Meta may ditch the idea of a cloud business for Muse. We think that’s a good trade

Wells Fargo raised its Meta price target to $1,000 from $796, implying nearly 35% upside, as it sees Muse’s potential as a larger long-term opportunity than reselling unused compute capacity. The analysts removed about $5 billion of expected 2027 high-margin capacity-resale revenue, cut 2027 EPS to $31.18, and forecast operating expenses of roughly $212 billion versus $202 billion consensus; they expect EPS to rise to $40.02 in 2028 as AI products contribute more meaningfully. Meta shares gained 27% in September, while Wells Fargo estimates third-quarter revenue of about $64 billion, up 25% year over year.

Analysis

The key re-rating question is whether Meta can turn agent usage into incremental economics—not simply move compute from resale into another high-cost workload. Muse could strengthen ad conversion and create enterprise or transaction revenue, but task completion may also reduce time spent in Meta’s ad-supported surfaces. Adoption alone therefore is not proof of value; watch for paid usage, advertiser ROI, retention, and evidence that agent activity adds rather than cannibalizes monetizable engagement.

Near term, rising infrastructure expense against delayed product revenue leaves estimates vulnerable to further cuts and makes the recent optimism sensitive to execution. Over 1–3 months, disclosures on agent usage and enterprise uptake matter more than launch announcements. Over 6–18 months, the bull case requires monetization to outrun incremental inference, power, and depreciation costs. If it works, owning compute could capture more upside than renting it; if it does not, Meta gives up resale revenue while carrying the cost and utilization risk. Neocloud competition may also cap the strategic value of a standalone compute-rental alternative.

The analyst upgrade is not independent evidence of product economics, and the long-dated earnings case should not be treated as near-term consensus support. Separately, MongoDB should verify succession and customer-execution plans following its CEO’s reported move; the article provides no basis to quantify financial impact.

Key Decisions for Investors

• META: favor staged exposure on weakness rather than chasing a sharp rerating. Reassess after the next results for ad-revenue durability, operating-expense trajectory, and measurable agent monetization; avoid treating downloads or usage claims as revenue evidence.
• Falsifier for the META thesis: further expense or capex escalation without improving revenue guidance, agent retention/paid conversion, or advertiser outcomes. That combination would imply compute is consuming earnings rather than financing a second growth engine.
• Do not trade a compute-resale short solely on this thesis: the article does not identify a listed direct comparable or establish that Meta would have been a material supplier. Revisit only with verified capacity, utilization, and rental-pricing data.
• MDB: place on watch for CEO succession, customer continuity, and any changes to guidance; the reported leadership move is a possible execution and sentiment overhang, not enough by itself to justify a short.