Morgan Stanley believes that negative sentiment towards Meta has bottomed out, with the current valuation creating a rare tactical buying opportunity. The firm has also designated Meta as a top pick.
According to analysts, Morgan Stanley's Brian Nowak maintained an Overweight rating on Meta Platforms in a research report dated March 29, while lowering the price target from $825 to $775. This still implies approximately 50% upside potential from the current share price of around $526.
The report indicates that due to concerns over GenAI investment returns, macroeconomic uncertainty, and regulatory pressures, Meta is currently trading at a valuation of about 15 times its estimated 2027 earnings per share. This is one standard deviation below its 10-year average and represents only the fourth time such undervaluation has occurred in the past decade. This valuation level, combined with the potential for the company's AI agent products to materialize, forms the core of Morgan Stanley's bullish thesis on Meta. Furthermore, Meta is advancing a plan to reduce its workforce by approximately 20%. Morgan Stanley estimates this could generate annual cost savings of $3 billion to $10 billion, potentially adding over $1 to the 2027 EPS, thereby providing stronger bottom-line support during a heavy investment cycle.
A key potential growth driver for Meta is its prospective AI agent product, MetaClaw. Leveraging the Llama large language model and its base of 3.5 billion active users, this tool aims to create a closed-loop "agent-based shopping" experience by integrating merchant inventories and payment capabilities. This could shift Meta's monetization from advertising display towards transaction facilitation. Morgan Stanley has concurrently trimmed its 2026 and 2027 advertising revenue forecasts by about 1% each to reflect more conservative macroeconomic assumptions. Even with these adjustments, the 2027 EPS estimate remains at $36.31, indicating that significant upside potential is still intact.
Meta's current valuation is at a notably discounted level compared to other large-cap technology stocks covered by Morgan Stanley.
Historically, Meta's forward P/E ratio is approximately 15x, below its 10-year average of 20.5x and even beneath the one-standard-deviation lower bound of 16.4x. Such a low valuation has occurred only four times in the past decade, including during the 2018 user growth and margin crisis, the impact of Apple's ATT privacy changes, and periods of heavy investment in Reality Labs.
On a comparative basis, Meta's 2027 PEG ratio is approximately 0.9x, while the median PEG ratio for Apple, Microsoft, Alphabet, Amazon, and Netflix is around 2.0x, indicating a discount of over 54%. Even based on Morgan Stanley's $775 price target, Meta's PEG would still trade at a discount of about 33% to the peer group median. Meanwhile, Meta's projected EPS compound annual growth rate from 2025 to 2027 is 16%, exceeding the median of 10% for the comparable companies mentioned.
Morgan Stanley's base-case price target of $775 implies a 2027 P/E of about 21x, roughly in line with the historical average. The bull-case target is $1,000, while the bear-case target is $450, suggesting potential upside of approximately 90%.
Amid the current valuation disconnect, a central market concern is whether Meta's estimated $190 billion in effective capital expenditures, including massive computing infrastructure procurement, can yield tangible commercial returns. Morgan Stanley posits that the answer lies in Meta's potential AI agent product, "MetaClaw."
The envisioned MetaClaw framework would be built upon Llama and more advanced large language models, with a potential intermediary layer for user-agent interaction. Combined with Meta's ecosystem—spanning over 3.5 billion daily active users across Facebook, Messenger, Instagram, and WhatsApp, 250 million businesses, and over 10 million advertisers—it could form a closed-loop AI agent platform for shopping and services. Reports suggest Meta is already testing integrations with services like Gmail and Google Calendar into its agent products.
Although Meta lacks a standalone browser, the inventory information and payment capabilities associated with its 250 million business users provide foundational infrastructure for enabling "agent-based commerce." Meta initiated testing of GenAI-driven shopping features less than a month ago. If MetaClaw is successfully developed, it could drive a multi-year revenue growth cycle similar to search, extending Meta's monetization from ad display into transaction completion.
Simultaneously, on the advertiser side, Meta is developing fully automated creative generation and ad placement tools for small and medium-sized businesses, potentially increasing its share of advertiser budgets.
Morgan Stanley's core advertising forecasts, while not overly aggressive, remain above current market consensus.
The firm predicts Meta's advertising revenue will grow approximately 28% in 2026 and 21% in 2027, reaching about $257.5 billion and $311.6 billion, respectively. These estimates are roughly 3% to 5% above consensus. This view is supported by the fact that user engagement, measured by daily time spent on Meta's apps, continues to grow, and the increasing share of video content enhances the monetization potential of the traffic.
Regarding the product pipeline, Meta's Q4 2025 earnings indicated at least 10 new features slated for 2026. These include enhancements to content recommendation driven by large models, expansion of ad inventory on WhatsApp, and the rollout of ads on Threads in markets like the UK, EU, and Brazil. Visibility for core business growth over the next 1-2 years remains clear, with the next significant growth inflection point anticipated around 2027. By then, large language models are expected to analyze Meta's proprietary data to further improve the quality of contextual advertising signals.
Morgan Stanley views the reported potential 20% workforce reduction as a positive development, suggesting it could establish a higher safety margin for EPS.
A reduction of approximately 15,773 positions, estimated at an average annual cost per employee of $200,000 to $600,000, could yield annual operational savings between $3.15 billion and $9.46 billion. This equates to roughly 3% to 9% of Morgan Stanley's 2027 EBIT forecast for Meta. On a per-share basis, this could contribute over $1 to the 2027 EPS beyond the base case, serving as a buffer if the advertising market weakens.
Notably, these potential savings are not included in the base-case model, which already anticipates a significant increase in depreciation and amortization to approximately $31.9 billion in 2026 and $51.3 billion in 2027, reflecting continued heavy infrastructure investment. If the workforce reduction is implemented, it could effectively offset some of these rising costs, adding greater certainty to the $36.31 EPS forecast for 2027.
Morgan Stanley identifies May and September 2026 as two potential key catalyst windows for Meta.
In May, following Meta's inaugural LlamaCon AI developer conference last year, a potential follow-up event in 2026 could feature announcements regarding new models and product progress, with some media reports also suggesting new model releases around that time. In September, Meta typically holds its annual Connect developer conference, which has historically been a platform for unveiling product roadmaps and significant technological milestones.
Product and model announcements during these periods, particularly any developments related to MetaClaw's capabilities, are seen as crucial variables that could drive a market reassessment of Meta's return on invested capital visibility and subsequently lead to valuation recovery. Currently, options market pricing implies approximately a 13.6% probability of Meta's share price reaching the $775 target within the next 12 months, a 0.9% probability of hitting the $1,000 bull-case scenario, and a 31.7% probability of falling to the $450 bear-case scenario.
Comments