META

Executive Summary

Cheapest Mega-Cap, Costliest Bet

Meta owns the strongest advertising engine it has ever operated and the weakest cash conversion in its public history, and the two facts are the same fact. Q2 2026 revenue reached $60.801 billion, up 28% year over year per the 10-Q filed July 30, with ad impressions up 14% and average price per ad up 12% across 3.60 billion daily active people. eMarketer now projects Meta passes Google as the largest digital ad seller in the world this year, $243.46B versus $239.54B. On demand, there is no problem to solve.

The bill is the problem. Capital expenditure of $31.1 billion in the quarter consumed roughly 97% of operating cash flow and collapsed free cash flow 91% to $784 million, against $8.55 billion in the year-ago quarter. Management narrowed full-year capex to $130–145 billion and total expenses to $165–169 billion, and declined to put any number on 2027 at all. Diluted EPS of $6.18 missed the ~$7.22 consensus by about 14%, snapping a six-quarter beat streak, with a $2.4 billion legal charge and $1.18 billion of severance inside the number. The shares now sit 28.1% below the 52-week high of $790.80, below both the 50-day ($597.13) and 200-day ($627.63) averages.

The reported earnings series is unusable without adjustment, and this is where most screens go wrong. Q3 2025 GAAP EPS of $1.05 carries a one-time $15.93 billion non-cash tax charge; ex-charge it was $7.25. Q1 2026 GAAP EPS of $10.44 contains an $8.03 billion tax benefit worth $3.13 per share; ex-benefit it was $7.31. Clean trailing four-quarter earnings power is therefore near $29.60, not the $26.54 that screens report, putting the stock at roughly 19x trailing operating earnings and 16–18x forward. That is the lowest multiple in the mega-cap set against the fastest growth in it: MSFT 24.4x, GOOGL 25.9x, AMZN 28.1x.

The Street has trimmed price without trimming conviction. Of 62 covering analysts the split is 55 buy-equivalent, 7 hold, zero sell, average target $754, but more than twenty firms cut targets after the July 29 print and the surviving range runs $595 to $1,000 on identical numbers. That dispersion is the honest signal: nobody knows the payback period on $140 billion a year. Our probability-weighted target of $639 sits well below consensus and carries a HOLD, because the two variables that decide the next twelve months are both unresolved. The first 2027 capex figure does not arrive until the Q4 print. The federal youth-harm bellwether trial opened in Oakland on August 18 and runs roughly six weeks, after a Los Angeles jury already found malice in March. Buy the franchise on weakness, not the tape.

Q2 2026 scorecard

Record Revenue, Vanishing Cash — Q2 2026 was the best top line in company history and the worst cash quarter in a decade. Revenue of $60.801B (SEC EDGAR) grew 28% year over year while capital expenditure of $31.1B absorbed almost all operating cash flow, leaving $784M of free cash flow against $8.55B a year earlier.

Revenue, Q2 2026

$60.801B

+28% YoY vs $47.516B in Q2 2025 · SEC EDGAR 10-Q filed Jul 30, 2026

Diluted EPS, Q2 2026

$6.18

Down from $7.14 · missed ~$7.22 consensus by ~14% · includes $2.4B legal, $1.18B severance

Operating margin, Q2 2026

31%

Operating income $18.78B, down 8% YoY while revenue rose 28% · total costs +55%

Family daily active people

3.60B

June 2026 average · ad impressions +14%, price per ad +12% YoY

Capex, Q2 2026

$31.1B

~97% of operating cash flow · $50.9B spent in H1 against a $130–145B FY guide

Free cash flow, Q2 2026

$784M

Down ~91% from $8.55B · sub-1% FCF yield on a $1.46T cap removes the buyback floor

Clean trailing EPS

$29.60

COMPUTED: $26.54 reported, add back the $15.93B Q3-25 tax charge, strip the $3.13 Q1-26 benefit

Forward P/E · EV/Revenue

17.8x · 6.4x

Cheapest mega-cap in the peer set · price $568.97, −28.1% from the 52-week high

Quarterly Revenue — Last 4 Quarters ($B)

$51.2B

Q3 2025

$59.9B

Q4 2025

$56.3B

Q1 2026

$60.8B

Q2 2026

Growth engine

The Ad Machine Is Not the Problem

3.60B

Family daily active people

June 2026 average, per the Q2 2026 release

$243.46B

2026 worldwide ad revenue (eMarketer)

Passes Google's $239.54B for the first time · Meta accelerating to 24.1% growth against Google at 11.9%

+14% / +12%

Ad impressions / price per ad

Q2 2026 YoY · volume and pricing rising together, the healthiest mix of the cycle

Sep 23–24, 2026

Meta Connect, Menlo Park

Two-day keynote on AI glasses, wearables and what Meta is teasing as its next computing platform. AI glasses sales more than tripled in 2025 and now absorb most Reality Labs investment, so this is the first read on whether the hardware pivot has a revenue line behind it.

Late Sep 2026

Oakland bellwether verdict window

The federal youth-harm trial (N.D. Cal., Judge Yvonne Gonzalez Rogers) opened Aug 18 and is expected to run roughly six weeks. The outcome is expected to shape resolution of some 1,600 pending family and school-district cases, and follows a March 2026 Los Angeles jury that found malice.

Late Oct 2026

Q3 print, the $61–64B guidance test

Management guided Q3 revenue to $61–64 billion against consensus near $63.1 billion, with Street EPS around $6.76. With $50.9B of capex already spent in H1, the FY guide implies a heavy H2 ramp. This is the first quarter where AI ad tools must show revenue against the spend.

Jan–Feb 2027

First 2027 capex guidance

No 2027 capex figure was given on the Q2 call. The Q4 print is the first scheduled moment for one, and it sets both the 2027 free-cash-flow path and the size of the depreciation step-up now working through the P&L. This is the single largest unquantified variable in the story.

2026–2027

Reality Labs peak burn and cloud optionality

Reality Labs lost about $19.2B in 2025 with a similar loss guided for 2026, which management frames as the peak. The Q2 segment loss was $4.62B on $431M of revenue. Separately, Zuckerberg has confirmed plans to sell excess AI compute, the first credible revenue story attached to the buildout.

Earnings & Signals

Forward Estimates, Surprises & Insider Activity

Forward Earnings Estimates

FY+1 EPS Consensus$31.05 (FY2026E consensus) · flattered by the $8.03B one-time Q1 tax benefit; clean basis closer to $27.90

FY+2 EPS Consensus$34.89 (FY2027E consensus) · Morgan Stanley models ~$33 after an ~8% cut; our estimate is $33.60

PEG Ratio0.85 (stockanalysis.com) · methodology-sensitive, GuruFocus publishes 3.68 on a different growth basis

Forward P/E17.8x on FY2026E · 16.3x on FY2027E consensus · ~19x on clean trailing EPS of $29.60

EPS Revisions (90d)↑1 ↓2 (Downward on the bottom line. Named actions since the Q2 print: Morgan Stanley cut FY26 and FY27 EPS ~8% each (Jul 30), Erste Group cut FY2027 EPS (Aug 5), Oppenheimer raised 2027–2028 revenue and EPS while staying at Perform. A Street-wide revision tally could not be sourced, so this counts only verified named actions.)

Guidance AccuracyRevenue guidance has been reliable, with Q2 2026 revenue of $60.80B landing above the $60.2–60.3B consensus for a sixth straight top-line beat. The credibility gap is on cost: FY expenses were raised to $165–169B and capex narrowed to $130–145B after two upward revisions, so management guides spend up faster than revenue.

Earnings Surprise Track Record

Q3 2025 Est: $6.70 Act: $7.25 adj. ($1.05 GAAP) +8.2%

Q4 2025 Est: $8.22 Act: $8.87 +7.9%

Q1 2026 Est: $6.66 Act: $10.44 +56.8%

Q2 2026 Est: $7.22 Act: $6.18 −14.4%

Beat Rate75% (3 of 4) · the Q1 beat is entirely the $3.13 tax benefit; on a clean basis Q1 was $7.31 and the streak still broke in Q2

Insider Activity (90 Days)

Net Buying/Selling$0 · zero insider purchases; 87 Form 4 transactions over the trailing 90 days were all sales, totaling roughly $50 million

Sell/Buy Ratio87 sales : 0 buys

Uniformly negative but low signal. Most sales run through Rule 10b5-1 plans, so they are scheduled rather than opportunistic, and the dollar amounts are trivial against a $1.46T cap. The informative fact is the absence: no officer or director has bought at $568.97, roughly 28% below the high, a level that would normally draw at least a token conviction purchase.

CFO Susan Li · 30 sales, 148,802 shares, ~$95.0M over six months, no purchases. COO Javier Olivan · 117 sales, 38,565 shares, ~$24.2M. CTO Andrew Bosworth · 11 sales, 15,936 shares, ~$9.9M. Separately, on Jul 31, 2026 CZI Holdings converted 591,690 Class B shares to Class A and transferred them as a bona fide gift to Chan Zuckerberg Biohub. That is a charitable transfer, not a market sale, and several aggregators misclassify it as insider selling.

Peer Comparison

Relative Valuation vs. Competitors

Company	P/E (fwd)	EV/Rev	Rev Growth	Gross Margin
Meta Platforms	17.8x	6.4x	+28.0%	81.8%
Alphabet	25.9x	9.2x	+24.2%	60.9%
Microsoft	24.4x	10.9x	+17.8%	67.9%
Amazon.com	28.1x	3.8x	+19.6%	50.8%
Snap Inc.	7.4x	1.6x	+18.9%	57.3%

META is the only name here growing faster than every peer while trading at the lowest forward multiple of the mega-caps, 17.8x against 24.4x for MSFT, 25.9x for GOOGL and 28.1x for AMZN, on the highest gross margin in the group at 81.8%. The discount is not a growth discount, it is a cash-conversion discount: Alphabet and Microsoft carry richer EV/Revenue because their AI spend still leaves visible free cash flow, while Meta's Q2 capex absorbed nearly all of it. SNAP is the cautionary comparator, similar top-line growth with negative operating margin and a 7.4x multiple, which is what the market pays for scale without economics. META revenue growth of +28.0% is the SEC EDGAR figure ($60.801B vs $47.516B); peer ratios from stockanalysis.com as of Aug 18, 2026.

Valuation

Price Targets & Scenarios

Scenario	Price Target	Assumptions	Probability
Capex converts, cloud optionality gets priced	$780	Ad revenue holds mid-twenties growth through 2027 as AI ranking and generative ad tools compound, and Meta confirms a 2027 capex plateau near $150B rather than another step function. The compute-resale business gets a disclosed revenue line, reframing the buildout as an asset rather than a sinkhole, and free cash flow recovers toward $40B as the first cohorts of GPUs finish depreciating against real monetization. FY2027 EPS lands near $34 on the current consensus and the multiple re-rates to roughly 23x, back toward the mega-cap average.	28%
Ads compound, cash flow stays the overhang	$655	Revenue grows roughly 20% into FY2027 as the comparison base hardens, ad share gains against Google continue, and Reality Labs burn stays near $19B. 2027 capex is guided up again but inside expectations, keeping free cash flow depressed and capping the multiple. Depreciation and continued legal and severance charges hold FY2027 EPS near $33.60, below the $34.89 consensus, and the stock trades at a justified 19.5x, a deliberate discount to Meta's five-year average forward multiple for the missing free-cash-flow floor.	47%
Spend outruns monetization, verdict lands badly	$450	2027 capex is guided past $170B with no compute-resale revenue line, extending near-zero free cash flow through 2027 while the depreciation wave from two years of buildout resets operating margin below 30%. The Oakland bellwether returns a large plaintiff verdict, forcing reserves well beyond the $2.4B already booked and pulling product changes into the ad stack. On roughly $30.70 of clean EPS at a de-rated 14x, the stock retraces to $450, through the 52-week low.	25%

Probability-Weighted Target: $639 (+12% upside vs $568.97)

$639

Weighted

Bull $78028%

Base $65547%

Bear $45025%

Analyst Consensus

HSBC (Nicolas Cote-Colisson)

$830

Lowered from $905 · Buy maintained · Jul 30, 2026

J.P. Morgan (Doug Anmuth)

$800

Lowered from $875 · Overweight maintained · Jul 30, 2026

Morgan Stanley

$750

Lowered from $820 · FY26/FY27 EPS cut ~8% each · Jul 30, 2026

Jefferies (Brent Thill)

$710

Lowered from $825 · Buy maintained · Aug 3, 2026

55 Buy · 7 Hold · 0 Sell — 62 covering analysts split 55 buy-equivalent, 7 hold, zero sell, with an average target of $754.14, about 33% above the $568.97 close. More than twenty firms cut targets after the July 29 miss and essentially none cut ratings, so the Street trimmed price rather than conviction. The surviving range of $595 to $1,000 is the widest mega-cap dispersion in the group.

Technical

Key Levels & Options Intelligence

S/RSupport & Resistance

52-week high

$790.80

200-DMA · primary trend resistance

$627.63

Max pain (Jul 23, 2026)

$610

50-DMA · overhead, sits below the 200-DMA

$597.13

Broken support, now resistance

$577.22

Last · 18% of 52w range, RSI 47.6

$568.97

52-week low · last structural floor

$520.26

OptOptions & Sentiment

Put/call volume ratio: 0.58 · call-dominant
Put/call ratio, open interest (150d): 0.2527 (Aug 11, 2026)
25-delta put IV (30d): 39.4% vs 38.2% call IV
25-delta put skew: +1.2 pts · a regime shift after ~3 years flat or call-skewed
Short interest: 30.91M shares · 1.42% of float
Days to cover: 1.7 · no squeeze fuel
Max pain: $610 (Jul 23, 2026 · stale)
IV rank / percentile: n/a · no reliable current reading sourced

Conviction Analysis

Systematic Conviction Score: 56/100 (Medium)

Analyst Alignment

30%

FCF Visibility

25%

Catalyst Clarity

20%

Valuation Safety

15%

Mgmt Quality

10%

Direction of sell-side alignment is near-unanimous (55 buy / 7 hold / 0 sell) but targets span $595 to $1,000, so alignment scores 72 rather than 90. Free-cash-flow visibility is the binding constraint at 20: Q2 FCF of $784M against $31.1B of quarterly capex, with 2027 capex unguided, makes the cash path essentially unmodelable. Catalysts are unusually well dated (Connect Sep 23–24, the Oakland verdict window, the Q3 print, the Q4 capex disclosure) at 68, and valuation safety of 65 reflects the cheapest forward multiple in the mega-cap set offset by a sub-1% FCF yield. Management scores 60: elite ad execution, unchecked capital allocation under 61% voting control. Weighted total 56 lands squarely in Medium, which is why the rating is HOLD rather than BUY despite 12% weighted upside.

Risk & Ownership

Risk Assessment & Insider Signals

OOwnership & Insider Signals

Institutional ownership ≈64.6%: WallStreetZen puts Meta at 64.58% institutional, 11.42% insider and 24.00% retail. The high institutional share means the marginal seller in a de-rating is a benchmark-aware fund, and Q2 2026 13F data shows that seller is already active.
Top three index holders control about a fifth of the shares: Vanguard (~8.5%), BlackRock (~7.5%) and State Street (~4%) hold roughly 20% of shares outstanding between them. That block is passive and price-insensitive, so it neither cushions a decline nor supplies an activist check on capital allocation.
Dual-class structure gives Zuckerberg ~61% of the vote on ~13% of the equity: Class B shares carry ten votes each and sit primarily with Zuckerberg. There is no governance mechanism by which shareholders can force a change to the $130–145 billion capex program or the Reality Labs burn, a structural risk that turned material the moment free cash flow disappeared.
Q2 2026 13F filings show net institutional trimming: Reuters' review of 6,371 institutional filers found nearly 44% trimmed Magnificent Seven holdings including Meta against 42% initiating or expanding, a narrow but negative gap. Tiger Global specifically disclosed cutting its Meta position alongside Microsoft and Nvidia.
Marquee discretionary funds exited earlier in 2026: Filings reported earlier in the year show Bridgewater cutting its Meta stake by about 46.3% and Point72 reducing by roughly 91.9%, effectively liquidating. Their exit removes a layer of conviction capital that is not replaced by index flows.
Director-level distribution in August: A Form 4 dated Aug 4, 2026 records a disposition by director Marc L. Andreessen through a16z Capital Management. It adds to a Form 4 record that has been one-directional across the last two quarters.

Risk Heatmap

Quantified Risk Assessment

Severity Risk Factor Prob. PT Impact

High Free cash flow has effectively evaporated under the AI capex program

Q2 2026 free cash flow fell to $784 million from $8.55 billion a year earlier even as operating cash flow rose to $31.9 billion, because Q2 capex of $31.1 billion consumed almost all of it. FY2026 capex is guided to $130–145 billion against an estimated ~$8.5 billion of full-year FCF, versus $43.6 billion in 2025. At a $1.46T cap that is a sub-1% FCF yield, which removes the buyback-and-return floor the equity has traded on for three years.

85% -20%

High Federal youth-harm bellwether trial is running right now

A landmark federal trial opened Aug 18, 2026 in the Northern District of California (Oakland, Judge Yvonne Gonzalez Rogers) over platform design features alleged to addict minors, and is expected to run about six weeks. The verdict is expected to shape roughly 1,600 pending family and school-district cases, and follows a March 2026 Los Angeles jury that found Meta and Google liable with malice, oppression or fraud. Meta booked $2.40 billion of legal charges in Q2 and flags youth-related trials as possibly material.

45% -15%

High Depreciation hits the P&L years before the revenue does

GPU-heavy infrastructure carries a roughly four-to-six year useful life, so each dollar of the $130–145 billion 2026 capex begins converting to depreciation almost immediately while AI monetization stays unproven. Q2 total expenses already rose 55% to $42.03 billion and operating income fell 8% to $18.78 billion despite revenue growing 28%. Management declined to quantify 2027 capex at all, which leaves the largest single input unmodelable until the Q4 print.

70% -12%

High Multiple compression while the market reprices AI capex without a visible return

The stock sits 28.1% below its 52-week high with the 50-DMA ($597.13) below the 200-DMA ($627.63) and price below both, and it fell 7–10% on the Q2 print. Screens show ~21.4x on reported trailing EPS of $26.54, but the clean operating figure of $29.60 and the absence of any 2027 anchor mean the multiple is set by sentiment, not by a cash-flow floor. A 6.6% single-day rally on Morgan Stanley commentary alone shows how sentiment-driven the tape currently is.

55% -18%

Medium EU DMA enforcement is unresolved and widening into AI

The European Commission found Meta in breach of the DMA over its pay-or-consent model for failing to offer a genuinely less data-intensive option. On June 3, 2026 the General Court partially annulled the Marketplace designation in Case T-1078/23 but upheld Messenger, so the core obligations stand, and the Commission has since opened an investigation into Meta's WhatsApp AI policy. Exposure is extending into the AI product line rather than closing out.

50% -8%

Medium Reality Labs still burns roughly $19 billion a year with no inflection

Reality Labs posted a Q2 2026 segment loss of $4.62 billion on revenue of $431 million, widening from the prior quarter. Management guided full-year 2026 losses roughly in line with 2025's ~$19 billion, a second consecutive year with no reduction in burn. This is a standing charge against earnings that now compounds on top of the AI capex rather than trading off against it.

90% -5%

Recommendation

Summary

Rating

HOLD

Conviction

Medium

Price Target

$639

Timeframe

12 mo

Upside

+12%

Position Size

2%-4%

Entry Strategy

Tranche 1 — 30%

~$569

A starter only, and only for investors underweight the name. At 16–18x forward against 28% revenue growth you are paid to hold the best ad asset in the market, but the tape is below both moving averages with the 50-DMA under the 200-DMA, so there is no technical reason to lead with size. Keep this tranche small enough that a verdict headline does not force a decision.

Tranche 2 — 35%

~$525

Add on a retest of the 52-week low at $520.26. That level is the last structural floor on the chart and a hold there, ideally on a non-fundamental headline rather than a capex revision, is the highest-quality technical add in the setup. If it breaks decisively, wait for the reclaim instead of averaging into the break.

Tranche 3 — 35%

~$470

Reserve the largest reserve tranche for the bear case actually happening: a 2027 capex guide past $170B, a large plaintiff verdict, or both. Near $470 the stock trades close to 14x clean earnings, a level Meta has only seen in genuine capitulation, and the asymmetry flips decisively. If the Q4 print instead brings a capex plateau, deploy this tranche into strength rather than waiting for a price that will not come.

1-Year Price & Key Levels