META — Meta Platforms, Inc.

As of 2026-09-13 the tape and the sell side are aligned bullishly: the stock is +11.95% over one month and +14.3% over three months, RSI(14) is 67.18, and price sits 10.5% above its 20-day SMA. The catalyst is company-specific, not macro: the September 2026 'Muse' agent/model launch convinced several prior bears that Meta's AI spend is producing sellable product, JPMorgan upgraded back to Overweight and raised its target from $640 to $820 on 2026-09-10, and Pershing Square (Ackman) added 3M+ shares in early September. Note the stock is still below its 52-week high of $790.80 and only +3.98% above its 200-day SMA, so this is a sentiment rotation and a recovery rally off a summer drawdown, not a fresh breakout to record highs. The bull case rests on narrative and optionality right now, not on reported earnings — Q2 2026 EPS actually fell 13.4% YoY.

Revenue growth has been genuinely accelerating and is now cooling slightly off a peak: YoY growth ran +16.1% (Q1 2025), +21.6% (Q2 2025), +26.3% (Q3 2025), +23.8% (Q4 2025), +33.1% (Q1 2026), and +28.0% (Q2 2026, $60.80B vs $47.52B). That is a $34.1B-to-$60.8B quarterly revenue ramp over 12 quarters. Margins are the problem, and it is structural, not a blip: operating margin peaked at 48.3% in Q4 2024, held at 40-43% through Q1 2026, then dropped to 30.9% in Q2 2026 — a 12.1pp YoY and 9.7pp sequential compression driven by R&D jumping to $21.66B (+67.3% YoY) from $12.94B a year earlier. The bottom line is being distorted by two one-offs in the TTM window: Q3 2025 EPS of $1.05 (net income of $2.71B against operating income of $20.54B implies roughly an $18B below-the-line hit) and Q1 2026 EPS of $10.44 (net income of $26.77B actually EXCEEDED operating income of $22.87B, a below-the-line gain). Strip both and normalized quarterly EPS runs in the $6.20-$8.90 band. Free cash flow has inflected hard negative-trending: $1.75B in Q2 2026 versus $9.02B in Q2 2025 and $13.23B in Q1 2026, as capex hit $30.12B (+82% YoY) in one quarter — TTM FCF is roughly $41.0B for a 2.5% FCF yield. Read the trend this way: top line accelerating, bottom line and cash flow decelerating sharply, and the gap is entirely the AI build.

Fair, with a mild expensive tilt on trailing numbers and a reasonable look on normalized ones. The headline P/E of 24.08x as of 2026-09-13 sits on a distorted TTM EPS of $26.779 — normalize the Q3 2025 tax/charge quarter and TTM EPS is closer to $29.1, which puts the stock at ~22.3x, essentially dead-on Meta's own 5-year average P/E of 22.52x. Forward P/E on the next-FY EPS estimate of $31.384 is 20.65x, which is not demanding for a company growing revenue 28% YoY, but is also not cheap given that EPS growth is now being throttled by AI opex and depreciation. P/S of 7.23x is only ~4% above the 5-year average of 6.97x. The PEG ratio of -4.937 is meaningless here — it is a negative number produced by a one-off quarter, not a signal. EV/EBITDA of 16.4x is unremarkable for the growth rate. Deterministic valuation_context labels this "Neutral vs. its own history" (score 43.3), and I agree. Nothing about this price is a bargain, but nothing about it is a bubble either — you are paying roughly the historical average multiple for a better-than-average grower with worse-than-average cash conversion.

Right now the price is being supported by narrative, not by the earnings print — and you should trade it that way. Q2 2026 revenue grew 28.0% while diluted EPS fell 13.4% ($6.18 vs $7.14), and full-year capex guidance of $130-145B against roughly $250B of revenue means the earnings line stays suppressed for at least several more quarters. So at $648.03 the market is capitalizing AI optionality (Muse agents, the 'Watermelon' frontier model, possible cloud/IaaS commercialization), not the current EPS stream — which is exactly why JPMorgan's target spiked to $820 on 2026-09-10 despite weak Q2 earnings. The actionable read: if Muse monetization starts showing up as visible revenue lines in Q3/Q4 2026 or management trims the top of the capex range in late October, the forward EPS estimate gets revised up and 20.6x starts to look cheap. If instead capex stays at $130-145B with no offsetting AI revenue, the multiple compresses toward 18x forward on a stagnant EPS number, which implies roughly $565. Watch the Q3 2026 print for exactly two things: the capex range and any new non-advertising revenue disclosure. Those two data points move this stock more than the headline revenue beat.

As of 2026-09-13, the technical setup is strong-but-extended. RSI(14) stood at 67.18 — high momentum, approaching but not yet at the classic 70 overbought threshold. Price is 10.5% above the 20-day SMA (implied level roughly $586), 7.84% above the 50-day (roughly $601), and 3.98% above the 200-day (roughly $623). Returns: +5.07% over one week, +11.95% over one month, +14.3% over three months, but only +5.59% over six months — meaning the last month did most of the work and the 6-month picture is far less exciting than the 1-month picture. At $648.03 the stock sits almost exactly mid-range within its 52-week band of $520.26-$790.80, so there is real room to the old high, but the double-digit gap above the 20-day SMA means chasing here is expensive. The prudent entry-timing implication: let it come back toward the 20-day/SMA50 zone (~$590-600) before adding; buying 10% above the short-term average right after a 12% month is how you get shaken out on the first capex headline.

• AI capex arms race: Meta's 2026 capex budget of $130-145B is the single largest swing factor in the story. Spend of this magnitude relative to a ~$250B revenue base invites direct comparison to the hyperscalers on ROI discipline, and every competitor's capex guidance now reads across to Meta's multiple.
• Interest rates and mega-cap duration: rates stabilizing removes a headwind but also removes a tailwind — there is no rate-cut catalyst to lean on. Higher-for-longer rates matter more for Meta now than in 2023-2024 because it is funding the build with debt, not just cash flow.
• Balance-sheet-financed buildout: long-term debt went from $28.83B (Q3 2025) to $58.74B (Q4 2025) to $83.66B (Q2 2026), while cash dropped from $35.87B to $15.46B. Net debt is roughly $68B. Leverage of this speed is a new macro-sensitivity for a name that was net-cash not long ago.
• Regulatory and antitrust overhang: continuing global antitrust scrutiny plus structural compliance frameworks (EU DSA/DMA and the state-AG settlements on platform governance and youth engagement) act as a persistent, non-diversifiable operating cost and headline risk.
• Advertising cyclicality and sector rotation: Meta is still overwhelmingly an ad business. A consumer-spending slowdown hits it directly, and within tech, capital is rotating between 'AI-infrastructure' and 'AI-application' buckets — Meta can get whipsawed by which bucket is in favor in a given month.
• Competitive ad-share pressure: TikTok, YouTube Shorts, and AI-native feed products compete directly for the same engagement minutes and ad dollars that underwrite the entire AI investment.

• Q3 2026 earnings (expected late October / early November 2026): management previously guided Q3 revenue to $61-64B. The two numbers that matter are any refinement of the $130-145B full-year capex range and any new disclosure on AI/agent revenue.
• Muse ecosystem commercialization: continued rollouts of Muse Spark 1.3, the 'Watermelon' frontier model, and Muse agent integration into Instagram and Facebook. This is the narrative that drove the September re-rating and would drive further multiple expansion if it produces billable product.
• Cloud / infrastructure-as-a-service commercialization: reports of Meta exploring direct cloud offerings would reframe the capex from pure cost to revenue-generating asset and is the single most powerful potential re-rating catalyst in the file.
• Continued institutional accumulation: Pershing Square added 3M+ shares in early September 2026 — follow-through buying from other large holders validates the AI-monetization thesis and supports the tape.
• Further sell-side target revisions post-Muse: JPMorgan at $820 and Morgan Stanley at $815 versus a consensus of ~$730-736 leaves room for the consensus to migrate upward if Q3 confirms the narrative.

• AI capex with no matching revenue: Q2 2026 free cash flow was $1.75B on $60.8B of revenue, down 80.6% YoY, with capex of $30.12B in a single quarter. If the $130-145B annual budget persists into 2027 without visible AI revenue, FCF stays structurally depressed and the market's patience gets tested.
• Operating margin compression is real and accelerating: Q2 2026 operating margin of 30.9% versus 43.0% a year earlier, with R&D up 67.3% YoY to $21.66B. Continued decline forces forward EPS estimate cuts, which mechanically compresses the multiple.
• Debt-funded expansion: long-term debt quadrupled from $28.83B to $83.66B in four quarters while cash fell to $15.46B. Debt-to-equity of 0.43 looks fine today, but the trajectory is the risk, not the level.
• Regulatory, antitrust, and youth-engagement litigation: ongoing global scrutiny and compliance obligations create unpredictable legal and operational costs and headline risk that a trailing financial formula cannot price.
• Valuation sensitivity to a narrative reversal: at 20.65x forward EPS on a business whose EPS just declined 13.4% YoY, the multiple is being earned by the AI story. If Muse underdelivers or delivers late, there is no earnings cushion to catch the stock — an 18x forward multiple on a flat EPS number is roughly $565.
• Advertising demand cyclicality and competitive share loss: the ad engine funds everything, and it faces direct competition from TikTok, YouTube, and AI-native feed formats.

Meta is still a genuinely high-quality franchise — 81.7% gross margin, 29.8% net margin, ROIC of 18%, a fortress current ratio of 2.23 — and revenue growth of 28.0% YoY at a $250B run-rate is exceptional for a company this size. But be honest about what changed: this is no longer the 2023-2024 story of margin expansion and buybacks. Q2 2026 showed revenue +28% while EPS fell 13.4% and free cash flow fell 80.6% to $1.75B, as capex hit $30.1B in one quarter and R&D jumped 67% YoY. You are now buying the ad franchise plus a very expensive, unproven AI option, funded increasingly with debt (long-term debt $28.8B to $83.7B in four quarters). At $648.03 you are paying roughly the historical average multiple (20.65x forward, ~22x normalized TTM) for that mix. I think the risk/reward is constructive but not compelling after a 12% one-month run — my calculated fair value of ~$722 implies only about 11% upside from here, and the entry is materially better on a pullback toward the $590-600 short-term moving-average zone than at today's price 10.5% above the 20-day. Own it for the ads engine and the Muse optionality; size it knowing the capex line, not the revenue line, is what will set the multiple for the next several quarters. This is analysis, not personalized financial advice.

Method: forward P/E applied to the consensus next-FY EPS estimate, cross-checked against P/S and EV/EBITDA. Key inputs: forward EPS of $31.384 (FMP analyst estimate), multiplied by a 23.0x multiple — roughly a 2% premium to Meta's own 5-year average trailing P/E of 22.52x, justified by accelerating revenue growth (+28.0% YoY in Q2 2026), an 81.7% gross margin, and the unquantified Muse/AI-monetization option, offset by the fact that capex intensity ($130-145B in 2026) and FCF conversion have deteriorated materially, which is why I did not award a full mega-cap-tech growth premium. 23.0 x $31.384 = $721.83, rounded to $722. Cross-check: applying the 5-year average P/S of 6.97 to FY2026 revenue of ~$250B and ~2.55B diluted shares gives ~$683/share, and the current 16.4x EV/EBITDA on a $1,719B enterprise value is unremarkable for the growth rate — so the P/E-based figure is the anchor and sits comfortably between the consensus target of $729.93 and my own lower bound, implying roughly 11% upside from the 2026-09-13 price of $648.03.