Summary

Buy, $774.27 target, 20.7% upside: AI ad pricing outpaces 2027 depreciation

InitiationBy Penmark Team
METABuy$774.27+21%
Page 1 of 28
Equity Research | Initiation of coverage | United States | 11 November 2026, 07:30 ET
Communication Services / Interactive Media & Services | Coverage: Initiated (Low conviction)
Penmark Team | research@penmarkresearch.com
Meta Platforms, Inc. (NASDAQ: META)

Buy, $774.27 target, 20.7% upside: AI ad pricing outpaces 2027 depreciation

Rating
Initiation · Low conviction
Buy
12M target
$774.27
Price (9 Nov 26)
$641.37
Price return*
+20.7%
Total return
+21.1%
Bull case (12M)
$1,066.85
Bear case (12M)
$496.39
*Rating = price return to target, excl. dividends: Buy above +10%, Hold −10% to +10%, Sell below −10%. Total return is for information.

We initiate on Meta Platforms at Buy (Low conviction) with a 12-month target of $774.27 (+20.7% price return; +21.1% total return). Meta monetises 3.64bn daily active people across Facebook, Instagram, WhatsApp and Threads via AI-ranked ads. At $641.37 (21.8x published forward P/E; 20.3x on our FY27E EPS), the stock double-discounts AI capex.

  • AI ranking sustains 21.0% FY27E ad growth, 8.5pp above the 12.5% price-implied rate. GEM and Andromeda lifted 3Q26 ad pricing +14.0% on +10.0% impressions, backing 22.94x peer P/E (+$75.47/sh vs price).
  • Assumed FY27E capex of $234.9bn (75.0% of sales) places EPS 7.0% below Street, not P/E. D&A ($47.6bn), cloud/serving costs (36.5%) and +2.8% dilution put FY27E EPS at $31.58 vs $33.97 Street (−$54.92/sh).
  • 4Q26E–FY28E cash needs (~$195bn: −$133.2bn FCF, $47.0bn RSU taxes, $15.1bn divs) still require ~$35bn of FY28E notes. Cash draw ($42.4bn), closed 4Q26 raises ($29.4bn), fwd/ATM equity ($88.7bn; +3.6% FY27E shs) and $34.9bn FY28E notes hold a $40bn floor.

What the market is missing. Consensus $312.6bn FY27E sales (+0.2% gap) and $33.97 EPS imply ~34% EBIT margin, under-modelling D&A, cloud costs and dilution (ours 32.5%, $31.58 EPS), while equity prices 20.55x.

Catalysts. Class C 1-for-1 dividend (~19 Jan 2027); 4Q26 & FY27 capex guide (late Jan 2027 est., +$4.80 / −$11.90/sh); AG trials (1H27); Muse & Meta One (+$8.20/sh).

Valuation and target. Peer-median 22.94x on $31.58 FY27E EPS less $7.51/sh legal accruals = $716.84 base; Bear ($448.05, 25%) / Base ($716.84, 55%) / Bull ($962.94, 20%) = $698.86, rolling 12M to $774.27 (DCF $502.26 on negative FCF; Low conviction: +1pt rate or 1.57 beta = Hold).

Risks. Policy slowing FY27E ad pricing to +6.5% (−$47.20/sh), FoA cash costs at 38.5% (−$45.00/sh), or FY27E capex above $250bn (−$11.90/sh).

Exhibit 1: Revenue compounds 22% to $368.4bn by FY28E; META trades at 20.3x on our FY27E EPS
Dec YE, $mFY24AFY25AFY26EFY27EFY28E
Revenue164,501200,966257,086313,139368,416
Growth y/y21.9%22.2%27.9%21.8%17.7%
EBITDA84,878101,892108,834149,431184,169
Margin51.6%50.7%42.3%47.7%50.0%
Net income62,36060,45874,44084,29591,517
Adj. EPS ($)*23.8623.4933.4231.5833.69
Growth y/y*60.4%−1.5%42.3%−5.5%6.7%
Free cash flow52,10343,585−1,421−80,555−40,151
At $641.37, 9 NovFY24AFY25AFY26EFY27EFY28E
P/E (adj.)26.9x27.3x19.2x20.3x19.0x
EV/EBITDA†19.3x16.0x15.0x10.9x8.9x
P/B9.2x7.6x5.3x3.7x3.0x
FCF yield3.1%2.6%−0.1%−4.7%−2.3%
Dividend yield0.3%0.3%0.3%0.3%0.4%
ROE (adj.)37.1%30.2%32.4%21.9%17.5%
Net debt / EBITDA−0.6x−0.2x0.0x0.4x0.5x
P/E at $774.27 target32.5x33.0x23.2x24.5x23.0x

Source: Company filings; Penmark Research model at 9 Nov 2026. *Adj. EPS ex $11.89bn FY26E AG charges (GAAP $28.89); FY26E includes a $5.0bn 1Q26 tax benefit (6.5% tax rate vs 16% FY27E). †Model EV $1,634.0bn (2,646.2m dil. shs × $641.37 − $63.2bn net cash/bridge; 3Q26 basic EV $1,640.7bn).

Base value is $716.84, or $794.19 rolled 12M; weighted $774.27 target offers 2.94x upside-to-downside

Exhibit 3: Weighted scenarios roll 12 months to our $774.27 target (+20.7%)
$400$600$800$1000$1200Jan-26Jul-26Jan-27Jul-27ActualEstimatesPrice $641Bull $1067 (+66%)Base $794 (+24%)Bear $496 (−23%)Target $774 (+21%)

Source: Nasdaq closes to 9 Nov 2026; Penmark Research model (rolled 12M at 10.79% net of 0.34% yield).

12M rolled scenarios and price target basis
Bull$1,066.85+66.3%20% prob.

FY27E ad +24.9% (+11.0% imp., +12.5% price), FoA cash cost 35.0%, capex 68% ($220.9bn), 27.20x P/E ($962.94 at 9 Nov).

Base$794.19+23.8%55% prob.

FY27E ad +21.0% (+9.0% imp., +11.0% price), FoA cash cost 36.5%, capex 75%, 22.94x peer P/E ($716.84 at 9 Nov).

Bear$496.39−22.6%25% prob.

FY27E ad +14.0% (+7.0% imp., +6.5% price), FoA cost 38.5%, capex 88% ($258.7bn), $15bn legal, 18.10x P/E ($448.05 at 9 Nov).

Price target (prob.-weighted) $774.27 +20.7% price return = Buy
Bull up / Bear down 2.94x +$425.48 / −$144.98
Exhibit 4: EBIT hits $113.6bn by FY28E despite $70.5bn D&A
Income stmt, $mFY24AFY25AFY26EFY27EFY28E
Revenue164,501200,966257,086313,139368,416
FoA & RL opex79,62399,074148,252163,708184,248
EBITDA84,878101,892108,834149,431184,169
Deprec. & amort.15,49818,61627,78347,63870,519
EBIT (GAAP)69,38083,27681,051101,793113,650
EBIT margin42.2%41.4%31.5%32.5%30.8%
Net fin. items1,2832,656−1,400−1,442−3,388
Income tax8,30325,4745,21116,05618,744
Net income*62,36060,45874,44084,29591,517
Exhibit 5: $234.9bn FY27E capex drives −$80.6bn FCF trough
Cash flow, $mFY24AFY25AFY26EFY27EFY28E
EBITDA84,878101,892108,834149,431184,169
SBC & working cap.13,47036,72636,35522,36740,967
Tax & net int. paid−7,020−22,818−6,610−17,498−22,133
Operating CF91,328115,800138,579154,300203,003
Capex incl. leases−39,225−72,215−140,000−234,854−243,155
Free cash flow52,10343,585−1,421−80,555−40,151
Divs & buybacks−35,197−31,572−5,440−5,766−6,112
Eq., notes & RSU†−4,494−8,23618,72637,66341,464
Chg. in cash & sec.12,4123,77711,865−48,658−4,799
Exhibit 6: PP&E scales 3.7x to $659.0bn while cash holds $40.0bn
Balance sheet, $mFY24AFY25AFY26EFY27EFY28E
Cash & securities77,81581,59293,45744,79940,000
Receivables16,99419,76926,43227,45332,300
Net PP&E121,346176,400299,129486,345658,981
Goodwill & other59,89988,260120,449120,449120,449
Total assets276,054366,021539,467679,047851,729
Gross debt28,82658,74498,66998,669133,541
Net cash / (debt)48,98922,848−5,212−53,870−93,541
Total equity182,637217,243310,215459,287588,125
Exhibit 7: FY27E EPS is $31.58 on +11.0% ad price growth
Per share & KPIsFY24AFY25AFY26EFY27EFY28E
Diluted shares (m)2,6142,5742,5762,6702,716
EPS, adjusted ($)*23.8623.4933.4231.5833.69
EPS, GAAP ($)23.8623.4928.8931.5833.69
Common div. / sh ($)1.942.072.112.162.25
FCF / share ($)19.9316.93−0.55−30.17−14.78
Book value / sh ($)69.8784.40120.42172.04216.51
Ad impressions y/y11.0%12.0%—9.0%8.0%
Avg price / ad y/y10.0%9.0%—11.0%8.0%

Source: Company filings; Penmark Research model. *FY26E adj. EPS ($33.42) sums 9M26A ($24.98) + 4Q26E ($8.45) ex $11,890m legal charges (GAAP EPS $28.89); FY26E includes a $5.0bn 1Q26 tax benefit (6.5% tax rate vs 16% FY27E). †Equity/preferred & notes issued less RSU settlement taxes, preferred dividends and other cash/securities items.

EBITDA margin rebounds to 50.0% by FY28E; our FY27E EPS is 7.0% below Street on D&A

Exhibit 8: Sales grow 21.8% in FY27E as capex hits 75.0%
Growth & marginsFY24AFY25AFY26EFY27EFY28E
Growth y/y (%)
Sales21.9%22.2%27.9%21.8%17.7%
EBITDA46.5%20.0%6.8%37.3%23.2%
EBIT (GAAP)48.4%20.0%−2.7%25.6%11.6%
Adj. EPS*60.4%−1.5%42.3%−5.5%6.7%
Common div. / dil. shn/a6.7%1.9%2.4%4.2%
Margins & coverage
FoA cash cost / FoA rev.26.5%28.8%34.3%36.5%36.0%
EBITDA margin51.6%50.7%42.3%47.7%50.0%
EBIT margin42.2%41.4%31.5%32.5%30.8%
FCF margin31.7%21.7%−0.6%−25.7%−10.9%
EBITDA / gross int.93.2x51.6x33.2x27.3x28.9x
Capex / sales23.8%35.9%54.5%75.0%66.0%
Exhibit 9: ROE eases to 21.9% on equity-funded AI PP&E
Ratios & valuationFY24AFY25AFY26EFY27EFY28E
Returns & leverage
ROE (adj.)37.1%30.2%32.4%21.9%17.5%
ROA24.7%18.8%16.4%13.8%12.0%
ROCE36.2%34.2%23.7%21.1%17.8%
Net debt / equity−26.8%−10.5%1.7%11.7%15.9%
Net debt / EBITDA−0.58x−0.22x0.05x0.36x0.51x
Valuation at $641.37 (9 Nov 2026)
P/E (adj.)26.9x27.3x19.2x20.3x19.0x
P/E (GAAP)26.9x27.3x22.2x20.3x19.0x
EV/EBITDA†19.3x16.0x15.0x10.9x8.9x
P/B9.2x7.6x5.3x3.7x3.0x
FCF yield3.1%2.6%−0.1%−4.7%−2.3%
P/E at $774.27 target32.5x33.0x23.2x24.5x23.0x

Source: Company filings; Penmark Research model at $641.37 (9 Nov 2026); FY27E yield 0.34%. *FY26E includes a $5.0bn 1Q26 tax benefit (6.5% tax rate vs 16% FY27E). †Model EV $1,634.0bn (2,646.2m dil. shs × $641.37 − $63.2bn net cash/bridge; 3Q26 basic EV $1,640.7bn).

Exhibit 10: In line on FY27E sales (+0.2%), 7.0% below Street EPS
Penmark vs Street4Q26EFY26EFY27E
Revenue ($m)
Penmark estimate75,373257,086313,139
Consensus75,155256,876312,630
Prior-year period59,893200,966257,086
Diff. vs consensus+0.3%+0.1%+0.2%
Growth y/y (Penmark)+25.8%+27.9%+21.8%
Diluted EPS ($, GAAP basis vs Street)
Penmark estimate8.4528.8931.58
Consensus—29.1233.97
Prior-year period—23.4928.89
Diff. vs consensus—−0.8%−7.0%
Growth y/y (Penmark)—+23.0%+9.3%

Source: Yahoo Finance consensus (10 Nov 2026; 44–59 analysts); Penmark Research model. 4Q26E is the modelled quarterly revenue/EPS stub (guide $73.0bn–$76.0bn).

Exhibit 11: Adj. EPS grows from $14.87 in FY23A to $33.69 in FY28E*
Street GAAP EPS ($29.12 FY26E / $33.97 FY27E)$0$10$20$30$40FY23AFY24AFY25AFY26EFY27EFY28EActualEstimates$34

Source: Filings (A); Penmark Adj. EPS ex AG charges (E); Street GAAP EPS ticks (10 Nov 2026). *FY26E includes a $5.0bn tax benefit booked in 1Q26; the FY26E tax rate is 6.5% vs 16% in FY27E.

At 21.8x published forward P/E, 5.0% below the 22.94x peer median and 20.3x on our FY27E EPS, META discounts peak capex

Exhibit 12: META trades ~5% below the 22.94x peer median forward P/E
At 9 Nov 2026 pricesMETAPeer med.Diff.
Valuation multiples
Forward P/E (published)21.8x22.94x−5.0%
P/E on our FY27E EPS20.3x22.94x−11.5%
P/E at $774.27 target24.5x22.94x+6.9%
Trailing P/E21.9x24.3x−9.9%
Operating & scale benchmarks
Latest qtr sales y/y26.1%21.8%+4.3pp
Latest op. margin (adj.)*35.4%31.8%+3.6pp
Latest op. margin (GAAP)20.1%31.8%−11.7pp
Market value ($bn)1,639.53,206.5−48.9%

Source: Penmark Research model (6 Nov published forward EPS at 9 Nov close: GOOGL 18.10x, APP 20.83x, AMZN 25.05x, MSFT 27.20x; median 22.94x). *META 3Q26 ex $9.89bn AG charge (20.1% GAAP; 3-peer op. margin median).

Peer dispersion & multiple choice: Alphabet (18.10x) anchors our Bear floor and Microsoft (27.20x) anchors our Bull ceiling; using the four-peer median (22.94x, alongside AppLovin 20.83x and Amazon 25.05x) avoids cherry-picking higher-multiple cloud peers.
Exhibit 13: Four dated catalysts over the next 12 months
  • 19 Jan 27Class C 1-for-1 non-voting stock dividend effective; expands equity currency without diluting Class B control.
  • Late JanEst. 4Q26 results ($75.4bn vs $75.2bn Street) and formal FY27 capex/opex guide (+$4.80 bull / −$11.90/sh bear).
  • 1H 2027Youth & state AG benchmark trials; tests $19.9bn accrued legal reserve (−$4.90/sh for $15.3bn extra charges).
  • 2H 2027Muse commerce & Meta One scaling toward $7.4bn FY27E FoA Other sales (+75% y/y; +$8.20/sh bull).

Source: Company 8-K/10-Q filings; Penmark Research model.

Exhibit 14: Key rating risks (including upside risk vs Low conviction)
  • Ad pricing slowdown. Regulatory limits cut FY27E ad price growth to +6.5%. Med; −$47.20/sh.
  • FoA opex creep. AI compute pushes FoA cash cost ratio to 38.5%. Med; −$45.00/sh.
  • Runaway FY27E capex. Spend exceeds $258.7bn (88% of sales) or 10Y rates +1pt. Med; −$11.90 to −$121.00/sh.
  • Upside risk vs Low conviction. Ad growth hits +24.9% and capex holds 68% ($220.9bn), rerating to 27.2x. 20% prob.; +$246.10/sh.

Source: Penmark Research model scenario & sensitivity matrix.

Exhibit 15: TTM operating P/E averaged 23.3x (28.0x at 3Q26)
10x15x20x25x30x35x4Q213Q222Q231Q244Q243Q252Q26Mean 23.3x+1 SD 28.9x−1 SD 17.7x28.0x, +0.8 SD

Source: Company quarterly filings and Nasdaq quarter-end closes, 4Q21–3Q26; Penmark Research model (3Q26 $725.18 close; at $641.37 on 9 Nov, 21.8x published fwd P/E or 20.3x on our FY27E EPS).

Exhibit 16: At 21.8x published fwd P/E on 26.1% growth, META sits in 18.1x–27.2x peer range
17.5x20.0x22.5x25.0x27.5x10%20%30%40%50%60%Published forward P/ELatest-quarter revenue growth y/yMETAGOOGLAMZNMSFTAPP

Source: Penmark Research model (META and four valuation peers on 6 Nov published forward EPS priced at 9 Nov 2026 close; META is 20.3x on our FY27E EPS).

Thesis monitoring framework · KPI signposts and rating-change triggers (Hold band: −10.0% to +10.0%)
Monitoring KPIThesis confirmed ifBroken (cut to Hold) ifPer-share model sensitivity
Ad price growth (D1)≥ +10% y/y through 1H27< +6.5% y/y for two quarters±$19.43/sh per ±2pp (−$47.20 Bear)
FY27E capex guide (D2/D3)≤ $235bn (≤ 75% of sales)> $255bn without ad-yield lift+$4.80/sh at $221bn / −$11.90 at $259bn
FoA cash cost ratio (D2)≤ 36.5% of FoA revenue≥ 38.5% on serving/cloud opex∓$44.86/sh per ±2pp (−$45.00 Bear)
WACC / Equity betaWACC ≤ 10.80% (beta 1.25)+1pt rate or 1.57 beta$650.78 (+1.5%) or $598.85 (−6.6%) = Hold

What would change our view: At $641.37, any target below $705.51 (+10.0%) moves META to Hold; we would upgrade conviction if 4Q26 guidance caps FY27E capex near $220bn while ad pricing holds ≥ +11%.

PART 1 · SUMMARY · 3 THE THESIS IN CHARTS

Six exhibits trace the call from TTM EBIT, a 0.91 correlation, and +14% ad pricing to our $716.84 base value

Meta’s share price tracks trailing operating profit over multi-year cycles, while quarterly sell-offs track capex surprises. Our model puts FY27E EPS 7.0% below Street ($31.58 vs $33.97) on D&A, cloud costs and dilution yet still derives $716.84 base value ($774.27 12M target) at peer-median 22.94x P/E.

Exhibit 17: Share price reached $725 at 3Q26 on $79.4bn TTM EBIT
TTM EBIT ($bn)Quarter-end price ($) (right)$0bn$20bn$40bn$60bn$80bn$100bn$0$200$400$600$8004Q213Q222Q231Q244Q243Q252Q26

Source: Company quarterly filings & Nasdaq quarter-end closes, 4Q21–3Q26 (descriptive r = 0.913, n = 20; 9 Nov model price $641.37).

Exhibit 18: Ad price hit +14.0% in 3Q26 on +10.0% impressions
Price per ad y/yAd impressions y/y+5%+10%+15%+20%4Q241Q252Q253Q254Q251Q262Q263Q26+14%+10%
Thesis pillar 1: GEM (+9.6% app-ad conv.) and Andromeda (+1.4%) accelerated ad price growth from +6% (4Q25) to +14% (3Q26).

Source: Meta Form 10-K and 10-Q MD&A, 4Q24–3Q26.

Exhibit 19: Assumed capex peaks at $234.9bn–$243.2bn in FY27E–28E
$0bn$50bn$100bn$150bn$200bn$250bnFY23AFY24AFY25AFY26EFY27EFY28EFY29EFY30EActualEstimates$28bn$39bn$72bn$140bn$235bn$243bn$233bn$220bnFY27 Street $217.2bn

Source: Company filings (A); Penmark Research capex incl. leases (E, 75% FY27E assumption); Street FCF-implied $217.2bn (~30 Oct 2026).

Exhibit 20: FY27E EPS ($31.58) sits 7.0% below Street ($33.97)
ConsensusPenmarkFY26E GAAP EPSFY27E Diluted EPS$28.89$29.12$31.58$33.97
Thesis pillar 2: On a like-for-like GAAP basis, FY26E EPS is −0.8% vs Street ($28.89 vs $29.12; adj. ex-legal $33.42*) while FY27E EPS sits −7.0% below Street ($31.58 vs $33.97; Street moved from $34.92 to $34.03 over 90d, now $33.97) on D&A, cloud costs and dilution.

Source: Yahoo Finance consensus (10 Nov 2026); Penmark Research model. *FY26E includes a $5.0bn 1Q26 tax benefit (6.5% tax rate vs 16% FY27E).

Exhibit 21: Price has tracked the 20x–30x TTM operating EPS corridor
PricePrice at each multiple of TTM operating EPS$0$200$400$600$800$10004Q213Q222Q231Q244Q243Q252Q2615x20x25x30x$725 (28.0x)

Source: Company quarterly filings and Nasdaq closes to 30 Sep 2026; Penmark Research model.

Exhibit 22: D&A (−$34.6), dilution (−$20.3) & legal (−$7.5) bridge to $716.8
$0$200$400$600$800Street EPS x22.94xMargin &D&ADilution &interestLegalaccrualsBase value$779.3−$34.6−$20.3−$7.5$716.8

Source: Penmark Research model at 9 Nov 2026.

PART 1 · SUMMARY · 4 KEY DEBATES

Four debates pit +$75.47/sh of P/E upside vs price against −$54.92/sh of D&A/cost drag vs Street

Equity pricing ($641.37) and sell-side consensus ($833.12 average target) commit opposite errors on Meta’s AI infrastructure cycle. The share price implies 20.55x our FY27E EPS (21.8x published forward P/E), whereas consensus EPS under-counts 2027 D&A, cloud costs and dilution by $2.39/sh.

Exhibit 23: Four debates govern Meta’s 12-month path to our $774.27 target
Debate Market view Penmark view Primary evidence What changes our view Resolves
1. AI ad pricing & multiple (Ch. 10) Stock prices 20.55x FY27E EPS (12.5% implied ad growth) on negative FCF. 22.94x peer-median P/E on +21.0% FY27E ad growth (+$75.47/sh vs price). 3Q26 ad price +14.0%; GEM +9.6% app conversions; 10m GenAI ad tool users. Two quarters of ad price growth below +7.0% y/y or ad growth <14%. Jul 2027
2. FY27E D&A & cost step-up (Ch. 11) Consensus $33.97 FY27E EPS implies ~34% EBIT margin on $312.6bn sales. $31.58 FY27E EPS (32.5% EBIT margin) on D&A, cloud costs & dilution (−$54.92/sh). FY26E capex $140bn & 3Q26 opex +68%; 39 Street EPS cuts in 30 days. Server useful-life extension or FoA cash costs below 35.0% of revenue. Q1 2027
3. Peak capex & funding plan (Ch. 12) Bear estimates of $250bn+ after the 3Q26 print ($711.38/sh at market view). Assumed $234.9bn FY27E capex (75% rev; +$5.47/sh vs bear); ~$195bn 4Q26–28E need incl. $35bn FY28E notes. $454.5bn commitments ($121.4bn due 2027); $14.4bn pref & $15bn notes closed; $88.7bn fwd/ATM eq. FY27 capex guide >$255bn (>80% rev) or 10Y Treasury yield +100bps. Q1–Oct 27
4. Youth & state AG litigation (Sec. 17) Headline fear of unquantified multi-state settlements after $9.89bn 3Q26 charge. Deduct $19.88bn accrued legal liabilities ($7.51/sh) directly in base equity bridge. 3Q26 10-Q accrues $11.58bn current + $8.30bn non-current; payments capped ~$18.6bn. Adverse benchmark trial verdicts adding $15.3bn unaccrued charges (−$4.90/sh). 1H 2027

Source: Meta Form 10-Q/8-K filings; Yahoo Finance consensus (10 Nov 2026); Penmark Research model.

Exhibit 24: Peer P/E adds +$75/sh; D&A trims −$55/sh vs Street
Market viewPenmark base ($716.84)D1 Multiple on FY27E EPSD2 FY27E D&A & coststep-upD3 FY27E capex ($250bn+vs $235bn)D4 Youth & AG litigation$641$717$717$772$711$717$712$717

Source: Penmark Research model (9 Nov 2026).

How the debates map to the thesis

Each confirmed stock driver in Section 9 links cleanly to our four debates. Debate 1 maps to Chapter 10 (22.94x peer P/E on +21.0% FY27E ad growth, +$75.47/sh vs $641.37). Debate 2 maps to Chapter 11 ($47.6bn D&A, cloud costs & dilution, −$54.92/sh vs Street EPS). Debate 3 maps to Chapter 12 ($234.9bn FY27E capex vs $250bn+ bear views, +$5.47/sh, & ~$195bn 4Q26–28E funding incl. $35bn notes to issue), and Debate 4 to Section 17 (+$4.84/sh vs $15.3bn new charges).

Why both Street and market pricing miss

Sell-side targets ($833.12 mean) apply peer multiples to unadjusted margins, while $641.37 applies 20.3x to our FY27E EPS. Once formal FY27 guidance aligns Street EPS toward ~$31.6, clearing capex fears supports rerating to 22.94x.

Exhibit 25: Five milestones between January and October 2027 resolve all four debates
19 Jan 2027Class C dividend1-for-1 non-voting stocksplitQ1 2027 (est.)4Q26 print & FY27 guideResolves D1, D2 & D3capex1H 2027Youth & AG trialsTests D4 contingent legalJul 20272Q27 ad & D&A checkTests +11% ad pricing &D&AOct 20273Q27 & FY28 capex signalTests Muse revenue &capex peak

Source: Company Form 8-K and 10-Q filings; Penmark Research catalyst calendar.

PART 2 · THE BUSINESS · 5 THE STORY

3.64bn people pay nothing to use Meta every day, yet AI auctions turn each ad dollar into 41¢ of operating profit

Meta runs the world’s largest attention and commerce auction across Facebook, Instagram, WhatsApp and Threads. With $454.5bn of infrastructure commitments booked ($121.4bn due in 2027), we model $234.9bn of 2027 capex—three-quarters of sales—to rebuild around generative AI.

22 yearsFounded Feb 2004
3.64bnDaily active people, 3Q26
$201.0bnRevenue, FY25A (+22.2%)
66,775Headcount, 3Q26 (−15% y/y)
$1,639.5bnMarket cap, 9 Nov 2026

The scene

In 3Q26, nearly 10 million advertisers logged into Meta’s self-serve tools and let generative AI write copy, crop video or target buyers without an agency. Over the same weeks, more than 10 million US and Canadian users set up Muse, Meta’s personal AI assistant, where commerce emerged as the top use case at a >$200 average basket. Behind both sits a daily habit shared by 3.64bn people.

What it actually does

Meta sells predicted outcomes rather than thirty-second ad slots. Whenever a user scrolls Reels or Feed, neural ranking models—led by Meta’s Generative Ads Recommendation Model (GEM) and Andromeda retrieval engine—auction that slot across millions of merchants. The winner is the ad most likely to make that specific viewer buy, install or message.

How the money comes in

Advertising supplied $196.2bn (97.6%) of Meta’s $201.0bn FY25A revenue, paid per impression or conversion across uncapped self-serve budgets. FoA subscriptions and messaging added $2.6bn (1.3%) and Reality Labs devices added $2.2bn (1.1%). In FY25A, $1.00 of sales left 82.0¢ gross profit, 41.4¢ operating profit ($83.3bn) and 21.7¢ free cash flow ($43.6bn).

Where the story turns

After cutting staff in 2023 to lift operating margin above 41%, Mark Zuckerberg launched an unprecedented AI compute build. Capex including leases jumped from $39.2bn (FY24A) to $72.2bn (FY25A), $140.0bn (FY26E), and an assumed $234.9bn in FY27E—pushing FY27E FCF to −$80.6bn.

Why it works, and why it might not

Advertisers stay because AI conversion lifts (+9.6% app conversions on GEM; +9% spend for Business Assistant users) outpace ad price inflation. Capital intensity is the risk: if $234.9bn of GPUs fail to hold +11% FY27E ad pricing (D1) or capex tops $255bn (D3), D&A crushes returns.

+14.0%Average price per ad y/y, 3Q26
AI ranking turned 10% more ad impressions into 26% revenue growth because advertisers paid 14% more per ad.Source: Meta Q3 2026 Form 10-Q. Pairing +11.0% FY27E ad pricing with +9.0% impressions drives +21.0% ad growth, clearing the 12.5% price-implied hurdle (+$75.47/sh).
1 · DAILY HABIT
3.64bnDaily active people open Facebook, Instagram, WhatsApp or Threads (+3% y/y).
2 · AI AUCTION
~10mAdvertisers deploy GenAI creative & Advantage+ bidding at $0 upfront.
3 · QUARTERLY YIELD
$17.72FoA revenue per daily person in 3Q26 ($64.16bn FoA sales), up from $14.25.
4 · PROFIT KEPT
41.4¢FY25A operating profit per $1 of sales (32.5¢ in FY27E at peak D&A).

Exhibit 26: One customer dollar, start to finish (built from FY25A and 3Q26 reported disclosures). Source: Meta filings; Penmark Research.

Why the story matters for the stockInvestors at $641.37 treat Meta’s −$80.6bn FY27E free cash flow trough as permanent value destruction rather than a finite GPU build. Chapters 10–12 show why +14% ad pricing and a ~$195bn 4Q26E–FY28E funding plan (incl. ~$35bn of FY28E notes still to issue) preserve per-share value through peak depreciation.
PART 2 · THE BUSINESS · 5 HOW THE COMPANY MAKES MONEY

Family of Apps generated 98.9% of $201.0bn FY25A revenue and >100% of Meta’s $83.3bn operating profit

Exhibit 27: Six-stage business model map: 3.64bn daily users feed $196.2bn of high-margin AI auction revenue (FY25A)
1. Customers
Global SMBs & brandsSelf-serve performance & brand advertisers; zero customer >10% of sales.
10m+ active97.6% of FY25A sales
Daily consumersFree social, video & messaging habit; paid Meta One subscribers.
3.64bn DAP15m+ Meta One subs
Hardware buyersRay-Ban Meta AI glasses & Quest VR users.
1.1% shareConsumer devices
2. Offerings
Family of Apps (FoA)Instagram, Facebook, WhatsApp, Messenger & Threads video/feed surfaces.
$198.8bn98.9% of FY25A rev
Muse & Business AIPersonal AI commerce (>$200 basket) & merchant agents.
10m+ setups2027E growth leg
Reality Labs (RL)Quest VR & AI wearables.
$2.2bn1.1% of FY25A rev
3. Revenue
FoA AdvertisingAuction impressions (+12% FY25A) × avg. price per ad (+9% FY25A).
$196.2bn97.6% share (+22.1%)
FoA OtherWhatsApp paid messaging & Meta One subscriptions.
$2.6bn1.3% share (+54%)
RL Hardware/Soft.Headsets & glasses.
$2.2bn1.1% share (+3%)
4. Channels
Self-serve AI Ads ManagerAdvantage+ & GenAI creative automation; instant global bidding.
>85% directZero agency lock-in
Partnership & CreatorCreator-backed ads growing >97% y/y in 3Q26.
$15bn+ run-rateHigh-conv. video
Retail & EssilorLuxotticaOptical & online stores.
Global retailWearables channel
5. Cost to serve
Cost of revenueData-centre energy, D&A ($18.6bn) & RL hardware COGS.
$36.2bn18.0% of FY25A rev
R&D (AI & RL)Foundation models, ranking engineers, SBC ($20.4bn total).
$57.4bn28.6% of FY25A rev
M&S and G&ASales, legal & policy.
$24.1bn12.0% of FY25A rev
6. What is left
FoA Operating Profit51.6% segment margin before RL investment.
$102.5bn123% of group EBIT
RL Operating LossAR/VR & glasses R&D drag.
−$19.2bn−9.6% of group rev
Group EBIT & FCF$83.3bn EBIT (41.4%); $43.6bn FCF after $72.2bn capex.
$83.3bn EBIT$43.6bn FY25A FCF
3.64bnFamily DAP (3Q26) · +3.0% y/y
+10.0%Ad impressions (3Q26) · +12% FY25A
+14.0%Avg price per ad (3Q26) · +9% FY25A
36.5%FoA cash cost / rev (FY27E) · +7.7pp vs FY25A
$234.9bnFY27E capex (75% rev) · +67.8% y/y

Source: Company filings; Penmark Research model. Figures are FY25A reported unless noted.

What Meta sells, and why 10m+ advertisers buy

Meta sells measurable customer acquisition across a unified 3.64bn-person daily social graph with zero single-customer concentration above 10%. Unlike open-web ad networks crippled by mobile identifier deprecation, Meta closes the loop inside first-party apps (Reels, Feed, Click-to-WhatsApp, Shops) using on-platform engagement and conversion APIs.

Pricing, operating leverage and cash conversion

Revenue equals Family of Apps ad impressions multiplied by auction-cleared price per ad, plus FoA subscriptions/messaging and Reality Labs hardware. Cash operating costs are largely fixed engineering payroll, while incremental cost is shifting from variable headcount (down 15% y/y in 3Q26 to 66,775) to capitalised GPU servers—turning FY25A’s $43.6bn FCF into −$80.6bn in FY27E despite $154.3bn of operating cash flow.

What this map shows for value Family of Apps earned $102.5bn of segment EBIT in FY25A (51.6% margin), masking a −$19.2bn Reality Labs drag. Guiding RL losses down from $18–19bn in FY26E to ~$15.0bn in FY27E cushions ~$3.4bn of next year’s AI depreciation step-up.
PART 2 · THE BUSINESS · 5 REVENUE FLOW, DRIVER TREE & GEOGRAPHY

Of each $1 of FY25A sales, 82.0¢ was gross profit and 41.4¢ EBIT; North America drives 45.1% of ad revenue

Exhibit 28: Advertising delivered $196.2bn of $201.0bn FY25A revenue, leaving $83.3bn operating profit
Advertising$196.2bnFoA Other$2.6bnReality Labs$2.2bnRevenue$201.0bnCost of revenue$36.2bnGross profit$164.8bnR&D$57.4bnM&S and G&A$24.1bnOperating profit$83.3bn

R&D ($57.4bn, 28.6% of sales) exceeds COGS ($36.2bn) and SG&A ($24.1bn) combined. As first-party Muse models cut content-processing cost >80%, cash opex ratios fall while server D&A rises inside COGS and R&D.

Source: Meta FY2025 Form 10-K ($bn).

Exhibit 29: Ad impressions (+9.0%) and price per ad (+11.0%) drive $303.3bn of FY27E ad sales
Ad impressions+9.0% FY27EAvg price per ad+11.0% FY27EFamily of Apps Ads$303.3bn (+21.0%)Meta One & WhatsApp15m+ subs/trialsMuse commerce10m+ setupsFoA Other$7.4bn (+75.0%)AI Smart GlassesMuse Dec 2026Quest VR-$15.0bn RL EBITReality Labs$2.4bn (+12.0%)Total Revenue$313.1bn (+21.8%)

Source: Penmark Research model; Q3 2026 earnings call transcript. Shaded nodes highlight primary Section 14 forecast drivers.

Exhibit 30: US & Canada drove 45.1% ($28.44bn) of 3Q26 ad sales
$63.06bn3Q26 ad revUS & Canada45%$28.44bnEurope23%$14.23bnAsia-Pacific18%$11.53bnRest of World14%$8.86bn

US & Canada ($28.44bn) and Europe ($14.23bn) supply 67.7% of ad revenue.

Source: Meta Q3 2026 Form 10-Q geographic KPIs.

Exhibit 31: DAP hit 3.64bn as FoA ARPP climbed to $17.72
Family DAP (bn)FoA ARPP ($) (right)0bn1bn2bn3bn4bn$12$14$16$184Q241Q252Q253Q254Q251Q262Q263Q26

Quarterly FoA ARPP rose 24.4% (+$3.47) from 4Q24 to 3Q26 as DAP expanded +8.7%.

Source: Meta Form 10-Q and earnings releases, 4Q24–3Q26.

Customer concentration & cash-vs-profit conversion No single advertiser accounts for ≥10% of revenue across >10m self-serve marketers, keeping working capital self-funding ($19.8bn FY25A receivables offset by accrued liabilities). Prior to the FY26E–28E GPU build, FY25A converted 52.3% of operating profit ($83.3bn) into free cash flow ($43.6bn after $72.2bn capex incl. leases).
PART 2 · THE BUSINESS · 6 THE MARKET

Global social advertising reaches $394.6bn in 2026E, up 21.3%, while Meta leads global digital ads at 26.8% share

Digital and social advertising are compounding nearly twice as fast as global all-media spend ($1.30tn–$1.34tn in 2026E). Generative AI creative tools and closed-loop conversion APIs are shifting performance budgets from open-web display and linear TV into first-party platforms.

Exhibit 32: Closed-loop integration across 3.64bn daily users, AI ranking and commerce captures margin
Users & Creators3.64bn daily usersReels, Feed, Threads,WhatsAppAd Creative10m+ AI tool usersAdvantage+ & GenAI toolsAI AuctionGEM & Andromeda+1.4% to +9.6% conv. liftCommerce & AIMuse & Business AI>$200 avg Muse orderAdvertisers10m+ SMBs & brandsZero >10% concentrationMeta direct platform monetisation ($201.0bn FY25A; 82.0% gross margin)

Source: Meta Q3 2026 Form 10-Q and earnings call transcript; Penmark Research.

Exhibit 33: Meta ($243.5bn) & Google ($239.5bn) lead 2026E digital ads
Meta PlatformsAlphabet (Google)AmazonOther digital platforms$243.5bn$239.5bn$82.1bn$343.4bn

Source: eMarketer 2026 global digital ad projections (26.8% Meta, 26.4% Google, 9.0% Amazon; top-3 = 62.3% share). Scope differs from GAAP ad revenue ($250.7bn Penmark FY26E).

Exhibit 34: Market benchmarks across ad definitions (2025–2026E)
Market benchmarkSizeGrowthMeta context
Global all-media (WARC 26E)$1,340bn+9.1%WPP ex-US pol. $1.30tn
Global social ads (WARC 26E)$394.6bn+21.3%Up from ~$325bn in 2025
Global digital ads (eMkt 26E)~$908.5bn*n/aMeta est. 26.8% share
US internet ads (IAB 2025A)$294.6bn+13.9%Secular shift from TV
US social media (IAB 2025A)$117.7bn+32.6%40.0% of US digital rev
Meta GAAP ad rev (FY26E)$250.7bn+27.8%FY27E $303.3bn (+21.0%)

Source: WARC, WPP Media, IAB/PwC, eMarketer (*implied from $243.46bn / 26.8%), Penmark model. Agency pools differ across China/retail media boundaries.

Bottom-up demand: units × price and cycle position

Social advertising is taking share because GenAI removes creative production bottlenecks for >10m small and mid-sized advertisers. In the US, IAB data show social revenue rose 32.6% to $117.7bn in 2025 (40.0% of internet ad spend), while WARC projects global social spend up 21.3% to $394.6bn in 2026. Meta’s +27.8% FY26E GAAP ad growth ($250.7bn vs $196.2bn FY25A) outpaces the broader market by >6pp.

Margin capture sits inside walled gardens, not open-web intermediaries. Because Meta owns consumer supply (Reels, Feed, Threads), creative automation (Advantage+), auction ranking (GEM/Andromeda) and messaging checkout, it bypasses third-party demand-side and supply-side take rates—earning an 82.0% FY25A gross margin.

Regulatory, AI search and substitution threats

Three structural forces could cap ad pricing power over 2027–2028: (1) European DMA / “consent-or-pay” rules restricting cross-app behavioural targeting across Europe’s $14.23bn quarterly ad pool; (2) US youth mental-health and state AG restrictions on teen engagement features; and (3) retail media networks (Amazon $82.1bn 2026E) and AI answer engines competing for bottom-funnel purchase intent.

How our model reflects market maturity We decelerate Meta’s ad growth from +27.8% in FY26E to +21.0% in FY27E (+9.0% impressions, +11.0% price) and +16.6% in FY28E (+8.0% impressions, +8.0% price), keeping Meta aligned with global social growth rather than assuming perpetual share gains.
PART 2 · THE BUSINESS · 7 COMPETITIVE POSITION

Meta outgrows Alphabet, 26.1% vs 24.7%, and trades at 21.8x forward P/E vs 22.94x peer median

Meta’s moat rests on three self-reinforcing advantages: a 3.64bn daily social graph, closed-loop SMB auction liquidity, and hyperscale AI ranking. Among platforms able to fund $100bn+ annual AI infrastructure, Meta pairs faster top-line growth than GOOGL, AMZN and MSFT with unmonetised messaging and Reels surfaces.

Exhibit 35: Meta scores 4 of 4 on daily audience reach, GenAI ad tools and self-serve SMB auction liquidity
METAGOOGLAMZNMSFTAPPDaily audience reachFirst-party intent/signalGenAI creative & rankingSMB self-serve auctionEnterprise cloud & searchBalance-sheet FCF bufferNoneWeakFairStrongLeading

Source: Penmark Research competitive assessment grounded in company filings and KPI disclosures across the four-company valuation peer set.

Exhibit 36: Competitive benchmark across the four-company AI & digital advertising peer set (plus reference social peers)
Company (Ticker) Primary overlap & business-model role Market cap ($bn) Latest rev y/y Op. margin Trail. P/E Fwd P/E
Meta Platforms (META)Subject: social, Reels, messaging & AI ads$1,639.526.1% (3Q26)35.4%*21.9x21.8x
Alphabet (GOOGL)YouTube video, Search intent & Gemini AI$3,538.024.7%31.8%15.0x18.1x
Amazon (AMZN)Closed-loop retail media ads & AWS cloud$2,875.012.1%13.1%21.5x25.1x
Microsoft (MSFT)Hyperscale AI capex, Copilot & LinkedIn ads$4,003.018.8%n/a28.4x27.2x
AppLovin (APP)AI mobile performance ad-ranking engine$135.855.5%77.8%27.1x20.8x
Peer median (4 peers)Equal-ranked valuation benchmark$3,206.521.8%31.8%24.3x22.94x
META vs peer medianPublished fwd P/E (memo: 20.3x on our FY27E EPS)−48.9%+4.3pp+3.6pp−9.9%−5.0%
Reddit (RDDT, ref. only)Excluded: fell 31% post-29 Oct print; no 9 Nov fwd P/E$19.553.0%29.4%19.6xn/a
Pinterest (PINS, ref. only)Excluded: no published forward P/E at 9 Nov price daten/an/an/an/an/a

Source: Penmark Research model (forward P/Es as published on 6 Nov and priced at 9 Nov 2026 close: GOOGL 18.10x, APP 20.83x, AMZN 25.05x, MSFT 27.20x; median 22.94x; statistics use four price-date peers only). *META 3Q26 op. margin ex $9.89bn AG charge (20.1% GAAP).

Why GOOGL, AMZN, MSFT and APP define the peer set

Only GOOGL, AMZN and MSFT match Meta’s $100bn+ annual AI data-centre commitment, while APP isolates pure-play AI ad-ranking economics. Smaller social platforms (Reddit at $19.5bn market cap, Pinterest at $19.06/sh) lack both hyperscale GPU budgets and published 9 November forward multiples. On the same 6 November published forward basis, Meta’s 21.8x P/E sits 5.0% below the 22.94x four-peer median despite faster revenue growth (+26.1% vs 21.8% median).

Durability of switching costs & unmonetised surfaces

Unlike GOOGL, AMZN and MSFT, Meta lacks an external enterprise cloud division (1/4 score), heightening free-cash-flow sensitivity during peak GPU capex. Offsetting that gap, nearly 10m advertisers use Meta’s generative creative tools, Partnership Ads exceed a $15bn run-rate (+97% y/y), and Threads plus WhatsApp business messaging expand auction supply organically without acquisition traffic costs.

PART 2 · THE BUSINESS · 8 MANAGEMENT, GOVERNANCE & CAPITAL ALLOCATION

Zuckerberg’s 60.8% vote enables a $234.9bn FY27E capex pivot while Class C preserves control

Meta is a founder-controlled company whose capital allocation has swung from $76.1bn of FY23A–25A buybacks to external infrastructure financing. Minority investors have zero voting check on capital intensity, making return on AI capex the central governance question.

Exhibit 37: Dual-class control, compensation structure and January 2027 Class C stock dividend
Governance dimension Reported structure & proxy disclosure Implication for minority shareholders
Founder voting control Mark Zuckerberg holds 639,347 Class A (1 vote) and 341,823,978 Class B shares (10 votes) = 99.8% of Class B and 60.8% total voting power. Unilateral approval of budgets, M&A, board slate and multi-year AI/Reality Labs spend without shareholder veto.
Class C non-voting dividend (~19 Jan 2027) Board & majority vote approved non-voting Class C stock and a 1-for-1 Class C stock dividend for each Class A/B share. Creates liquid non-voting equity for $60bn FY27E–28E ATM and RSU grants without diluting Zuckerberg’s 60.8% vote.
CEO pay & alignment 2025 total CEO pay $25.13m ($1 base salary, $0 bonus/equity awards, $25.13m personal security & aircraft). Direct ~$219bn equity stake aligns long-term wealth with share price, though time horizon exceeds 3–5 yr fund cycles.
Board & leadership bench Combined Chair/CEO role with Lead Independent Director; CFO Susan Li leads disciplined hybrid capital structure. Strong execution track record since 2023 efficiency reset (headcount 66,775 in 3Q26, −15% y/y), offset by key-person risk.
Capital shift (FY26E–28E) Buybacks paused since 4Q25 ($0 in 9M26A and $0 modelled FY26E–30E); $14.4bn mandatory preferred + $15.0bn senior notes issued Oct–Nov 2026. FY27E diluted shares are +3.6% vs FY26E (+2.8% vs consensus share count); by FY29E shares are +6.4% vs FY26E (2,740m).

Source: Meta 2026 DEF 14A Proxy Statement (filed 16 Apr 2026); Form 8-K filings (21 Oct, 29 Oct, 5 Nov 2026); Penmark model.

Exhibit 38: FY23A–25A CFO ($278.2bn): $139.5bn capex, $76.1bn buybacks
Capex incl. leasesShare repurchasesRetained FCF & cashCommon dividends$139.5bn$76.1bn$52.2bn$10.4bn

Source: Company filings; Penmark Research model (FY23A–FY25A CFO $278.2bn: $139.5bn capex incl. leases, $76.1bn buybacks, $10.4bn dividends).

Exhibit 39: Revenue beat 3Q26 top end ($64.6bn vs $64.0bn)
Actual or estimateGuidanceConsensus1Q26A rev2Q26A rev3Q26A rev4Q26E rev$56.3bn vs guide $53.5bn–$56.5bn$60.8bn vs guide $58.0bn–$61.0bn$64.6bn vs guide $61.0bn–$64.0bn$75.4bn vs guide $73.0bn–$76.0bn

Source: Q1–Q3 2026 earnings releases; Yahoo Finance 4Q26 consensus ($75.16bn); Penmark Research model.

Guidance record: consistent top-line beats, rising capex

Quarterly revenue landed near the top of guidance in 1Q26 ($56.3bn vs $53.5bn–$56.5bn) and 2Q26 ($60.8bn vs $58.0bn–$61.0bn) and beat the top end in 3Q26 ($64.60bn vs $61.0bn–$64.0bn), while management lifted FY26 capex guidance to $135bn–$145bn. Our $75.37bn 4Q26E sits near the top of the $73bn–$76bn range (~1% FX headwind), whereas four of the last five results windows coincided with −5.8pp to −10.3pp 3-day excess sell-offs alongside higher spend guides.

Controlled-issuer governance trade-off

History cuts both ways on Zuckerberg’s unilateral bets. Reality Labs has accumulated >$70bn of operating losses (still −$18bn to −$19bn in FY26E, narrowing to −$15.0bn in FY27E), yet the 2022–2023 pivot to AI Reels ranking and staff discipline doubled TTM EBIT from $36.8bn to $79.4bn while trimming 3Q26 headcount 15% y/y to 66,775.

Class C dividend mechanics (~19 January 2027 expected): All per-share prices, EPS estimates and targets in this report are stated pre-dividend ($641.37 price, $774.27 target); upon payment of the 1-for-1 non-voting Class C stock dividend, nominal per-share figures halve mechanically with zero impact on intrinsic value, percentage return (+20.7%) or Buy rating.
PART 3 · THE THESIS · 9 WHAT MOVES THE STOCK (1/2)

TTM EBIT drives 3-year returns of +$371.5/sh, while capex hikes coincided with four drops of −5.8pp to −10.3pp

Over three years (Nov 2023–Nov 2026), META doubled (+100.1% to $641.37) almost entirely on TTM EBIT expansion (+$371.54/sh), while multiple compression cost −$69.01/sh. Across four empirical tests, four company-specific drivers govern both earnings-day swings and medium-term re-rating.

Exhibit 40: Four confirmed stock drivers tested across event reactions, 3Y bridge, factor R² and Q&A share
Confirmed driver Tests passed Observed market reaction Penmark model value Debate Next number to watch
1. AI ad pricing & yield Event, 3Y bridge, Q&A (3/4) 4Q25 +10.5pp excess rally on +24% rev & +6%→+12% price acceleration ±2pp FY27E ad price = ±$19.4/sh (±2.7%); Bear +6.5% = −$47.2/sh D1 4Q26 ad price y/y (est. late Jan 2027; Penmark ~+12%)
2. AI capex & D&A path Event, 3Y multiple, Q&A (3/4) −5.8pp to −10.3pp 1-day excess drops when FY capex guide rose +$5.47/sh vs $250bn+ bear capex view; −$11.90/sh above $250bn D2, D3 Initial FY27 capex guide (est. late Jan 2027; Penmark $234.9bn)
3. Muse & FoA AI revenue Event, Call Q&A #1 (2/4) 55.4% of analyst Q&A turns; market assigns ~0x to AI assistants FoA Other +75% to $7.40bn FY27E ($23.19bn by FY30E) D3 2027 Muse commerce GMV & Meta One paid subs (>15m base)
4. Cost of equity / beta Factor R², Valuation (2/4) 57.6% weekly variance from Market + XLC; rate spikes compress P/E +1pt rate cuts base −$121.0/sh to $595.9 (wtd $582.1 / $650.8 tgt = Hold) D1, D3 Hybrid financing spreads & 10Y UST / sector beta regime

Source: Nasdaq daily/weekly closes, company filings/transcripts and Penmark Research model.

Exhibit 41: Event-day reactions across the last five earnings releases (close-to-close next trading session)
Release (Date) Primary operational & capex surprise vs expectations META 1D S&P 500 Excess
3Q25 (29 Oct 25 AMC)Ad rev beat, but management flagged materially higher FY26 AI capex growth−11.33%−0.99%−10.34pp
4Q25 (28 Jan 26 AMC)Strong 1Q26 rev guide ($53.5bn–$56.5bn) outweighed initial FY26 capex frame+10.40%−0.13%+10.53pp
1Q26 (29 Apr 26 AMC)Rev $56.3bn near top of guide (+12% price), but FY26 capex ceiling raised−8.55%+1.02%−9.57pp
2Q26 (29 Jul 26 AMC)Rev $60.8bn beat (+14% imps, +12% price), offset by steeper 2H26/FY27 infra commentary−7.95%+1.66%−9.61pp
3Q26 (21 Oct 26 AMC)Rev $64.6bn beat (+14% price), FY26 capex lifted to $135bn–$145bn + $9.89bn AG charge−6.98%−1.21%−5.77pp

Source: Nasdaq unadjusted daily closes around Form 8-K earnings timestamps; Penmark Research.

Exhibit 42: Capex hikes coincided with 4 of 5 negative excess moves
3Q25 (30 Oct 25)4Q25 (29 Jan 26)1Q26 (30 Apr 26)2Q26 (30 Jul 26)3Q26 (22 Oct 26)−10.3pp+10.5pp−9.6pp−9.6pp−5.8pp

Source: Nasdaq and S&P Dow Jones Indices next-day excess return vs S&P 500. Notice diminishing sell-off (−10.3pp → −5.8pp) by 3Q26 as $140bn capex entered consensus.

Exhibit 43: 3Y bridge: +$371.5 EBIT growth vs −$69.0 de-rating
$0$200$400$600$800Price 9 Nov23TTM EBITgrowthSharereductionMultiplechangePrice 9 Nov26$320.6+$371.5+$18.3−$69.0$641.4

Source: Penmark decomposition (9 Nov 23 $320.55 → 9 Nov 26 $641.37; TTM EBIT $36.77bn → $79.38bn; P/TTM EBIT/sh 23.03x → 20.79x). Excludes ~$6.1/sh cumulative dividends.

PART 3 · THE THESIS · 9 WHAT MOVES THE STOCK (2/2)

Idiosyncratic variance is 42.4% weekly; Muse/AI at 55.4% and capex at 32.3% top multi-coded Q&A

Exhibit 44: Market + sector explain 57.6% of weekly variance; 42.4% META
Sector (XLC)Market + SectorCompany idiosyncraticMarket (S&P 500)55.9%57.6%42.4%27.4%

Source: Penmark OLS regression of weekly log returns (n = 262 weeks, Nov 2021–9 Nov 2026) on S&P 500 and XLC closes. Note: META (~22% XLC weight) inflates standalone XLC R².

Exhibit 45: Muse/AI (55.4%) & capex (32.3%) led 65 analyst Q&A turns
Muse & Meta AImonetisationAI capex, infra & financingEngagement, Reels &ThreadsAI ad ranking & pricingOperating expenses &marginsReality Labs & smartglassesMacro, China, EU & legal55.4%32.3%24.6%10.8%9.2%9.2%4.6%

Source: Penmark analyst topic coding across the last 8 quarterly earnings calls (n = 65 multi-part analyst Q&A turns; shares sum >100%).

Driver 1 · Ad pricing & yield beats set upside (Debate 1)

Claim: Sustained >10% ad price growth proves GPU capex earns an immediate cash return. Market prices 12.5% FY27E ad growth ($641.37), whereas our 21.0% build (+11% price, +9% imps) adds +$75.47/sh to our $716.84 base value (+11.8%). Watch: 4Q26 ad price y/y (est. late Jan 2027) — Penmark +12%; >+13% likely +4–6% move.

Driver 2 · Capex & D&A step-up compresses multiples (Debates 2 & 3)

Claim: Four of the last five prints (excluding 4Q25’s +10.5pp rally) coincided with −5.8pp to −10.3pp excess drops as capex guidance rose. Our $234.9bn FY27E capex puts EPS −7.0% below Street ($31.58 vs $33.97) yet leaves +20.7% target upside. Watch: FY27 capex guide (est. late Jan 2027) — Penmark $234.9bn vs $250bn+ bear views.

Driver 3 · Muse & Business AI shift from Q&A to revenue (Debate 3)

Claim: Analysts asked about Muse/Meta AI on 36 of 65 turns (55.4%) with confident management answers on 10x compute efficiency and >$200 Muse commerce orders, yet price assigns ~$0. We model +75% FoA Other to $7.40bn in FY27E ($23.19bn by FY30E). Watch: FY27 FoA Other run-rate ($1.5bn+ quarterly confirms +75% ramp).

Driver 4 · Discount-rate & beta shifts govern our rating (Debates 1 & 3)

Claim: Because FCF is negative through FY28E (−$80.6bn FY27E, −$40.2bn FY28E; +$6.6bn FY29E), equity duration is long. Raising discount rate +1pt cuts our rolled target −$123.5/sh to $650.8 (+1.5%, Hold); high-beta cuts it to $598.9 (−6.6%, Hold). Watch: Credit spreads on $15bn Nov 2026 notes (5.85%–7.50%) & FY27 ATM execution.
Exhibit 46: Candidate drivers tested and dropped (failed ≥2 empirical confirmation threshold)
Candidate tested Common market assumption Why empirical tests rejected it as a primary driver
China cross-border ad spendTariff/de-minimis shocks derail ad growthOnly 4.6% of Q&A turns (combined macro/China/EU); global auction re-allocated impressions to domestic & AI-assisted SMBs with 3Q26 ad price accelerating to +14%.
Reality Labs quarterly revenueHeadset unit swings move META sharesRL is 1.1% of FY25A sales ($2.2bn); stock reacts only to annual RL operating-loss cap (~−$15bn FY27 guide), folded into Driver 2/margin floor.
One-off legal accruals & tax$9.89bn 3Q26 AG charge drives valuationMarket looked through $3.83 GAAP vs $7.40 ex-charge EPS on 3Q26 print (−5.8pp move tracked capex guide); model deducts $19.88bn balance-sheet liability directly.
DAP headline user additionsMature user saturation caps revenueDAP grew steadily +3% to 3.64bn with low variance; ARPP (+22% y/y in 3Q26) and price/ad explain >85% of top-line variance.

Source: Penmark Research driver screening across company KPIs, 8 earnings transcripts and Nasdaq event windows.

Methodology & limits: Quarterly operating correlations (n = 20 quarters) and event windows (n = 5–8 prints) describe co-movement rather than causality; Q&A shares reflect Penmark’s internal coding of 65 analyst question turns across company earnings transcripts.

PART 3 · THE THESIS · 10 PILLAR 1: AI AD RANKING & PRICING YIELD

AI ranking drives +21.0% FY27E ad growth, from +11% price and +9% impressions, clearing the 12.5% implied bar by 8.5pp

1 · The claim (rests on Section 9 Driver 1 / Debate 1): Meta’s GPU-trained recommendation engines (GEM, Andromeda, Advantage+) structurally lift advertiser conversion rates, enabling +21.0% FY27E ad growth ($303.3bn) versus the 12.5% implied by $641.37.
Exhibit 47: Base FY27E ad growth (21.0%) clears 12.5% implied rate
10%15%20%25%30%FY23AFY24AFY25AFY26EFY27EActualEstimatesBull 24.9%Base 21.0%Bear 14.0%FY27 price-implied 12.5%

Source: Penmark Research model (FY23A–FY27E FoA ad revenue y/y growth; FY27E scenario rates labelled at right).

2 · What the market believes

Sell-side revenue consensus ($312.6bn FY27E, +21.7%) looks aligned with our $313.1bn (+21.8%), yet the share price ($641.37; 21.8x published forward P/E, 20.3x on our FY27E EPS) discounts a sharp slowdown. Holding peer multiple and base margins fixed, $641.37 solves for just 12.52% FY27E ad growth—an immediate 15.3pp deceleration from FY26E’s +27.8%. Bears argue Reels ad load is full and 2026 pricing comps (+12% to +14%) will roll over in 2027.

3 · What it is worth per share

Clearing the 12.5% implied hurdle at our 21.0% base ad forecast generates +$75.47/sh (+11.8%) of present value ($716.84 vs $641.37). In our sensitivity matrix, each ±2pp shift in FY27E ad price growth moves primary P/E value by ±$19.43 per share ($697.4 to $736.3, ±2.7%), while our Bear pricing path (+6.5%, −4.5pp) cuts value −$47.2/sh. Bull ad growth of 24.9% lifts value toward $962.94.

Exhibit 48: 4 · Three independent evidence streams supporting sustained double-digit ad pricing in FY27E
Evidence source Observed quantitative proof (period & filing) Why it supports +11.0% price & +9.0% impressions in FY27E
1. Reported KPI series (Form 10-Q) Ad price growth accelerated from +6% (4Q25) to +12% (1Q26), +12% (2Q26) and +14% (3Q26) while impressions grew +10% in 3Q26. Proves price acceleration is auction-conversion driven, not supply scarcity: impressions still expanded double digits simultaneously.
2. Model-level AI conversion lifts (3Q26 call) GEM expansion lifted offsite campaign conversions +1.7% and app-ad conversions +9.6%; Andromeda added +1.4%; Feed generative retrieval added ~+2%. In a second-price/VCG performance auction, higher conversion probability mechanically raises advertiser bids per impression without hurting ROAS.
3. Advertiser tool adoption & agency benchmarks Nearly 10m advertisers used GenAI creative tools in 3Q26; Business Assistant users raised ad spend +9%; Partnership Ads hit >$15bn run-rate (+97% y/y); IAB US social spend rose +32.6%. Automated creative and creator partnerships broaden auction density across millions of SMBs, insulating pricing from single-vertical pullbacks.

Source: Meta Q3 2026 Form 10-Q, Q3 2026 earnings transcript, and IAB/PwC Internet Advertising Revenue Report.

5a · Signpost confirming Pillar 1 (4Q26–2Q27) 4Q26 revenue prints ≥$75.0bn (vs $73bn–$76bn guide) in late Jan 2027, followed by 1Q27–2Q27 ad price growth holding ≥+9.5% y/y alongside ≥+8.0% impressions.
5b · Signpost disproving Pillar 1 (Bear trigger) Two consecutive quarters of ad price growth <+6.0% y/y or ad revenue growth <14% in 1H27, signalling diminishing returns on GPU ranking models and pulling value toward our $448.05 Bear case.
PART 3 · THE THESIS · 11 PILLAR 2: DEPRECIATION STEP-UP VS OPERATING MARGIN FLOOR

Consensus under-models FY27E D&A of $47.6bn, yet $31.58 EPS, 7.0% below Street, still backs $774.27

1 · The claim (rests on Section 9 Driver 2 / Debate 2): Even after penalising FY27E EPS −7.0% below consensus ($31.58 vs $33.97) for a 2.6x surge in depreciation ($18.6bn FY25A → $47.6bn FY27E) and a +7.7pp step-up in FoA cash costs (36.5% vs 28.8% FY25A), 21.8% revenue growth and narrowing RL losses hold EBIT margin above 30.8% and support a Buy.
Exhibit 49: D&A hits $90.7bn by FY29E; EBIT margin floors at 30.8%
D&A ($bn)EBIT margin (%) (right)$0bn$50bn$100bn$150bn30.0%32.5%35.0%37.5%40.0%42.5%FY24AFY26EFY28EFY30EFY32EActualEstimates

Source: Company filings; Penmark Research model (FY24A–FY32E D&A and GAAP operating margin; FY26E includes $9.89bn 3Q26 legal charge).

2 · What the market believes

Sell-side FY27E EPS drifted from $34.92 to $34.03 over 90 days and sits at $33.97 today (39 cuts vs 7 raises in 30 days), while equity investors fear a 2022-style margin collapse toward 20%. Published FY27E EBITDA estimates span ~$173bn–$183bn (vs our $182.3bn EBITDA+SBC), indicating our −7.0% EPS gap vs Street stems from D&A ($47.6bn), cloud/AI serving costs (36.5% FoA cash cost ratio), and 2,670m diluted shares (+2.8% vs consensus).

3 · Valuation impact (what it is worth)

Absorbing the −$2.39/sh consensus EPS haircut upfront removes negative revision risk while preserving $716.84 base value ($774.27 rolled 12m target). Each ±2pp move in FoA cash operating cost ratio (36.5% base) shifts P/E value by ∓$44.86/sh (±6.3%). Because EBITDA margin rebounds from 42.3% (FY26E) to 47.7% (FY27E) and 50.0% (FY28E), cash earnings absorb the D&A wave.

Exhibit 50: 4 · Three independent evidence streams anchoring Meta’s >30.8% operating margin floor
Evidence source Observed quantitative proof (period & filing) Why it prevents a 2022-style margin collapse
1. Non-capex headcount & AI efficiency (3Q26 release) Headcount fell −15% y/y to 66,775 in 3Q26; first-party Muse models cut content-processing costs >80%, with management targeting >10x compute efficiency gains. Caps the FoA cash cost peak at 36.5% in FY27E (+7.7pp vs FY25A on cloud/serving) before easing 50–100bps/yr from FY28E.
2. Reality Labs explicit loss cap (3Q26 guidance) Management guided FY26E Reality Labs operating loss to $18bn–$19bn and explicitly framed FY27E RL loss around −$15.0bn. Narrows RL drag from 9.6% of sales in FY25A ($19.2bn) to 4.8% in FY27E ($15.0bn / $313.1bn) and 3.5% in FY28E ($13.0bn).
3. Model PP&E D&A schedule (10-Q / Model) Modelled at 12.0% of opening net PP&E plus 5.0% of in-year capex (blending 5.5-yr servers and 25-yr buildings), D&A rises to $47.6bn (FY27E) and $70.5bn (FY28E). Spreads the FY26E–28E GPU cohort across multi-year revenue compounding rather than expensing it in a single fiscal year.

Source: Company filings and Q3 2026 transcript; Penmark Research model.

5a · Signpost confirming Pillar 2 (Jan–Jul 2027) Initial FY27 total expense guidance (late Jan 2027) implies ~31–33% GAAP operating margin (~$208bn–$214bn total expenses) with Reality Labs loss narrowing toward $15bn.
5b · Signpost disproving Pillar 2 (Bear trigger) FY27 expense guidance implies <28% operating margin (FoA cash cost ratio >39.0%) or server useful-life cuts accelerate D&A above $55bn in FY27E.
PART 3 · THE THESIS · 12 PILLAR 3: CAPEX PEAK, FINANCING & AI REVENUE OPTIONALITY

Assumed 75% FY27E capex (−$80.6bn FCF) needs ~$195bn 4Q26–28E funding incl. $35bn FY28E notes

1 · The claim (rests on Section 9 Drivers 2–3 / Debate 3): Our 75.0% FY27E capex-to-revenue ratio ($234.9bn—an explicit analyst peak assumption, not historical fact) creates ~$195bn of 4Q26E–FY28E cash outflows bridged by cash, hybrid equity and ~$35bn of FY28E notes before capex eases to 25% of sales.
Exhibit 51: FCF troughs at −$80.6bn (FY27E), rebounds to +$146.0bn
−$100bn−$50bn$0bn$50bn$100bn$150bnFY24AFY26EFY28EFY30EFY32EActualEstimates$146.0bnBreakeven $0bn

Source: Penmark Research model (company/model FCF = CFO less capex incl. finance leases, FY24A–FY32E).

2 · What the market believes

Bear estimates of $250bn+ after the 3Q26 print ($711.38/sh implied value) treat negative FCF as permanent value destruction, leaving our $234.9bn base +$5.47/sh less bearish. Simultaneously, the market assigns near-zero standalone value to Muse, Meta One and WhatsApp Business AI—even though they featured in 55.4% of analyst Q&A.

3 · Valuation gap & Low-conviction honesty

This capex hump explains the $214.58 gap between our primary P/E value ($716.84) and 2040 DCF cross-check ($502.26): DCF discounts −$221.8bn of negative unlevered FCF over 4Q26E–FY29E (−$133.2bn company FCF over 4Q26E–FY28E) upfront at 10.80% WACC, whereas P/E captures that build via the multi-year D&A step-up. Because +1pt discount rate ($650.78 target) or high beta ($598.85) flips our Buy to Hold, conviction is Low.

Exhibit 52: 4 · Three independent evidence streams supporting peak FY27E capex funding & new AI monetization
Evidence source Observed quantitative proof (period & filing) How it anchors our model forecast & funding bridge
1. Commitments footnote (Q3 2026 Form 10-Q) $454.5bn of non-cancelable purchase/data-centre commitments ($121.4bn due in 2027) plus $18.48bn venture leases (and $41.49bn non-probable residual guarantees). Supports our aggressive 75.0% FY27E capex assumption ($234.9bn), fading to 66% (FY28E), 55% (FY29E), 46% (FY30E) and 25% terminal.
2. Oct–Nov 2026 capital raises (Form 8-K filings) $82.45bn 3Q26 cash/securities + $14.37bn 6.50% mandatory convertible preferred (23m shs) + $15.0bn senior notes (5.85%–7.50%) + Class C framework. Covers ~$195bn of 4Q26E–28E cash needs (−$133.2bn FCF, $47.0bn RSU taxes, $15.1bn divs) via $42.4bn cash draw, $29.4bn 4Q26 raises, $88.7bn fwd/ATM eq. & $34.9bn FY28E notes to issue.
3. Muse & Meta One early KPIs (3Q26 call) >10m US/Canada Muse setups in ~6 weeks (~25% DAU; >25% shopping at >$200 basket); Meta One >15m subs/trials; Meta AI engagement/user >2x since Apr. Backs FoA Other revenue growing +64% to $4.23bn (FY26E), +75% to $7.40bn (FY27E) and $23.19bn by FY30E.

Source: Meta Q3 2026 Form 10-Q, Form 8-K filings (Oct–Nov 2026), Q3 2026 transcript, and Penmark Research model.

5a · Signpost confirming Pillar 3 (FY27–FY28) Management frames FY27E as peak capex/sales (~65–75%), executes Class C ATM cleanly, and scales FoA Other above $6.0bn annualised run-rate by 2H27.
5b · Signpost disproving Pillar 3 (Rating-change trigger) FY27E capex exceeds $255bn without revenue upside or FY28E capex stays >70% of sales, pushing our rolled target below $705.51 (+10% Buy threshold) into Hold/Sell.
PART 4 · THE NUMBERS · 13 TRACK RECORD & QUALITY OF REPORTED NUMBERS

Operating cash flow doubled to $115.8bn between FY21A and FY25A, while SBC ($20.4bn) & $454.5bn commitments require care

Meta’s GAAP earnings convert cleanly to operating cash flow (191.5% CFO/net income in FY25A), because advertising collects in ~35 days with zero inventory. Three accounting items dominate analysis: SBC, legal accruals and off-balance-sheet GPU commitments.

Exhibit 53: Five-year reported track record, volume/price decomposition and FCF definition bridge ($m)
Metric ($m unless noted) FY21A FY22A FY23A FY24A FY25A 3Q26A
Revenue$117,929$116,609$134,902$164,501$200,966$64,601
Revenue growth y/y (ad imps / ad price)+37.2%−1.1%+15.7%+21.9%+22.2%+26.1% (+10/+14%)
GAAP operating income (EBIT)$46,753$28,944$46,751$69,380$83,276$12,988*
GAAP operating margin39.6%24.8%34.7%42.2%41.4%20.1% (35.4% ex-AG)
Stock-based compensation (SBC)$9,164$11,992$14,027$16,690$20,427~$7,200
Operating cash flow (CFO)$57,683$50,475$71,113$91,328$115,800$32,790
Capex incl. finance-lease principal (model/co.)——$28,103$39,225$72,215$34,905
Free cash flow (company/model definition)——$43,010$52,103$43,585−$2,120
Memo: Purchases of property & equipment (GAAP)$18,567$31,431$27,266$37,256$69,691$33,942
Diluted EPS (GAAP, $) / CFO-to-NI conversion$13.77 / 147%$8.59 / 218%$14.87 / 182%$23.86 / 146%$23.49 / 192%$3.83 / 333%

Source: Company filings; Penmark Research model (FY21A–FY25A & 3Q26A; lease-inclusive capex and company FCF reported from FY23A onward). *3Q26 includes $9.89bn AG legal accrual ($7.40 adj. EPS vs $3.83 GAAP).

Exhibit 54: CFO doubled from $57.7bn to $115.8bn vs $60.5bn profit
Operating cash flowGAAP net income$0bn$25bn$50bn$75bn$100bn$125bnFY21AFY22AFY23AFY24AFY25A$57.7bn$39.4bn$50.5bn$23.2bn$71.1bn$39.1bn$91.3bn$62.4bn$115.8bn$60.5bn

Source: Company Form 10-K GAAP cash flow and income statements (FY21A–FY25A; FY25A net income reflects $25.47bn tax provision).

Exhibit 55: Accounting quality & off-balance-sheet items in our model
ItemReported size & Penmark treatment
SBC ($20.4bn FY25A; 10.2% rev)Real economic cost: Never added back to GAAP EPS; deducted as a cash-equivalent cost in our 2040 DCF ($32.9bn FY27E).
PP&E D&A (12% open / 5% capex)Model blends 5.5-yr server & 25-yr building lives via 12.0% opening PP&E + 5.0% in-year capex D&A, lifting D&A to $47.6bn (FY27E) and $90.7bn (FY29E).
Legal accruals ($19.88bn 3Q26)3Q26 booked $9.89bn AG charge (FY26E GAAP EPS $28.89 vs $33.42* ex-charge); we deduct full $19.88bn ($11.58bn curr. + $8.30bn non-curr.) from equity value.
Commitments ($454.5bn)$121.4bn due 2027 + $18.48bn venture leases anchored in our capex/lease build; $41.49bn max residual guarantees excluded as not probable.

Source: Meta Q3 2026 Form 10-Q; Penmark model. *FY26E includes a $5.0bn tax benefit booked in 1Q26; the FY26E tax rate is 6.5% vs 16% in FY27E.

What reported earnings quality means for value per share Because we never add back $32.9bn of FY27E SBC to EPS and we deduct all $19.88bn of balance-sheet legal accruals (−$7.52/sh) upfront, our $31.58 FY27E EPS and $716.84 base value reflect clean GAAP shareholder economics rather than inflated “non-GAAP” earnings.
PART 4 · THE NUMBERS · 14 FORECASTS (1/2: REVENUE EQUATION & CONSENSUS BRIDGE)

Revenue scales from $201.0bn in FY25A to $313.1bn in FY27E and $477.6bn in FY30E on AI ad yield & FoA Other

Exhibit 56: Ad pricing tapers from +11.0% (FY27E) to +5.5% (FY30E) as impressions ease from +9.0% to +6.0%
Avg price per ad y/yAd impressions y/y0%+5%+10%+15%FY24AFY25A9M26AFY27EFY28EFY29EFY30EActualEstimates

Source: Meta Form 10-K and 3Q26 10-Q (FY24A–9M26A); Penmark Research model (FY27E–FY30E Family of Apps ad volume & pricing assumptions).

Exhibit 57: Ad sales reach $450.9bn by FY30E + $23.2bn FoA Other
$0bn$100bn$200bn$300bn$400bn$500bnFY24AFY26EFY28EFY30EActualEstimatesAdvertisingFoA OtherReality Labs

Source: Penmark Research model (FY24A–FY30E segment revenue in $bn).

Reasoning behind each revenue line

Family of Apps Advertising ($250.7bn FY26E → $303.3bn FY27E → $450.9bn FY30E): After +27.8% growth in FY26E, we model ad impressions decelerating smoothly from +9.0% (FY27E) to +8.0% (FY28E), +7.0% (FY29E) and +6.0% (FY30E) as Reels and Threads ad load matures. Average price per ad tapers from +11.0% (FY27E) to +8.0%, +6.5% and +5.5% as GEM/Andromeda conversion gains lap tough comps.

FoA Other ($4.23bn FY26E → $7.40bn FY27E → $23.19bn FY30E): Grows +75% in FY27E, +60% in FY28E, +45% in FY29E and +35% in FY30E on WhatsApp paid messaging, Meta One (>15m base) and Muse commerce take-rates. Reality Labs grows 12–15%/yr ($2.41bn → $3.57bn) on Ray-Ban/AI glasses.

Exhibit 58: Line-by-line Penmark forecasts vs management guidance and Yahoo Finance consensus
Line item Penmark Guide / Cons. Gap Primary analytical reason for difference
4Q26E revenue ($m)$75,373$75,155 cons+0.3%Near top of $73,000–$76,000m guide (incl. ~1% FX headwind) on +14% 3Q ad pricing momentum.
FY26E total expenses ($m)$176,035$175,000–$179,000 guideIn-lineIncludes $9.89bn 3Q26 AG legal accrual; FY26E GAAP EPS $28.89 vs $29.12 cons (−0.8%).
FY27E revenue ($m)$313,139$312,630 cons+0.2%59-analyst consensus ($312.6bn, 10 Nov 26) matches our +21.0% ad & +75% FoA Other build.
FY27E EBITDA + SBC ($m)$182,311$183,283*−0.5%Published EBITDA listings span ~$173bn–$183bn; cash operating profitability matches Street.
FY27E diluted EPS ($)$31.58$33.97 cons−7.0%Higher D&A ($47.6bn), FoA cloud/serving costs (36.5%), net interest (−$1.44bn) & 2,670m diluted shares (+2.8% vs Street).
FY27E capex / FCF ($m)$234,854 / −$80,555$217,172 / −$46,956+8.1% capex75.0% capex/rev peak assumption vs ~30 Oct 26 FCF-implied Street capex ($170.2bn CFO − $217.2bn).

Source: Penmark Research model; Yahoo Finance consensus (rev/EPS 10 Nov 2026; NI/OCF/FCF ~30 Oct 2026; *workbook consensus EBITDA+SBC benchmark).

Why our revenue forecast matches Street (+0.2%) while FY27E EPS sits 7.0% lower Both Penmark ($313.1bn) and 59-analyst consensus ($312.6bn) model ~21–22% FY27E top-line growth, isolating our −$2.39/sh EPS gap ($31.58 vs $33.97) entirely to non-operating and capital-intensity lines: $47.6bn D&A, 36.5% FoA cash costs, −$1.44bn net financing drag and 2,669.6m diluted shares.
PART 4 · THE NUMBERS · 14 FORECASTS (2/2: COSTS, BALANCE SHEET & SENSITIVITIES)

EBITDA margin rebounds to 54.6% by FY30E while gross debt peaks at $158.6bn to defend a $40bn cash floor

Meta’s cost base splits cleanly between scalable FoA cash costs (36.5% → 34.0% of FoA revenue in FY27E–30E) and semi-fixed GPU depreciation ($18.6bn → $107.1bn). Hybrid financing bridges negative FY26E–28E FCF without breaching investment-grade leverage (peak 0.53x net debt/EBITDA).

Exhibit 59: Cost structure, capex/depreciation schedule and balance-sheet funding bridge ($bn)
Metric ($bn unless noted) FY24A FY25A FY26E FY27E FY28E FY29E FY30E
FoA cash operating cost % FoA rev26.5%28.8%34.3%36.5%36.0%35.0%34.0%
Reality Labs operating loss ($bn)−$17.7−$19.2−$18.4−$15.0−$13.0−$11.0−$9.0
EBITDA ($bn) / EBITDA margin$84.9 / 51.6%$101.9 / 50.7%$108.8 / 42.3%$149.4 / 47.7%$184.2 / 50.0%$222.0 / 52.4%$260.9 / 54.6%
Depreciation & amortisation ($bn)$15.5$18.6$27.8$47.6$70.5$90.7$107.1
Operating income ($bn) / margin$69.4 / 42.2%$83.3 / 41.4%$81.1 / 31.5%$101.8 / 32.5%$113.6 / 30.8%$131.3 / 31.0%$153.8 / 32.2%
Capex incl. leases ($bn) / % rev$39.2 / 23.8%$72.2 / 35.9%$140.0 / 54.5%$234.9 / 75.0%$243.2 / 66.0%$232.9 / 55.0%$219.7 / 46.0%
Free cash flow ($bn)$52.1$43.6−$1.4−$80.6−$40.2+$6.6+$56.5
Cash & securities / Gross debt ($bn)$77.8 / $28.8$81.6 / $58.7$93.5 / $98.7$44.8 / $98.7$40.0 / $133.5$40.0 / $158.6$63.4 / $158.6
Diluted shares (m) / Net debt/EBITDA2,614 / −0.58x2,574 / −0.22x2,576 / 0.05x2,670 / 0.36x2,716 / 0.51x2,740 / 0.53x2,740 / 0.36x

Source: Penmark Research model (FY26E EBITDA/EBIT reflect $9.89bn 3Q26 legal charge).

Exhibit 60: Gross debt peaks at $158.6bn (FY29E) to hold $40bn cash
$0bn$50bn$100bn$150bn$200bnFY23AFY26EFY29EFY32EActualEstimatesGross debt $99bnCash & securities $100bnCash floor $40.0bn

Source: Company filings; Penmark Research model ($14.37bn preferred + $15.0bn Nov 26 notes + $28.7bn fwd + $60bn FY27–28 ATM + $34.9bn FY28E & $25.0bn FY29E cash-floor notes).

Exhibit 61: Primary value spans $653.3–$781.9 on ad-price × FoA cost
$741.5$751.6$761.7$771.8$781.9$719.4$729.4$739.3$749.2$759.1$697.4$707.1$716.8$726.6$736.3$675.4$684.9$694.4$703.9$713.4$653.3$662.7$672.0$681.3$690.6Price-per-ad growth shift-2pt px-1pt pxBase (+11%)+1pt px+2pt pxFoA cash cost shift-2pt cost-1pt costBase (36.5%)+1pt cost+2pt cost

Source: Penmark Research model (9 Nov 2026 primary P/E base value across ±2pp FY27E ad-price & FoA cash cost shifts; all cells exceed $641.37).

Working capital, capex & financing mechanics

Receivables scale with sales ($19.8bn FY25A → $27.5bn FY27E), while payables/accruals partially offset working-capital absorption. Including $14.37bn Oct preferred (23m shs), 46m forward-sale shares and $60bn FY27–28 ATM, diluted shares reach 2,739.7m by FY29E.

Operating sensitivity takeaway

Across every cell of our ±2pt operating matrix ($653.3 to $781.9), present value stays above $641.37 (and rolled 12m value exceeds $723). Operational misses alone barely dent Buy; multiple/discount-rate shifts carry the real downside.

PART 4 · THE NUMBERS · 15 VALUATION (1/2: PRIMARY P/E, DCF CROSS-CHECK & RECONCILIATION)

Peer-median 22.94x FY27E P/E yields $716.84 base value and a $774.27 rolled target, +20.7% and a Buy

We value Meta primarily on peer-median forward P/E (22.94x FY27E EPS ex legal charges, less per-share legal accruals), cross-checked against an explicit FCFF DCF to FY2040 ($502.26). Peer P/Es are as published on 6 Nov and priced at the 9 Nov close (META 21.8x published forward; 20.3x on our FY27E EPS).

Exhibit 62: Primary P/E build & Page 1 target reconciliation
Primary P/E & target buildValue
FY27E diluted EPS (ex legal charges)$31.58
Peer-median forward P/E (GOOGL, AMZN, MSFT, APP)22.94x
Gross operating equity value per share$724.36
Less: 3Q26 legal accruals ($19,881m / 2,646.2m shs)−$7.51
Base-case primary value (at 9 Nov 2026)$716.84
Bear ($448.05 × 25%) + Base ($716.84 × 55%) + Bull ($962.94 × 20%)$698.86
1-yr roll-forward to 9 Nov 2027 (Ke net of div ~10.79%)×1.1079
12-month price target (probability-weighted)$774.27
Share price (9 Nov 2026) / Price return (Buy >+10%)$641.37 / +20.7%
FY27E dividend yield / 12m total return0.34% / +21.1%

Source: Penmark Research model (exact unrounded return +20.72%).

Exhibit 63: P/E ($716.84) & target ($774.27) vs DCF ($502.26)
Primary P/E (9 Nov value)12m rolled scenariotargetsDCF cross-check (to 2040)Discount rate ±1pt(weighted)Street 12m targets (57analysts)$0$250$500$750$1000$1250$448$963$496$1067$246$721$582$853$650$1000Price $641Target $774

Source: Penmark Research model; Yahoo Finance Street targets.

Exhibit 64: Explicit 2040 FCFF DCF cross-check ($m / $/sh)
2040 FCFF DCF componentBase value
WACC / Terminal growth / Terminal RONIC10.80% / 3.0% / 15.0%
PV of explicit FCFF (4Q26E–FY40E; SBC = cash cost)$498,560m
PV of terminal value (at FY40E)$752,070m
DCF enterprise value$1,250,630m
Plus cash, investments & hybrid / less debt & legal+$63,165m adj.
DCF equity value / share (2,646.2m shs, 30 Sep 2026)$496.48
DCF base value rolled to 9 Nov 2026$502.26
Gap vs primary P/E base ($716.84 − $502.26)−$214.58 (−29.9%)

Source: Penmark Research model.

Exhibit 65: Weighted 9 Nov value is $698.86 across Bear/Base/Bull
$0$200$400$600$800$1000$448$717$963Bear25%Base55%Bull20%Weighted $699Price $641

Source: Penmark Research model (column width = probability weight).

Why primary P/E ($716.84) exceeds our 2040 DCF ($502.26) by $214.58 (+42.7%) — and why P/E leads In FY26E–FY28E, Meta’s assumed AI capex ($618.0bn cumulative) exceeds depreciation ($145.9bn) by $472.1bn, turning FCFF deeply negative (−$80.6bn FCF in FY27E). DCF penalises the entire upfront cash outflow at 10.80% WACC while capping terminal RONIC at 15.0%, whereas mega-cap peers (GOOGL, AMZN, MSFT) also spend heavily on AI GPUs and are priced by public markets on forward GAAP P/E (where capex enters via multi-year D&A). We keep DCF visible as our downside floor and assign Low conviction because rate/beta stress pulls the target into Hold.
PART 4 · THE NUMBERS · 15 VALUATION (2/2: REVERSE DCF, SCENARIOS & SENSITIVITIES)

At $641.37 the market prices 20.55x FY27E P/E or 12.5% ad growth; +1pt rate cuts value −$121/sh

Exhibit 66: 2040 DCF grid spans $410.2–$644.8 (WACC × g)
$588.6$600.0$612.9$627.6$644.8$535.1$543.8$553.5$564.6$577.2$488.2$494.8$502.3$510.6$520.0$446.8$451.9$457.6$463.8$470.8$410.2$414.1$418.4$423.1$428.3Terminal growth2.0%2.5%3.0%3.5%4.0%WACC9.8%10.3%10.8% (Base)11.3%11.8%

Source: Penmark Research model (DCF value rolled to 9 Nov 2026).

Exhibit 67: Reverse DCF & price-implied solves at $641.37
Variable solved for $641.37Price-impliedPenmark Base
FY27E P/E (incl. $7.51 legal adj.)20.55x22.94x peer med
FY27E P/E (simple price / EPS)20.31x24.52x at target
FY27E ad revenue growth12.52%21.00% (+8.5pp)
DCF WACC (at 3.0% term. g)9.58%10.80% (+122bps)
DCF terminal growth (at 10.8% WACC)7.28%3.00%
EV / FY27E EBITDA (actual / target)10.93x13.28x at target

Source: Penmark Research model.

Exhibit 68: Bull, Base and Bear scenario matrix tied to Section 9 confirmed stock drivers
Scenario (subjective weight) FY27E ad imps / price FoA cash cost FY27E capex Target P/E 9 Nov value 12m target Price ret. / Call
Bear case (25.0% weight)+7.0% / +6.5% (+14.0%)38.5%$258.7bn (88%)18.10x (GOOGL)$448.05$496.39−22.6% (Sell)
Base case (55.0% weight)+9.0% / +11.0% (+21.0%)36.5%$234.9bn (75%)22.94x (median)$716.84$794.19+23.8% (Buy)
Bull case (20.0% weight)+11.0% / +12.5% (+24.9%)35.0%$220.9bn (68%)27.20x (MSFT)$962.94$1,066.85+66.3% (Buy)
Probability-weighted outcome———23.0x FY28E (24.5x FY27E)$698.86$774.27+20.7% (Buy)

Source: Penmark Research model (2040 DCF cross-check values: Bear $246.36, Base $502.26, Bull $720.65, Weighted $481.96). Limits note: scenario weights (25% / 55% / 20%) are subjective analyst judgements reflecting right-skewed AI ad upside vs capex/multiple downside.

Exhibit 69: Discount-rate moves (−$121.0 / +$64.5) and ad pricing (−$47.2) swing base $716.84 value most
10Y yield / discount rate (+1.0pt / -0.6pt)FY27E ad price (+6.5% bear / +12.5% bull)FoA cash cost ratio (38.5% bear / 35.0% bull)FY27E ad impressions (+7% bear / +11% bull)FY27E capex intensity ($259bn bear / 68% rev bull)Muse/Meta One & legal (-$4.9 legal / +$8.2 Muse)Base $716.8−$121.0+$64.5−$47.2+$14.6−$45.0+$33.7−$20.4+$21.4−$11.9+$13.2−$4.9+$8.2

Source: Penmark Research model (one-at-a-time driver shocks around $716.84 base value at 9 Nov 2026).

Target reconciliation to Page 1 & Page 2 call box ($774.27 12-month target; +20.7% Buy) Weighting our 9 Nov 2026 scenario values ($448.05 × 25% + $716.84 × 55% + $962.94 × 20%) yields $698.86/sh today. Rolling $698.86 forward 12 months by cost of equity less dividend yield (×1.107902) gives our $774.27 12-month target—an unrounded +20.72% price return above $641.37 (+21.06% total return incl. 0.34% FY27E dividend yield), clearing Penmark’s +10.0% Buy threshold.
PART 4 · THE NUMBERS · 16 CATALYSTS

January 2027 Class C dividend & estimated late-Jan 4Q26/FY27 capex guide anchor the 12-month catalyst path

Because four of the last five earnings prints coincided with capex-related sell-offs (4Q25 rallied +10.5pp), clearing initial FY27E capex guidance (est. late January 2027) is the top catalyst for our $774.27 target (+20.7%).

Exhibit 70: Dated 12–18 month catalyst calendar — Penmark expectations vs Street consensus
Expected window Catalyst event Penmark expectation vs Street / guidance Potential share-price / rating impact
~19 Jan 2027 (expected) Class C 1-for-1 non-voting stock dividend effective 1 Class C share issued per Class A/B share; halves nominal share price mechanically while creating currency for $60bn FY27–28 ATM. Neutral on intrinsic value; removes founder-control uncertainty around equity funding.
Late Jan 2027 (unconfirmed est.) 4Q26 results & initial FY27 capex / expense guidance 4Q26E rev $75.37bn vs $75.16bn cons ($73bn–$76bn guide); FY27 capex framed near $220bn–$240bn vs bear estimates of $250bn+ after 3Q26. High (±6–10%): Capex ≤$235bn + 1Q27 ad growth >21% resets EPS floor; capex >$255bn threatens Buy.
Late Apr 2027 (est.) 1Q27 results & 1H27 AI ad-pricing durability check Ad price y/y holds ~+10% to +12% against +12% 1Q26 comp; FoA Other accelerates toward $1.5bn+ quarterly run-rate. Medium-High (±5–8%): Confirms GEM/Andromeda conversion gains persist into year two.
2Q27–3Q27 (May–Sep 2027) Llama / Muse commerce & Meta Connect 2027 showcase Expanded US/international Muse checkout (>$200 avg order), Meta One paid tier conversions (>15m base) & next-gen AI glasses. Medium (+3–6%): Shifts 55.4% Q&A focus into tangible FY28E Street revenue upgrades.
Late Oct 2027 (est.) 3Q27 results & first FY28E capex-to-sales peaking signal Management signals FY28E capex moderating as % of sales (Penmark 66.0% vs 75.0% FY27E) and RL loss holding ≤$15bn. High (+8–12%): Catalyst for multiple expansion from 20.3x on our FY27E EPS toward 23.0x target P/E on FY28E EPS.

Source: Meta Form 8-K (21 Oct 2026) for expected 19 Jan 2027 Class C distribution; earnings dates are Penmark estimates based on historical reporting cadence (not yet formally scheduled by Meta).

Exhibit 71: Section 9 driver watch-list & rating-change thresholds
Section 9 driver (Debate) Next metric & reporting date Penmark vs Street Threshold that changes our Buy rating (Hold band: −10.0% to +10.0%)
Driver 1: AI ad pricing (D1) 4Q26 ad price y/y (est. late Jan 27) ~+12.0% vs ~+11% Sustained FY27E ad price <+6.5% cuts rolled target toward $705.51 (+10.0% Buy/Hold boundary at $641.37).
Driver 2: Capex & D&A (D2, D3) FY27 capex guide (est. late Jan 27) $234.9bn vs $250bn+ bear FY27E capex >$255bn without revenue upside + 2.0x peer multiple compression moves target below $705.51 (Hold).
Driver 3: Muse / FoA Other (D3) 4Q26 / 1Q27 FoA Other ($m) $7.40bn FY27E (+75%) Upside option: faster Muse GMV ramp lifts Bull weight above 20%, widening buffer above +10% Buy threshold.
Driver 4: Cost of equity (D1, D3) 10Y UST / beta regime (ongoing) 10.80% base WACC Immediate trigger: +1pt discount rate cuts target to $650.78 (+1.5%, Hold); 1.57 beta cuts target to $598.85 (−6.6%, Hold).

Source: Penmark Research model.

Price-driven vs fundamental rating-change boundaries (Penmark rating rule) Holding our $774.27 12-month target unchanged, a pure share-price rally above $703.88 reduces 12-month price upside to ≤+10.0% (Hold), while a rally above $860.30 pushes return below −10.0% (Sell). Conversely, at $641.37, any fundamental revision cutting our weighted 12m target into $577.23–$705.51 triggers a downgrade to Hold (below $577.23 = Sell).
PART 4 · THE NUMBERS · 17 RISKS

Two robustness tests flip Buy to Hold, anchoring Low conviction: a 1pt rate rise gives $651 and high beta $599

As a Buy-rated initiation, our risk analysis centres on downside threats to our $774.27 target and +20.7% expected return. Heavy near-term cash demands (~$195bn over 4Q26E–FY28E, incl. −$133.2bn FCF and ~$35bn of FY28E notes to issue) magnify sensitivity to discount rates, capex discipline and regulatory rulings.

Exhibit 72: Rate (+1pt) & beta (1.57) stress cut target into Hold
Scenario Bear / BullPublished beta 1.57 / 1.23Discount rate +1pt / -1ptTarget $774$496$1067$599$788$651$936

Source: Penmark Market Check robustness tests ($650.78 target at +1pt rate = +1.5% Hold; $598.85 at 1.57 beta = −6.6% Hold).

What would end the Buy thesis outright

A simultaneous AI capex overrun (>$255bn FY27E) and ad-pricing stall (<+6.5% y/y) triggers our $448.05 Bear value ($496.39 rolled 12m target, −22.6% Sell). In that scenario, P/E on our FY27E EPS compresses from 20.3x to Alphabet’s 18.1x floor while FY27E EPS drops toward ~$25.2 on unrecovered D&A and financing dilution.

What moves Buy to Hold first: Even on base operations, +100bp WACC ($650.78 target, +1.5%) or 1.57 beta ($598.85, −6.6%) pulls expected return inside −10% to +10%.

Exhibit 73: Principal downside risks, likelihood, per-share value impact and governance/legal exposures
Risk category Likelihood Value / share impact Mechanism, filing evidence & mitigation
1. AI capex & commitment escalation Medium-High −$120 to −$269/sh (Bear $448.05) $454.5bn non-cancelable commitments ($121.4bn in 2027) + $18.48bn venture leases + $41.49bn max residual value guarantees could lock in >70% capex/sales past FY28E if GPU obsolescence accelerates.
2. Rate / beta multiple compression Medium −$123.5 to −$175.4/sh on 12m target +1pt WACC cuts rolled target to $650.78 (+1.5% Hold); upper-end published beta (1.57) cuts target to $598.85 (−6.6% Hold). Wider credit spreads on ~$60bn of FY28E–29E cash-floor notes ($34.9bn FY28E, $25.0bn FY29E) raise interest drag, while lower equity prices increase share dilution on the $60bn FY27E–28E Class C ATM.
3. Ad pricing & macro / AI substitution Medium −$19.4/sh per −2pp (−$47.2 Bear) SMB ad budgets are macro-cyclical; retail media (Amazon $82.1bn) and external AI answer engines could siphon high-intent commercial queries.
4. Legal, EU DMA & youth-safety rulings Medium −$7.51/sh accrued + EU margin risk 3Q26 booked a $9.89bn AG charge (up to ~$18.59bn cash outflows; $19.88bn total balance-sheet legal accruals). EU “consent-or-pay” remedies threaten European ARPP ($14.23bn 3Q26 sales).
5. Founder control, key-person & power ESG Structural Governance discount (~1–2x P/E) Zuckerberg’s 60.8% vote + Class C structure remove external checks on capital allocation; grid-power & water permitting bottlenecks for multi-GW AI clusters represent a material operational ESG risk.
Upside risk against Low conviction 20% Bull +$292.6/sh to $1,066.85 target Faster 10x inference efficiency cuts FY27E capex toward 68% of sales ($220.9bn) while Muse/Meta One scale, re-rating META to MSFT’s 27.2x multiple.

Source: Meta Q3 2026 Form 10-Q, 2026 DEF 14A, and Penmark Research model.

Contingent liability watch — Venture residual value guarantees ($41.49bn maximum exposure) Per Q3 2026 Form 10-Q commitments disclosures, Meta carries up to $41,493m of maximum residual value guarantees on data-centre joint ventures. Following GAAP probability thresholds, our base model deducts $0 because payment is not currently probable; a 25% realisation would reduce equity value by ~$3.90/sh.
PART 4 · THE NUMBERS · 18 OWNERSHIP & TRADING

Institutions held 76.8% of Class A at 2Q26, short interest is 1.3% at 1.8 days, and 88% of Street rates META Buy

Exhibit 74: Short interest fell to 28.4m shs on 15 Oct (1.8d cover)
0m10m20m30m40m30 Jun15 Jul31 Jul14 Aug31 Aug15 Sep30 Sep15 Oct34.5m37.8m31.5m28.0m28.8m31.0m33.6m28.4m

Source: FINRA bi-weekly short interest through 15 Oct 2026 settlement (−15.4% vs 30 Sep; 1.80 days to cover).

Exhibit 75: Penmark $774 target vs $833 Street mean ($650–$1,000)
Price $641Penmark $774High $1000Low $650Mean $833

Source: Yahoo Finance Street ratings (50 Buy incl. 3 Strong Buy, 7 Hold, 0 Sell; 10 Nov 2026).

Liquidity, index membership & supply overhangs

META trades $14.02bn/day (21.29m 3m avg daily shares) across S&P 500, Nasdaq-100 and XLC (~22% weight). Near-term equity supply comes from 23m mandatory preferred shares, 46m forward-sale shares, and our modelled $60bn FY27E–28E Class C ATM (~4.3 days of ADV).

Street positioning & crowding check

With 50 of 57 analysts rating META Buy (including 3 Strong Buy, 7 Hold, 0 Sell; $833.12 mean target and $650 Street low above $641.37), sell-side positioning is crowded bullish—reinforcing why our $31.58 EPS (−7.0% vs Street) and $774.27 target stay conservative.

Exhibit 76: Top 10 institutional 13F holders (quarter ended 30 Jun 2026)
Institution (Q2 2026 13F)Shares (m)% Class A*
Vanguard Group204.869.29%
BlackRock173.407.86%
FMR LLC (Fidelity)95.904.35%
State Street92.034.17%
Geode Capital Management54.952.49%
JPMorgan Chase45.942.08%
Capital World Investors40.881.85%
Capital Research Global39.621.80%
Morgan Stanley39.441.79%
Invesco38.371.74%
Top 10 / Total 13F reported825.39 / 1,693.6337.4% / 76.8%

Source: SEC Form 13F filings for quarter ended 30 June 2026 (not Q3/Q4 2026; *% of 2,205.13m Class A shares outstanding basis).

Insider activity & Form 4 interpretation

Recent Form 4 filings reflect routine liquidity, derivative settlements and pre-arranged Rule 10b5-1 plans rather than directional insider views. Filings show a 4 Nov 2026 Zuckerberg derivative-leg disposition (97,935 shs at $633.80), 2 Nov Marc Andreessen sales ($606.85–$613.88), and Javier Olivan sales executed under a pre-arranged Rule 10b5-1 trading plan. Because Form 4 footnotes do not report plan indicators uniformly across every filer, we draw zero directional signal from aggregate insider sales.

Share-class mechanics, Class C dividend adjustment (~19 Jan 2027) & equity supply calendar Meta’s capital structure comprises publicly traded Class A shares (1 vote/sh; ~2,214m shares) and founder-held Class B shares (10 votes/sh; 342.4m shares, 99.8% controlled by Mark Zuckerberg = 60.8% total voting power). The approved one-for-one non-voting Class C stock dividend (expected ~19 January 2027) doubles total shares without changing enterprise or intrinsic equity value (mechanically halving per-share metrics) while providing liquid non-voting currency for the 46m forward-sale shares, 23m mandatory convertible preferred shares (6.50% coupon) and our modelled $60bn FY27E–28E ATM issuance (+3.6% FY27E diluted shares vs FY26E; +2.8% vs consensus share count; +6.4% by FY29E).
PART 5 · APPENDIX · 19 FINANCIAL MODEL (ANNUAL SUMMARY FY23A–FY30E)

Summary statements, balance sheet and valuation ratios: EPS reaches $44.54 by FY30E

Exhibit 77: Consolidated income statement summary (FY23A–FY30E, $m except per-share)
Income statement ($m) FY23A FY24A FY25A FY26E FY27E FY28E FY29E FY30E
Advertising revenue$131,948$160,633$196,175$250,701$303,323$353,796$403,168$450,862
FoA other + Reality Labs rev$2,954$3,868$4,791$6,385$9,817$14,621$20,367$26,762
Total revenue$134,902$164,501$200,966$257,086$313,139$368,416$423,535$477,624
Revenue growth y/y+15.7%+21.9%+22.2%+27.9%+21.8%+17.7%+15.0%+12.8%
EBITDA$57,929$84,878$101,892$108,834$149,431$184,169$221,988$260,887
EBITDA margin42.9%51.6%50.7%42.3%47.7%50.0%52.4%54.6%
Depreciation & amortisation$11,178$15,498$18,616$27,783$47,638$70,519$90,725$107,129
Operating income (EBIT)$46,751$69,380$83,276$81,051$101,793$113,650$131,263$153,758
Operating margin34.7%42.2%41.4%31.5%32.5%30.8%31.0%32.2%
Net interest/other & tax−$7,653−$7,020−$22,818−$6,610−$17,498−$22,133−$26,806−$31,720
GAAP net income$39,098$62,360$60,458$74,440$84,295$91,517$104,457$122,037
Diluted EPS (GAAP / ex-legal $)*$14.87$23.86$23.49$28.89 / $33.42$31.58$33.69$38.13$44.54
Exhibit 78: Cash flow statement and balance sheet summary (FY23A–FY30E, $m)
Cash flow & balance sheet ($m) FY23A FY24A FY25A FY26E FY27E FY28E FY29E FY30E
Stock-based compensation (SBC)$14,027$16,690$20,427$28,440$32,880$36,842$40,236$42,986
Operating cash flow (CFO)$71,113$91,328$115,800$138,579$154,300$203,003$239,529$276,170
Capex incl. finance leases$28,103$39,225$72,215$140,000$234,854$243,155$232,944$219,707
Free cash flow (model/co. def.)$43,010$52,103$43,585−$1,421−$80,555−$40,151$6,585$56,463
Dividends paid / Buybacks$0 / $19,774$5,072 / $30,125$5,324 / $26,248$5,440 / $0$5,766 / $0$6,112 / $0$6,479 / $0$6,868 / $0
Cash & marketable securities$65,403$77,815$81,592$93,457$44,799$40,000$40,000$63,373
Property & equipment, net$96,587$121,346$176,400$299,129$486,345$658,981$801,200$913,778
Gross debt ($m)$18,385$28,826$58,744$98,669$98,669$133,541$158,564$158,564
Net debt (net cash) ($m)−$47,018−$48,989−$22,848$5,212$53,870$93,541$118,564$95,190
Total stockholders’ equity$153,168$182,637$217,243$310,215$459,287$588,125$701,211$833,145
Exhibit 79: Per-share data, returns, leverage and multiples at $641.37 share price (FY23A–FY30E)
Ratios & valuation FY23A FY24A FY25A FY26E FY27E FY28E FY29E FY30E
Diluted shares (m) / Div./dil. sh2,629 / $0.002,614 / $1.942,574 / $2.072,576 / $2.112,670 / $2.162,716 / $2.252,740 / $2.362,740 / $2.51
Book value / share ($)$58.26$69.87$84.40$120.42$172.04$216.51$255.94$304.10
ROE (GAAP) / ROCE28.0% / 30.4%37.1% / 36.2%30.2% / 34.2%28.2% / 23.7%21.9% / 21.1%17.5% / 17.8%16.2% / 16.6%15.9% / 16.6%
ROA / Net debt to EBITDA18.8% / −0.81x24.7% / −0.58x18.8% / −0.22x16.4% / 0.05x13.8% / 0.36x12.0% / 0.51x11.3% / 0.53x11.4% / 0.36x
P/E (ex-legal) at $641.37*43.1x26.9x27.3x19.2x (22.2x GAAP)20.3x19.0x16.8x14.4x
EV / EBITDA / P/B at $641.3728.2x / 11.0x19.3x / 9.2x16.0x / 7.6x15.0x / 5.3x10.9x / 3.7x8.9x / 3.0x7.4x / 2.5x6.3x / 2.1x
FCF yield / Dividend yield2.6% / 0.0%3.1% / 0.30%2.6% / 0.32%−0.1% / 0.33%−4.7% / 0.34%−2.3% / 0.35%+0.4% / 0.37%+3.2% / 0.39%
P/E at $774.27 target price52.1x32.5x33.0x23.2x24.5x23.0x20.3x17.4x

Source: Company filings; Penmark Research model (cut-off 9 Nov 2026; buybacks $0 in 9M26A & FY26E–30E). *FY26E includes a $5.0bn tax benefit booked in 1Q26; the FY26E tax rate is 6.5% vs 16% in FY27E.

PART 5 · APPENDIX · 19 SEGMENT & STUB DETAIL · 20 REFERENCE & METHODOLOGY

Quarterly/segment progression, GAAP-to-adjusted reconciliation, methodology and data cut-offs

Exhibit 80: Quarterly, 9M26A/4Q26E stub, segment KPIs and GAAP-to-adjusted EPS reconciliation ($m)
Segment, KPI & reconciliation 1Q25A 2Q25A 3Q25A 4Q25A 9M26A 4Q26E FY26E FY27E
Advertising revenue ($m)$41,392$46,563$50,082$58,137$177,448$73,253$250,701$303,323
FoA Other / Reality Labs rev ($m)$510 / $412$583 / $370$690 / $470$801 / $955$2,989 / $1,276$1,242 / $879$4,231 / $2,155$7,403 / $2,413
Total revenue ($m)$42,314$47,516$51,242$59,893$181,713$75,373$257,086$313,139
DAP (bn) / Ad imps / Ad price y/y3.43 / +5/+10%3.48 / +11/+9%3.54 / +14/+10%3.58 / +18/+6%3.64 / +14/+13%3.67 / +26% ad3.67 / +27.8% ad3.74 / +9/+11%
GAAP operating income ($m)$17,555$20,441$20,535$24,745$54,635$26,416$81,051$101,793
Plus: Legal settlement charges ($m)$0$0$0$0$11,890*$0$11,890*$0
Operating income ex-legal ($m)$17,555$20,441$20,535$24,745$66,525$26,416$92,941$101,793
GAAP EPS / EPS ex-legal ($)$6.43 / $6.43$7.14 / $7.14$1.05 / $1.05~$8.87 / $8.87$20.44 / $24.98$8.45 / $8.45$28.89 / $33.42†$31.58 / $31.58

Source: Company filings; Penmark Research model (*9M26A legal charges = $9.89bn 3Q26 + $2.00bn 2Q26; 3Q25A EPS reflects $18,954m total tax provision, 87.5% effective rate). †FY26E includes a $5.0bn tax benefit booked in 1Q26; the FY26E tax rate is 6.5% vs 16% in FY27E.

Exhibit 81: Methodology, archetype & glossary
Topic / termPenmark convention & known limits
ArchetypeAdvertising platform: users × engagement × price; revenue built as ad impressions growth × price-per-ad growth, plus FoA Other & RL hardware.
Primary vs rejectedPrimary: Peer-median P/E (22.94x) on FY27E EPS ex-legal, less $7.52/sh legal accruals. Cross-check: FCFF DCF to 2040. Rejected: EV/EBITDA (ignores D&A step-up & venture leases), SOTP (RL is shared AI cost), FCF yield (negative through FY28E).
Share & SBC basisPre-Class C split basis (2,556.2m basic + 21m RSU + 46m fwd + 23m pref + ATM). SBC expensed in GAAP EPS & treated as cash cost in DCF.
Departures & limitsAnnual + 9M26A/4Q26E stub replaces 4-quarter forecast grid (company guides FY expenses/capex & 1Q ahead revenue). Published 1.25 beta used (1.23–1.57 range tested).
GlossaryDAP: Family daily active people. ARPP: Quarterly rev / DAP. GEM / Andromeda: GPU ad ranking & retrieval models. Muse: Meta consumer/commerce AI assistant.
Exhibit 82: Peer set, ratios & data cut-off dates
ReferenceDefinition & as-of date
Valuation peers (9 Nov 26)Alphabet (GOOGL, 18.10x), Amazon (AMZN, 25.05x), Microsoft (MSFT, 27.20x), AppLovin (APP, 20.83x); median 22.94x (fwd EPS published 6 Nov, priced 9 Nov).
Peer P/E basisPeer forward P/Es are as published on 6 Nov and priced at the 9 Nov close (META 21.8x on published forward basis vs 22.94x peer median; 20.3x on our $31.58 FY27E EPS).
Key ratio formulasModel FCF: CFO − capex incl. finance-lease principal. ROCE: EBIT / average (equity + gross debt). Model EV: $1,634.0bn (2,646.2m dil. shs × $641.37 − $63.2bn net cash/bridge; 3Q26 basic EV $1,640.7bn).
Model cut-offModel run 11 Nov 2026 on 9 Nov 2026 closing price ($641.37); USD reporting currency; FYE 31 Dec.
External data datesSEC filings/KPIs: through 3Q26 10-Q (22 Oct 26) & 5 Nov 26 8-K. Consensus: Yahoo Finance rev/EPS 10 Nov 26; NI/FCF ~30 Oct 26. 13F: QE 30 Jun 26. Short int.: 15 Oct 26.
Per-share basis & Class C 1-for-1 stock dividend convention (~19 January 2027) All historical and forecast per-share figures, scenario values ($448.05 Bear / $716.84 Base / $962.94 Bull) and our $774.27 12-month target are stated on a pre-Class C dividend share basis anchored to the 9 November 2026 close ($641.37). Upon payment of the approved 1-for-1 non-voting Class C stock dividend (~19 January 2027), total share count doubles and all nominal per-share metrics adjust mechanically by 0.5x with zero change to enterprise value, percentage price return (+20.7%) or Buy rating.
PART 5 · APPENDIX · 21 DISCLOSURES

Analyst certification and important disclosures

Ratings

Ratings are based on the upside from the share price to our 12-month price target, excluding dividends. The price target is weighted across bull, base and bear scenarios.

BuyUpside to price target above +10%
HoldUpside to price target from −10% to +10%, inclusive
SellUpside to price target below −10%

We review ratings at each set of results and whenever the facts change materially. A share price can move outside its band before the rating changes, and targets may not be reached.

Analyst certification

The Penmark Team certifies that (1) the views in this report accurately reflect its views about the subject company and its securities, and (2) no part of its compensation was, is or will be directly or indirectly related to the specific ratings, price targets or views in this report.

Contact

Penmark Research

Questions about our research:
research@penmarkresearch.com

All our research and coverage:
penmarkresearch.com

Basis of preparation

This report uses information publicly available up to its publication date. Share prices are closing prices on the trading day before the publication date, unless stated otherwise. Forecasts, valuations and price targets are Penmark Research’s own, and forecast periods are marked E. Sources are shown beneath each exhibit.

Valuation and price targets

Price targets are based on the valuation methods set out in each report and are weighted across bull, base and bear scenarios. The main risks to each rating and price target are described in the report. Penmark Research updates its coverage after each set of results but is not obliged to update a report at any other time.

Independence and conflicts of interest

Penmark Research is independent. It does not provide investment banking, brokerage, market-making or asset management services, does not trade securities, and does not accept payment from any company for research coverage. Companies do not see or review Penmark Research reports before publication.

No member of the Penmark Team holds a long or short position, directly or through derivatives, in the securities of any company Penmark Research covers.

Regulatory status

Penmark Research publishes general, impersonal investment research on a regular basis. It is not registered as an investment adviser with the US Securities and Exchange Commission or any state securities authority, is not a registered broker-dealer and is not a member of FINRA. Penmark Research does not give advice tailored to any person’s circumstances.

Not investment advice or an offer

This report is for information only. It is not personalised investment advice and is not an offer or solicitation to buy or sell any security or other financial instrument. It does not take into account any reader’s objectives, financial situation or needs. Readers should make their own investment decisions and seek independent professional advice where appropriate.

Forward-looking statements and risk

Forecasts, valuations and price targets are estimates based on assumptions that may prove wrong, and they may not be achieved. Opinions are current as of the publication date and may change without notice. Investing in equities involves risk, including the loss of capital. Share prices can be volatile, and past performance is not a reliable indicator of future results.

Data and sources

Company financials, operating data and filings come from company filings, earnings releases and earnings calls. Market data and any consensus estimates come from third-party sources named beneath the relevant exhibit; consensus figures from aggregators are unofficial. Penmark Research does not use other brokers’ research. Third-party data belong to their providers and may not be extracted from this report.

Accuracy and liability

Information is taken from sources believed to be reliable, but Penmark Research makes no representation or warranty as to its accuracy, completeness or timeliness. To the fullest extent permitted by law, Penmark Research accepts no liability for any loss arising from use of this report. Descriptions of litigation and regulatory matters reflect allegations and public records, not findings, unless stated. If a published report is corrected, the corrected version carries an “Updated” date at penmarkresearch.com.

Distribution and access

Penmark Research reports are available to registered users of penmarkresearch.com. Downloads and programmatic access are governed by the Penmark Research subscription terms. This report is not directed at, or intended for, any person in Singapore or India, and access from those countries is blocked. It is not intended for distribution to, or use by, any person in any jurisdiction where that would be contrary to law or regulation or would subject Penmark Research to registration or licensing requirements there.

Copyright and permitted use

© Penmark Research. All rights reserved. Penmark Research is a trade name of Brightriver Labs. Readers may share a link to this report but may not copy, reproduce, redistribute, sell or use its content, including for text and data mining, without written consent. Contact: research@penmarkresearch.com.

Keep reading

Sign in to Penmark