GOOG Fundamental Valuation

Core Question: Will AI search replace Google, or will Google absorb AI? This determines the growth trajectory of $265B in revenue (>60% of GOOG total) — the single largest valuation variable (18.4%).

Why it matters now: Two forces are colliding. On one hand, Q4'25 revenue of $63B (+17%) proves search is still accelerating — management says AI Overviews doubled daily queries per user, with 3x longer query length and 1-in-6 queries now voice/image. AI is expanding search use cases, not replacing them. On the other hand, share has fallen from 92% to 89.6% (20-year low), ChatGPT now captures ~17% of digital queries, Gartner predicts a 25% decline in search volume by 2026, and 93% of AI Mode searches result in zero clicks.

The real debate isn't "is AI good or bad" — it's about RPM direction: If ad revenue per query (RPM) rises because AI improves ad relevance, then even declining query volume can be offset by higher monetization per query. But if zero-click rates keep climbing, RPM compresses — declining volume × declining RPM = accelerating revenue erosion. Current data supports the former (revenue is accelerating), but the trend supports the latter (zero-click is expanding).

Valuation impact: Bear(-12pp) → growth falls from 12% to 0% → 2028E revenue $215B → -$40/share. Bull(+5pp) → growth rises to 17% → 2028E $295B → +$21/share.Alphabet IR— StatCounter, Gartner, First Page Sage, Q4 Earnings

Core Question: Can Cloud close the gap with AWS/Azure in the AI era? The answer determines $500-700B of GOOG's EV attribution.

Why this is an inflection — three things happening simultaneously: ①Growth is accelerating, not decelerating (31%→48%) — extremely rare at $70B ARR scale. AWS at comparable scale had already decelerated to 29%. ②OPM went from losses to 30% in three years — margin inflection confirmed. ③$240B in signed contract backlog (not pipeline), +55% QoQ, $1B+ deals exceeding the prior three years combined, existing customers spending 30%+ above commitments.

TPU isn't just "cheaper": Google is the only company with proprietary chips (TPU v6) + proprietary foundation model (Gemini) + cloud infrastructure. End-to-end optimization delivers 4x better price-performance vs GPU solutions. Evidence: Gemini inference costs fell 78% in 2025. Anthropic signed for hundreds of thousands of TPUs (scaling to 1M by 2027), Meta in talks for multi-billion TPU lease. AWS depends on NVIDIA = costs controlled by a third party.

Bears have valid arguments: 13% share vs AWS 29% — absolute gap is still large. OPM 30% vs AWS 37%. Azure+OpenAI combo is compelling in enterprise. Multi-cloud strategies mean customers won't go all-in. How much of the $240B backlog is low-margin infrastructure contracts?

Alva judgment: Not whether Cloud is good (data proves it is), but whether 48% growth can be sustained. Bear → falls to +16% = -$17/share. Bull → accelerates to +46% = +$18/share.Cloud AI Blog · Synergy— CNBC, Investing.com, Q4 Earnings

Core Question: Can YouTube close the 60x Shorts monetization gap?Shorts RPM $0.03 vs long-form $3.00. Every 10% closure means billions in incremental revenue. This is the biggest internal lever in YouTube's $60B business.

Why YouTube is undervalued:$60B+ revenue (+17% YoY), 33% larger than Netflix ($45B), yet Netflix commands $370B market cap (8x P/S). As an Alphabet subsidiary, YouTube gets almost no standalone pricing. At 5-6x P/S, YouTube's standalone value would be $300-360B.

Three growth engines:①CTV is the fastest-growing surface — 700M hours of podcasts watched on TV (+70% YoY), shoppable ads launching, advertisers shifting from traditional TV budgets. ②Shorts monetization gap is 60x but closing — now 22% of ad distribution (15% last year), CPM +10-25% YoY. ③NFL Sunday Ticket + Oscars exclusive (2029-2033) locking in high-value households.

Risks:Q4 Q4 ad revenue $11.4B missed estimate of $11.8B (post-election political ad pullback). Shorts monetization gap may be structural (short-form ads inherently worth less). TikTok still has engagement advantage among younger users.

Alva judgment:Shorts 60x How much can the gap close? Bear → gap persists = YouTube growth drops to +3% → -$5/share. Bull → CTV+Shorts+NFL = +21% → +$5/share.YouTube Blog— Variety, Nielsen, Earnings

Core Question: Is $175B CapEx the next AWS moment, or the next Metaverse? Weight 15.5% — third-largest valuation driver. CapEx in this model simultaneously affects FCF (directly) and valuation multiples (indirectly: high CapEx = low FCF = investors demand lower multiples).

Why this is the biggest current debate: FY2025 CapEx of $91B is already massive. FY2026 guidance of $175-185B nearly doubles it — 60% servers, 40% data centers. At ~$200B operating cash flow, 2026 FCF will approach zero or go negative. D&A already up 38% to $21B/quarter and will "meaningfully increase" in 2026. Alphabet also raised $24.8B in debt (Nov 2025) and committed $16B to Waymo.

Bull logic chain: AWS's 2014-2017 CapEx surge was also questioned — it later became Amazon's profit engine. Key difference: Google has $240B in signed backlog = return visibility is clear. Gemini inference costs fell 78% in 2025 = investment efficiency is improving rapidly. Pichai: "The risk of underinvesting is dramatically greater than overinvesting." This isn't speculation — it's investment backed by signed contracts.

Bear logic chain: CapEx/Revenue has gone from 15% to 25% and is heading to 35%+. The key question: how much is "offensive" (Cloud expansion) vs "defensive" (Search AI transformation to counter ChatGPT)? Defensive CapEx doesn't create incremental revenue. If AI monetization lags CapEx burn, FCF will be negative for two consecutive years — for a $3T company, this is a serious signal of capital misallocation. Compare with Meta's Metaverse: $36B invested, zero revenue to date.

Alva judgment: Not whether CapEx is high (it definitely is), but CapEx quality — the offensive vs defensive ratio. Bear → continues rising to 30%+ = -$32/share (FCF + multiple double whammy). Bull → normalizes to 18% = +$19/share.Alphabet IR— Earnings Call, Macrotrends

: $126B of — Waymo Uber Segway?Alphabet 2026 year 2 month $126B $16B. $284M yearrevenue, 443x P/S — .

L4 vs L2+ are two fundamentally different approaches:Waymo operates L4 fully driverless (no one in the car), while Tesla FSD is L2+ supervised autonomy (requires human oversight). This isn't a tech gap — it's a different product. L4's barriers include: ①Regulatory approval is city-by-city ②Insurance pricing fundamentally favors fully autonomous vehicles on liability ③SF has 1,000+ fully driverless Waymo vehicles vs Tesla's 156 with safety monitors.

But the competitive landscape works against Waymo:①Tesla's million-vehicle fleet data flywheel — collecting hundreds of times more real-world driving data daily than Waymo's 3,000 vehicles. ②Per-ride cost: Waymo $17.25 vs Uber $14.94 vs Tesla $7.39 — 2x+ cost gap. ③Google doesn't make cars, relying on Zeekr/Jaguar manufacturing = no hardware scale advantage. ④New city expansion is extremely slow — Austin took 9 months to reach 8% of ride volume. ⑤The fundamental question: the industry may trend toward 'autonomy as a feature in every car' (Tesla/BYD/XPeng approach) rather than 'dedicated robotaxi fleets.' If that path wins, Waymo's asset-heavy model gets marginalized.

Alva judgment:L4 fully driverless safety and regulatory advantages offset Tesla data flywheel and cost advantages?Bear $20B(technology path loses)= -$5/share. Bull $160B(first-mover in $2T TAM robotaxi market)= +$7/share. Waymo Blog— Awisee, Om Malik, Automotive World

Core Question: Where are we in the AI investment cycle?Weight 16.1% (second-largest). This isn't about whether AI is useful — it's about whether current investment levels can generate proportionate returns. The answer affects GOOG's overall EV/EBITDA multiple.

Bubble signals — the data is striking: AI $1.5T(2025)→ $2.52T(2026E), +44%. Mag7 $680B CapEx . CapEx/Revenue . revenue: Sequoia Annual Revenue Gap $600B. MIT 2025: 95% AI ROI. McKinsey: 80% uses GenAI 90% usesat pilot. Deloitte: 2-4 year( 7-12 month of 3-4 ).

But this time is fundamentally different from 2000:The telecom bubble was debt-funded ($500B+, AT&T CapEx at 72% of EBITDA) — when it burst, debtors went bankrupt. Mag7 are self-funded from cash flow — even if returns disappoint, they won't go bankrupt. More critically: in 2000, the builders and operators were different entities (telecoms built the networks, internet companies used them). This time, the builders ARE the operators.

GOOG-specific:GOOG GOOG's AI CapEx has clear customer backing ($240B Cloud backlog) — this isn't blind construction. But at 19x EV/EBITDA vs peer median 13.5x, has the AI premium already been fully priced in?

Alva judgment:Current AI is the investment level sustainable?Bear → Bubble phase, multiples should revert toward peers (-8%)= -$28/share. Bull → Early innings, multiples have upside (+8%)= +$26/share. Sequoia— Gartner, MIT, McKinsey, Sparkline

Core Question: How much is the Apple $20B distribution deal risk worth?Weight 6.6%. The core of the DOJ ruling isn't about breakup (unlikely) — it's about whether exclusive default agreements can continue. This directly impacts Google's $20B+/year distribution payments and 60-80% of Search query volume on Apple devices.

What has already happened:2024.8 Judge Mehta Judge Mehta ruled Google illegally monopolized search. Sept 2025 remedies ruling (230 pages): ①6-year ban on exclusive default agreements ②Mandatory sharing of search index data with competitors ③Technical oversight committee established. Dec 2025 addendum: 1-year contract renegotiation limits. However — both sides appealed (Jan 2026), actual enforcement may be delayed to 2027-2028.

Quantifying the real financial impact:Google's internal estimate: losing Apple default = $28-33B revenue loss + 60-80% query volume loss on Apple devices. But this is the most extreme scenario. The EU DMA choice screen provides a reference: Google's European share fell from 87% to 78% (-9pp), but globally still at 89.6%. A choice screen doesn't mean going to zero.

What the market may be underestimating:Even if exclusive defaults are banned, Google remains the active choice of most users (EU data proves this). The $20B distribution fee saved would actually improve margins. The real risk is search index sharing — if ChatGPT can use Google's index for free, the competitive moat weakens significantly.

Alva judgment:enforcement intensity and timeline. Bear(strong enforcement)→ 6% risk discount = -$22/share. Bull(mild resolution/appeal succeeds)→ 0% discount = uncertainty premium release.DOJ Filing— NPR, CNBC, NYT

Core Question: Is the ad market pie still growing?Search + YouTube in GOOG revenue 63%. If the overall ad market contracts, even with stable share, Google's revenue declines. Weight 5.2%.

Structural tailwinds persist:Global ads surpassed $1T (2024, +9.5%), digital at 73% (2029E reaching 77%). Retail Media (Amazon Ads etc.) at $177B will overtake TV — incremental for Google Shopping. CTV shifting from TV budgets benefits YouTube.

An overlooked positive:Google cancelled Chrome cookie deprecation in July 2024. CMA found publisher revenue dropped -30% under Privacy Sandbox. Cancelling deprecation = preserving Google's ad targeting precision advantage. Walled gardens (Google/Meta/Amazon) are actually consolidating their position in the privacy era.

Risks:Ad spending elasticity is ~1.5x GDP — a 1% GDP decline means ~1.5% ad decline. If the US enters recession, Search and YouTube revenue will be directly impacted. AI-generated content may reshape ad distribution (users reading AI summaries instead of seeing ads).

Alva judgment:the next 2 Will the ad market expand or contract?Bear → Recession drags = Search/YT growth -2pp each. Bull → CTV + AI targeting acceleration = +1.5pp.IAB— GroupM, eMarketer, CMA

Core Question: When will AI revenue catch up with AI investment?Weight 6.1%. This is a different question from 'is AI a bubble' (dimension 6) — even if AI is real, the monetization pace determines Google's near-term ROIC and multiple justification.

The reality is harsh: CapEx/Revenue 6:1. Sequoia Annual Revenue Gap $600B. MIT: 95% Enterprises Zero AI ROI. McKinsey: 90% usesat pilot. Deloitte: 2-4 year( 7-12 month of 3-4 ).

But Google is closer to monetization than peers:Cloud AI revenue already accounts for >30% of Cloud total (growing >50%). Search AI (AI Overviews) is already generating ad revenue. Gemini API call volume is growing exponentially. $240B backlog = customers are already paying for AI. In contrast, Meta's AI investment ($36B Metaverse) has generated zero revenue to date.

Critical time window:If 2027 year AI revenue explodes(S curve inflection), then current CapEx ahead of schedule. If 2028 yearat pilot, then $360B+ investment(2025+2026)became . OpenAI 2026E revenue $12B+ — at, but supply is severely oversaturated.

Alva judgment:AI revenue S ofat?Bear(delayed)→ -3% = -$11/share. Bull (ahead of schedule) → +3% = +$10/share. Sequoia— MIT, McKinsey, Deloitte

Core Question: How much further can rates fall?Weight 4.4%. Tech stocks are long-duration assets — every 1% change in discount rate moves the theoretical valuation of a high-growth company like GOOG by >10%.

Current state (March 2026):Fed funds at 3.50-3.75% (held March 18). Dot plot median expects just 1 rate cut of 25bp. 14 of 19 members see 0-1 cuts. Core PCE revised up to 2.7% — inflation stickier than expected. 10Y Treasury ~4.35% (Middle East conflict + oil prices).

Transmission to GOOG:BlackRock confirms tech sector has the highest interest rate sensitivity across all sectors. Current ERP at just 5.0% (Kroll) — meaning even with rates unchanged, any earnings miss triggers a sharp valuation adjustment. GOOG at 25x PE sits at the 33rd percentile historically — if rates rise, multiples have room to compress.

Alva judgment:Will inflation remain sticky or subside?Bear(hawkish hold/hike)→ 4% discount. Bull(easing cycle begins)→ 0% discount, valuation expansion. CME FedWatch— Schwab, Kroll, BlackRock

Core Question: Has geopolitical risk already been priced in? 3.3%. Google 51% revenue from outside the US(EMEA 29%, APAC 16%). Geopolitical conflict impacts valuation through two channels: ①Direct hit to cross-border ad business ②Indirect lift in risk aversion and Treasury yield.

CurrentRisks:EU 2025 yearFine €2.95B(). DMA compliance costs keep rising — Search -9pp. : Google (/)not under export controls(vs NVIDIA/AMD ). Cloud APAC expansion faces policy uncertainty. Middle East conflict pushes up oil prices and 10Y Treasury, indirectly compressing tech valuations.

Google's unique situation:Already exited China search (2010) — the largest risk exposure is closed. India is the fastest-growing market but local competition (Jio) is intensifying. The primary risk isn't direct business impact, but risk-aversion sentiment transmission.

Alva judgment:Will existing geopolitical risks escalate or ease?Bear → 3% discount. Bull → 0%. Eurasia— Statista, EU Commission

Core Question: USD direction — tailwind or headwind?Weight 4.0%. Google derives 51% of revenue from overseas with no systematic FX hedging. Every 1% DXY move ≈ ~0.5% reported revenue change. FY2024 FX drag was approximately -1pp of growth.

DXY trajectory:2025 early in the year 110 → 10 month 97.7(-10%)→ Current ~100.5(). on a full-year basis the weak dollar is a GOOG tailwind for GOOG.

: MS/ING/MUFG Most are bearish on USD for 2026 — rate cut expectations + expanding US fiscal deficit + twin deficit concerns. But if the Fed stays hawkish (dot plot shows just 1 cut), the dollar could strengthen.

The transmission to GOOG is direct:51% Int'l Revenue × no hedging = FX flows directly through to financials. FY2024 growing 15% vs reported 14% — 1pp gap. Ifweakening 5%, 2026 reportedrevenuegets a ~$10B tailwind. Conversely ifstrengthening 5%, headwind ~$10B.

Alva judgment:2026 USD direction?Bear(strengthening)→ segment growth -1.2pp. Bull(weakening)→ +0.6pp. DXY— MS, ING, MUFG, IMF