- Meta has launched $35 billion of equity-related financing, with a potential $60 billion nonvoting Class C ATM program from 2027; the $95 billion capacity is not fully raised.
- AI spending is consuming cash: 3Q26 revenue rose 26% to $64.6 billion, but capex and lease payments exceeded operating cash flow by $2.12 billion; 2026 capex is guided to $135–145 billion.
- Meta’s distribution—3.64 billion daily users, advertising, commerce and messaging—could monetize AI agents through conversion and transactions rather than subscriptions; 18x forward P/E provides valuation support.
- The author remains long META, but returns depend on incremental ROIC and utilization; dilution, fixed infrastructure commitments, obsolete hardware and weak monetization could separate revenue growth from per-share value.
Investment pitches
filter:
89 69 - MSCI is consulting on moving Indonesia directly from Emerging Market to Standalone Market, bypassing Frontier; the author sees this as more damaging than the previously discussed downgrade.
- The bearish case rests on opaque ownership, questionable free-float quality, suspected coordinated trading and unreliable price discovery—not valuation; new rules have not proved implementation or enforcement.
- Standalone status could remove Indonesia from a broader universe of benchmark-driven institutional portfolios, pressuring equities, especially stocks with substantial foreign institutional ownership.
- This remains a consultation, not a decision; MSCI is expected to announce by end-June 2027, while existing restrictions on Indonesian securities remain.
85 - Technology represents roughly 55% of U.S. capital spending versus 15% in 1960, nearly half of S&P 500 profits, and about three-quarters of earnings growth.
- AI spending links compute to servers, data centers, power, construction and commodities; cash-rich hyperscalers can sustain this investment despite high rates, creating a broader industrial multiplier.
- The author expects AI’s later phase to expand margins through cheaper knowledge work, better monetization and productivity—not only direct AI revenue.
- The author remains long technology, but warns that capex outrunning useful AI demand could trigger a broader downturn; GPU utilization, backlog, inference and infrastructure returns are key monitors.
74 - Meta’s consumer-AI thesis is that Muse faces a less crowded market while OpenAI, Microsoft, Google and xAI prioritize enterprise and professional agents.
- Distributed through WhatsApp, Instagram, Facebook and Messenger, Muse could reach billions of existing users at moments of intent, building context, personalization, frequency and switching costs.
- Meta need not monetize a $20/month subscription alone: subsidized usage could support advertising, commerce, business messaging and premium tiers as Muse connects attention to transactions.
- The author is long $META; upside depends on strong product reception, aggressive distribution and consumer agents evolving from chatbots into tools that act continuously on users’ behalf.