are encouraging, but it also cautioned about the large total addressable market (TAM) implied by early enthusiasm, noting that prior sizing of Facebook Shops and the metaverse was overly optimistic. But what is certain: first, ROIC at the strategic level is greater than zero; second, investment levels are rising; third, Meta has ample resources to adjust direction flexibly based on market conditions and changes in consumer behavior. Bernstein elaborated on the Muse strategy from three dimensions. For consumers, the Muse flywheel has already started: the app has consistently ranked among the top downloads in the U.S., and real-time voice, customizable personalities, avatars, email, computer use, and expanding commerce and productivity integrations are driving Muse beyond the traditional chat Siasun Robot&Automation experience. At launch, dozens of publishing partners and Shopify product catalog integration were already in place, and the partner list continues to expand. The firm said this list should continue to grow, while the hardware strategy also fits the "use Muse in any way" framework. For developers, the firm said the goal is not just to have the best model; the key to winning developers is rapid innovation iteration and high accessibilitygetting new innovations into developers' hands as quickly as possible and making capabilities usable wherever developers want to build. Developer ecosystem adoption stems from lowering the barrier to use, and Meta is focused on eliminating every friction point between developers and users. For investors, this is "an AI ROI story in the making." At a minimum, ROIC on consumer AI investment is now greater than zero, easing long-standing investor concerns related to sustained metaverse investment and losses. Meta's AI investment scale ranks among the industry's largest, but unlike hyperscale cloud providers, it cannot rely on cloud demand to justify returns. Muse provides a consumer-facing revenue stream that can concretely demonstrate the output of large investments. Even if commercialization starts small, investors can at least begin to connect AI investment with measurable results: a few million subscribers, billions of dollars in gross merchandise value, app downloads, user engagement, and a range of other metrics can be used to measure the growth of the consumer AI business. Bernstein also said Muse is no longer just a model or another app, but is becoming the intelligence layer at the center of Meta's ecosystem. Glasses, VR, companion devices, the standalone Muse app, and integrations within the family of Meta apps all look like different endpoints of the same assistant. Hardware is the container; Muse is the product. From this perspective, the device itself matters less than the assistant. The firm said that last year the industry often felt like it was searching for devices consumers would adopt, while this year it feels more like Meta is saying "assistant first," with users able to access it in any way. Meta is no longer forcing consumers to choose a specific form factor, but building the intelligence layer and letting hardware follow. Short-term earnings: strong revenue, profit dragged by $10 billion legal settlement Looking ahead to fiscal Q3 2026 earnings, Bernstein expects Meta revenue of $64.058 billion, above the consensus of $63.312 billion, up 25% year over year; DAU/MAU forecast at 3.646 billion, 1% above consensus, implying/guidance of 19%-25% year-over-year growth. But EBIT is expected to be $11.338 billion, 40% below the consensus of $19.016 billion, mainly due to $10 billion in legal costs related to a teen media usage settlement recorded in the quarter. EPS is expected at $3.95, above the consensus of $3.56; free cash flow is expected at -$13.756 billion versus consensus of -$8.953 billion; capex is expected at $41.958 billion versus consensus of $41.011 billion. The firm said Q3 revenue fundamentals remain strong, with Meta's digital advertising engine continuing to run at full speed, benefiting from AI improvements in targeting and ad creative. Even with an approximately -1% FX headwind, the firm still expects Meta to exceed its Q3 revenue guidance of $61-64 billion (19%-25% year-over-year growth). However, the bar for Q4 is rising: the company faces a high 24% year-over-year comparison base from the prior-year period, while FX pressure is expected to persist. Fundamentals remain solid, but the high base combined with continued FX headwinds will make Q4 guidance harder to achieve. Bernstein raised its 2026 and 2027 revenue forecasts but lowered EBIT and EPS forecasts, and significantly cut its free cash flow forecast. Additionally, the firm noted that discussion of investment and capex is expected to remain a market focus. Last quarter, Meta raised the low end of its 2026 capex range by $5 billion to $130 billion while keeping the high end unchanged at $145 billion, suggesting spending may trend toward the upper end of the range. Looking ahead, the bigger focus is 2027. Meta has consistently said investment levels will remain high to continue ensuring adequate AI capacity, while also continuing to invest in Muse marketing, and investors increasingly expect AI-related spending to rise again next year. The firm also said that as early Muse enthusiasm builds, the market has begun discussing equity financing as another means of funding ambitious infrastructure plans. Bernstein said early Muse enthusiasm gives Meta the key evidence it has lacked throughout this capex cycle: visible evidence that AI investment is translating into consumer demand, and management can now point to a consumer AI product demonstrating explosive user growth. Investors are increasingly focused on whether management has this supporting evidence to launch a new, larger AI capex cycle. How high is Muse's growth ceiling? Finally, Bernstein raised a question: how high is Muse's growth ceiling? The firm said consumer adoption speed is hard to predict, and despite strong momentum, history shows that meaningful behavior change is often slower and less predictable than adoption curves suggest. Bernstein cited several historical examples: e-commerce has grown for more than 20 years yet still accounts for less than 20% of total U.S. retail sales, while smartphone adoption was faster and now exceeds 80% of the U.S. population. ChatGPT, despite being one of the fastest adoption curves in consumer tech history, remains mainly a useful but relatively narrow productivity tool three years after launch, with commercialization concentrated in limited applications; its agentic commerce push appears to have had limited success so far. Compared with TikTok, ChatGPT grew users faster, but engagement has already shown signs of peaking, with daily time spent per user stabilizing at 15-20 minutes versus about 90 minutes for TikTok. TikTok plays a larger role in consumers' daily lives, has more obvious commercial success, and generates significantly higher consumer revenue, at roughly twice ChatGPT's revenue per DAU per year. The firm believes product ease of use and strong contextual value can significantly accelerate consumer product adoption and reshape competitive dynamics. History shows that superior technology alone is often not enough; winning products are often those that make the experience simpler, more intuitive, or more immediately valuable to users. Facebook displaced MySpace with a cleaner, simpler, more useful social networking experience; Gemini last year successfully captured significant user mindshare through the viral Nano Banana image editing feature. Muse may represent a similar inflection point: if it can unlock a clearly simpler, more engaging, or uniquely differentiated consumer use case versus AI chat Siasun Robot&Automation such as ChatGPT, it could stand out and accelerate mainstream adoption.