Meta's AI narrative is improving, but high capital expenditures keep the rating at Neutral
AI summary card
Meta's AI narrative is improving, but high capital expenditures keep the rating at Neutral
JPMorgan believes that Muse Spark 1.1, the Meta Model API, and monetization of computing power have increased Meta's potential to capture value from the AI economy, but rapidly rising capital expenditures in 2026-2027 and insufficient validation of AI monetization still limit an upgrade.
- Meta's share price has risen approximately 23% from its late-June low, significantly outperforming the SPX's approximately 3% gain over the same period, mainly driven by improving sentiment toward its AI strategy.
- Muse Spark 1.1 has shown significant improvements in agentic and coding capabilities, potentially narrowing the gap between Meta and leading AI labs such as Anthropic, OpenAI, and Google.
- The Meta Model API has entered public preview and is viewed as Meta's first form of external AI monetization, with pricing at approximately 25% of that of leading AI models, which should help attract developer and enterprise adoption.
- The report forecasts Meta's 2026 capital expenditures at $142B, up 104% year over year, and 2027 capital expenditures at $202B, up 42%; 2027 capital expenditures could still be revised upward.
- The target price is $725 for December 2026, based on approximately 21x 2027E GAAP EPS of $34.19 and supported by a DCF model.
Report interpretation
Overview
This report focuses on the strategic inflection in Meta Platforms Inc's AI strategy. JPMorgan believes that progress in the Muse Spark 1.1 model, attempts to monetize the Meta Model API externally, and a potential AI product cycle have clearly improved the market's view of Meta's AI narrative. At the same time, the report maintains a Neutral rating because AI monetization remains in the early validation stage, while large-scale computing investments are driving a significant increase in capital expenditures and free cash flow pressure in 2026-2027.
Core views
The core view is that Meta's AI outlook is more attractive than before, but not yet sufficient to support a more positive rating. Positive factors include progress in model capabilities, API commercialization for developers and enterprises, internal AI product opportunities based on a user base of approximately 4B, and potential incremental ROI from leasing computing power or developing cloud computing capabilities. Limiting factors include the excessive scale of AI investment, the risk of delayed monetization, the possibility that 2027 capital expenditures will continue to be revised upward, and competition from Google, TikTok, and OpenAI in advertising and AI products.
Analysis framework
The report uses a company fundamental research framework, combining AI model progress, external API monetization, computing supply, and capital expenditure plans to assess Meta's earnings upside and cash flow pressure. The valuation centers on 2027E GAAP EPS, applying a multiple of approximately 21x and cross-checking the $725 per-share target price with a DCF model.
Methodology notes
The target price is based on approximately 21x 2027E GAAP EPS of $34.19.
This method combines Meta's earnings forecasts with a relative valuation multiple, reflecting the report's assessment of its revenue growth, cost efficiency, and premium relative to the SPX.
The DCF model likewise produces a per-share valuation of approximately $725, assuming a WACC of approximately 14% and a terminal growth rate of 10%.
The DCF is used to validate the reasonableness of the target price while explicitly reflecting the impact of high capital expenditures on future free cash flow.
AI investment returns are assessed through Muse Spark 1.1, the Meta Model API, enterprise adoption, and computing power leasing opportunities.
The report believes that model progress and the API pricing strategy could drive external AI monetization, but developer, enterprise, and internal product adoption still require monitoring.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- Meta Platforms Inc (META.O)Core covered security
- Strengths
- Possesses a social graph, a user base of approximately 4B, advertising targeting capabilities, improved AI-driven advertising ranking and recommendation, and rapidly expanding AI infrastructure.
- Weaknesses
- AI monetization remains less fully validated than that of some internet peers, while substantially higher capital expenditures in 2026-2027 are putting pressure on free cash flow for multiple years.
- Comparison
- The report compares Meta's model progress with leading AI labs such as Anthropic, OpenAI, and Google, and believes Muse Spark 1.1 could narrow the gap.
- Risks
- Overinvestment in AI and delayed monetization, failure of product and model launches to improve engagement, intensifying advertising competition, regulatory litigation, and related expenses.
Key data
- RatingNeutralThe report maintains a neutral rating.
- Current price$669.21The price date is 2026-07-10.
- Target price$725.00The target price corresponds to December 2026.
- Implied upsideApproximately 8.3%Calculated based on the $725.00 target price and the $669.21 current price.
- 2027E GAAP EPS$34.19Valuation basis for the target price.
- 2027E Adj. EPS$43.28Disclosed in the key financial forecast table.
- 2026E capital expenditures$142BThe report forecasts 104% year-over-year growth.
- 2027E capital expenditures$202BThe report forecasts 42% year-over-year growth and believes the figure could still be revised upward.
- 2026E computing power plan7GWMeta's 2026 computing power target reported by Reuters.
- 2027E computing power plan14GWReuters reported that computing capacity could double in 2027.
- Meta Model API pricingApproximately 25% of the cost of leading AI modelsBased on Bloomberg interview information cited in the report.
- User baseApproximately 4B usersThe report believes Meta can deploy core AI products across its vast user base.
Impact & implications
The implication for investment judgment is that Meta's AI narrative is gradually shifting from a pure cost center toward a potential source of revenue and improved ROI. If Muse Spark 1.1 and subsequent Watermelon models gain adoption in internal products, the developer ecosystem, and enterprise customers, the market may become more willing to accept higher capital expenditures and reassess Meta's AI monetization capabilities. However, if capital expenditures continue to rise while commercialization signals remain insufficient, declining free cash flow and earnings pressure could offset the improvement in the AI narrative.
Risks
- The large scale of AI investment and delayed monetization could depress GAAP OI, GAAP EPS, and free cash flow.
- Revenue growth below expectations would limit earnings upside from AI monetization.
- If AI products and model launches fail to gain user, developer, or enterprise acceptance, they could weigh on engagement, revenue, and margins.
- Google, TikTok, OpenAI, and other online advertising companies pose competition for advertising budgets and the AI product ecosystem.
- Regulatory litigation and related expenses could create additional uncertainty.
- 2027 capital expenditures could continue to be revised upward; without corresponding monetization signals, market acceptance could decline.
What to watch
- The performance, release cadence, and real-world product deployment of Muse Spark 1.1 and subsequent Watermelon models.
- Developer adoption, enterprise customer conversion, and revenue scale of the Meta Model API.
- Whether Meta uses Muse Spark 1.1 and subsequent models more deeply in internal advertising, recommendation, and FOA AI products.
- Whether external AI monetization can demonstrate higher steady-state operating margins.
- Whether 2027 capital expenditures continue to rise and whether the market accepts higher investment due to improved monetization.
- Changes in Meta's LLM market share, enterprise trust, and developer ecosystem.
- Whether potential discussions of equity financing intensify, as well as the market's reaction to the use of proceeds and AI investment posture.