Alphabet
The AI Winner, Re-Rated on Cloud

Executive Summary

Alphabet's Q2 CY26 print delivered ~$119.8B in revenue (+24% YoY)² and reframed the entire story. Google Cloud grew +82% to $24.8B with operating margin expanding to ~36% (from ~21%), backed by a $514B backlog. The investment frame is three segments, Google Services (~$320B, Search + YouTube), Google Cloud ($58.7B TTM, +82% in Q2), and Other Bets (~$2B with Waymo optionality)¹, and the two debates have shifted from antitrust to AI economics: (i) does the $195-205B 2026 capex ramp convert to high-return AI revenue rather than compressing free cash flow, and (ii) is the ~27x forward multiple, re-rated from ~22x, now fully pricing the Cloud and Gemini inflection. We answer "the backlog and margin trajectory support it" to the first and "earned, not stretched" to the second, which underwrites the Buy rating even after a near-doubling off the 52-week low.

We rate Alphabet Buy with a 12-month price target of $420, derived from a DCF / SOTP blend: segment-level discounted cash flow with independent revenue and margin trajectories for Services, Cloud, and Other Bets, cross-checked against a sum-of-the-parts that values Services on a mid-teens earnings multiple, Cloud on an EV/Sales that reflects its now-proven 30%+ operating margin, and Other Bets as option value. Bull $500 assumes Cloud sustains 40%+ growth off the $514B backlog at 30%+ margins, Gemini compounds enterprise and consumer share, and the capex converts to durable AI revenue. Bear $300 assumes the $195-205B capex out-runs AI-revenue conversion, Cloud decelerates toward the low-20s, and the AI-leadership shakeup slows the model roadmap, compressing the multiple to ~20-22x. Probability weights of 35 / 50 / 15 (Bull / Base / Bear) reproduce a blended fair value of ~$430, which we round to the $420 base PT after a modest haircut for capex-execution risk.

Tactical: GOOGL at $362 sits ~16% below our $420 12-month PT. Blended FV ~$430 on 35/50/15 weights. Rating Buy: at ~27x NTM EPS the multiple has re-rated to reflect the Cloud (+82%) and Gemini inflection, so the premium is now earned rather than cheap. The swing factor is whether the $195-205B capex converts to AI revenue. Add on Cloud-growth and margin prints; trim if capex steps up again without a matching revenue and backlog conversion.

Investment Thesis Summary

Overall Rating: Buy · 12-month PT $420 (Blended FV ~$430)

We rate Buy on Alphabet with a $420 12-month price target. The base case underwrites Google Cloud growing in the 30-40% range off a $514B backlog at 30%+ operating margins, Search ad revenue compounding as AI Overviews stay neutral-to-accretive to blended RPM, and Other Bets remaining option value rather than P&L contribution. At ~27x NTM EPS Alphabet has re-rated from the ~22x it traded at during the DOJ overhang, and now sits closer to the MAG7 cohort; we view the premium as earned by the Cloud inflection and the $514B of contracted demand rather than as multiple risk. The bear case is real and non-trivial: the $195-205B capex could out-run its AI-revenue conversion and compress free cash flow, and an AI-leadership shakeup adds roadmap risk, but the 15% weight reflects our reading of the backlog visibility and the self-funded balance sheet.

1 · Business Overview

What Alphabet Is

Alphabet reports under three segments that are deliberately asymmetric in current revenue contribution but reasonably balanced in terms of where consensus debate sits: Google Services at ~$320B (Search ads, YouTube ads, Network ads, Subscriptions / Platforms / Devices), Google Cloud at $58.7B TTM and growing +82% YoY in Q2 CY26, and Other Bets at ~$2B dominated by Waymo¹. The mix understates the strategic re-orientation that has been underway since the AI-platform transition began in 2023: Search has gone from "the only thing that matters" to "the thing that has to be defended while a second hyperscaler franchise scales," and Cloud has now inflected from a low-margin also-ran into a ~36%-operating-margin engine growing faster than any peer, which is the single change that re-rated the stock off its 52-week low.

Q2 CY26 Segment Breakdown

Key Franchises within Each Segment

• Google Search: the gravitational centre of the equity story. ~90% global Search share; the question is not share, it is whether AI Overviews change the per-query economics.
• YouTube: ~$44B annualised ad revenue (Q2 CY26 ads $11.06B, +12.6%) plus a growing subscription line (YouTube Premium + Music + TV). The TikTok comparison is real, but YouTube's CTV penetration (Connected-TV is now the majority of watch-time) is the differentiator.
• Google Cloud: the franchise that justifies the re-rating. GCP infrastructure + Vertex AI + Workspace; Q2 CY26 revenue +82% with operating margin expanded to ~36% (from ~21% a year ago) and a $514B backlog.
• Android + Play: mobile distribution moat reported inside Services as Platforms & Devices revenue. EU DMA implementation is the watch item.
• Waymo: not material to FY26 numbers but the cleanest public-market expression of full-stack autonomous-driving deployment economics.

Segment Growth & Margin Profile (Q2 CY26)

Inside Google Services

Services is the segment that funds everything else and the segment around which most of the bear narrative has converged. Search advertising, the largest franchise inside the segment at ~$224B TTM, is the asset where AI Overviews collide with classic Search RPM. The mechanical question: of the ~30% of queries that now surface an AI Overview at the top of the results page, what fraction of clicks (and therefore ad revenue) does the Overview substitute for, and what fraction does it augment by holding the user inside the Google session and driving incremental query volume from confident searchers? Our base case underwrites a neutral-to-slightly-positive blended RPM, with the AI Overview itself becoming a monetisable surface (sponsored answers, sponsored shopping carousels) over the FY26 disclosure cycle. YouTube, the second franchise in the segment, has now passed Netflix in measured CTV watch-time in the US and reports ad revenue of roughly $44B annualised (Q2 CY26 ads $11.06B, +12.6%) with a growing subscription contribution from YouTube Premium / Music / TV. The TikTok comparison preoccupies the sell-side but understates YouTube's structural advantage on long-form CTV; the competitive pressure is more focused on Shorts vs Reels vs TikTok at the format level and on creator-economics share. Network ads (programmatic via AdSense / DV360 / Google Ad Manager) is the slowest-growing franchise inside Services and the most exposed to identity / cookie deprecation. Subscriptions / Platforms / Devices runs at ~$46B annualised and is a high-margin sub-line, it grows fast inside Services on a percentage basis and is where the Gemini subscription tiers land.

Inside Google Cloud

Cloud is the segment that anchors the re-rating thesis. At $58.7B TTM and growing +82% YoY in Q2 CY26, Cloud is the fastest-growing hyperscaler franchise, and its operating margin has already expanded to ~36% (from ~21% a year ago), with a $514B backlog underwriting the forward trajectory. The debate is no longer whether Cloud is profitable, it is how much of that $514B converts to revenue and how durable the 30%+ margin proves as capacity scales. The segment decomposes roughly 70/25/5 between GCP infrastructure (the IaaS / PaaS layer where Vertex AI sits), Google Workspace (the productivity suite, direct competitor to Microsoft 365), and a long tail of vertical industry solutions. Vertex AI is the disclosed growth engine: a fully managed platform for training and deploying foundation models, including Gemini variants and a curated library of open-weight models, with consumption disclosed in management commentary as a key driver of Cloud growth. The TPU v6 / Trillium economics with Broadcom (covered in detail in the next section) are the gross-margin lever behind that margin step-up; absent the custom-silicon economics, Cloud margin would have plateaued and the re-rating thesis would be far weaker.

Inside Other Bets

Other Bets generates roughly $2B in revenue and runs at a meaningful operating loss, explicitly modelled by management as long-duration investment rather than as a near-term contributor. Waymo is the dominant story inside the segment: the autonomous ride-hail service has now scaled across Phoenix, San Francisco, and Los Angeles with a measured expansion path into additional markets¹², and the unit economics on incremental rides have crossed into positive territory on a contribution basis in the most mature markets. We frame Waymo as option value rather than as a number in the consolidated forecast, a probability-weighted addition to the SOTP if the service reaches ride-volume scale comparable to Uber/Lyft in a defined geography, but with zero credit in the base-case PT. Verily (life sciences), Wing (drone delivery), and Isomorphic Labs (AI drug discovery, leveraging DeepMind's AlphaFold work) round out the segment with measured optionality but materially below Waymo's economic potential.

2 · Cloud & AI: The +82% Inflection, Gemini, TPU v6, and the CapEx Question

The Alphabet thesis now collapses to two questions: (1) does the $195-205B 2026 capex convert to high-return AI revenue, and (2) is the Cloud + Gemini re-rating already fully priced at ~27x. The base case answers "yes, the $514B backlog and 30%+ margins support it" to the first and "earned, not stretched" to the second; the bull case strengthens both; the bear case answers "the capex out-runs the revenue" and "de-rate." Cloud grew +82% YoY in Q2 CY26² to $24.8B in segment revenue with operating margin ~36%, management attributing the strength to "broad-based growth in core GCP plus accelerating Vertex AI and Gemini consumption from enterprise customers."

The Gemini Model Family & Consumer Share

Gemini is the consumer-facing AI surface and the model family that powers Vertex AI inside Cloud. The lineup spans Gemini 2.5 Pro (the flagship multimodal model), Gemini 2.5 Flash (the cost-optimised inference SKU), and Gemini Nano (on-device, Pixel + Android distribution). Consumer assistant share against ChatGPT is the relevant battlefield; third-party panel data triangulates Gemini at roughly 12-15% of consumer assistant queries against ChatGPT's ~55%, with the remainder distributed across Claude, Perplexity, Copilot, and the long tail. Our base case underwrites Gemini holding ~15% share, consistent with management's framing on the Q1 CY26 call of Gemini consumer momentum and AI Overviews monetisation testing³, with the Workspace Gemini Ultra SKU providing the B2B revenue layer; bull case is 25% on better Gemini 3 generation reception, bear case is 8-10% on ChatGPT consolidating consumer default behaviour.

TPU v6 / Trillium & the Broadcom Partnership

Google's TPU programme, co-designed with Broadcom (AVGO) on the silicon side¹¹, is the single most underappreciated unit-economics story among the megacaps. The current generation, TPU v6 / Trillium, delivers (based on disclosed performance benchmarks and Google's own training-throughput claims) roughly 2x perf-per-dollar versus an H200-class GPU for transformer training workloads at scale, with the gap widening for inference. The economics are sharper than the perf-per-dollar number suggests once you factor in: (i) Google pays AVGO a design / IP fee on the silicon plus a per-chip royalty, rather than paying NVIDIA's gross margin on top of a fabless-merchant model; (ii) TPU pods are tightly integrated with Google's data-centre networking fabric, which removes the inter-node bandwidth tax that affects rented GPU clusters; and (iii) the inference workload mix on Vertex AI runs predominantly on TPU, which compounds the cost advantage at the segment-margin level. The strategic value: as the internal-mix share of training compute on TPU grows above ~60% in the FY27 base case, the segment-level gross margin steps up by a multi-hundred-bp amount, and the operating margin moves toward the mid-teens path we underwrite.

AI Overviews: the Search Cannibalisation Debate

AI Overviews is the AI-generated summary that now surfaces at the top of roughly 30% of Search results pages for queries where Google's models judge the answer to be summarisable with confidence. The disclosed monetisation status as of Q1 CY26: the Overview itself is currently not directly monetised: there are no sponsored answers, no sponsored placement inside the AI summary box, though sponsored shopping carousels and standard text ads still surface above and below the Overview on the same page. The bear argument: the Overview gives the user a confident answer without a click, so the click-through rate on the surrounding ads drops, the auction depth thins, and the per-query RPM compresses. The bull argument: the Overview holds the user inside the Google session rather than sending them to ChatGPT or Perplexity, increases the confident-search use case (which expands query volume on the relevant intents), and provides a future inventory surface (sponsored answers, sponsored shopping) that will be monetised over the FY26-FY27 disclosure cycle. Our base case: blended RPM is neutral-to-slightly positive in FY26 with the inventory expansion offsetting the click-deflection drag, and turns clearly positive in FY27 as the sponsored-answer surface goes live.

Vertex AI Growth & DeepMind Production Footprint

Vertex AI is the disclosed managed-AI platform inside Cloud and the segment line that has accelerated most visibly in FY25-FY26. Management commentary on the Q1 CY26 call attributed a meaningful portion of the +82% segment growth to Vertex AI consumption from both enterprise customers (financial services, retail, healthcare) and from external consumer-AI workloads, including, notably, a portion of OpenAI's inference that runs on Google Cloud TPU under a separate commercial agreement. DeepMind, the AI research arm, has now had production deployments cross from research milestone to revenue contributor: AlphaFold 3 is licensed through Isomorphic Labs and the Vertex AI Biology workbench; Gemini's reasoning architecture is DeepMind-derived; and DeepMind talent is now embedded in product teams across Search and Cloud. The strategic implication: Alphabet's research-to-product distance has shortened materially in the past 18 months, and the FY27 base case underwrites that shorter loop translating into measurable revenue.

Alphabet spent ~$85B on capex in FY25 and has raised its FY26 guide to $195-205B⁴, heavy on AI data-centre build, TPU procurement (the AVGO partnership flows through capex), GPU procurement for the customer-funded NVIDIA capacity inside Vertex AI, and grid-power infrastructure. This is the single biggest swing factor in the thesis, but it is a question of return, not affordability: on ~$200B+ of annual operating cash flow, a ~$95B cash position, and only ~$28B of debt, Alphabet self-funds the build without leverage stress (though it compresses free cash flow to a low level while capex peaks). The debate is whether that spend out-earns its cost as Cloud capacity comes online.

3 · Financial Health

Quarterly & Annual Summary

Q2 CY26 Print Recap & FY26 Guidance Posture

The Q2 CY26 print delivered ~$119.8B in revenue (+24% YoY), comfortably ahead of consensus, with the headline being Google Cloud at +82% to $24.8B and Cloud operating margin expanding to ~36% (from ~21% a year ago). Consolidated gross margin held at ~58%, a signal that the AI-mix shift has not shown up as a measurable gross-margin drag at the consolidated level, the TPU economics are doing real work here, since the same revenue mix on a pure NVIDIA-GPU stack would have compressed segment GM by ~150-200bps. Operating margin landed at ~34%. The single most important change was on capital: management raised the FY26 capex guide to $195-205B (from ~$85B spent in FY25), an increase large enough that free cash flow compresses to a low level in FY26E even as operating cash flow keeps growing. The market read the Cloud acceleration and the $514B backlog as validation that the spend is demand-backed, which is what drove the re-rating; the open question the guide raises is FY27, where management has not yet set a ceiling.

Calendar-Year Reporting & Quarterly Seasonality

Alphabet reports on a calendar-year fiscal basis, which matters for how the quarterly seasonality maps to consumer behaviour. Q4 (October-December) is the largest revenue quarter, the holiday-shopping period drives Search ad revenue and YouTube ad revenue simultaneously, typically running 8-12% above the trailing-three-quarter average. Q1 (January-March, the most recent print) is the cleanest read on underlying ad-market momentum: post-holiday consumer normalisation, lower seasonality noise, and the period that most accurately reflects the AI-Overviews monetisation trajectory because the Overview surface is fully deployed and the click-rate data is reportable on a like-for-like basis. Q2 and Q3 are typically the steady-state quarters where management commentary on segment trajectories carries the most predictive weight. The implication for our model: a beat in Q1 is a higher-quality signal than the equivalent beat in Q4 would have been, and the Q2 CY26 print delivered the Cloud acceleration and $514B backlog the market had been waiting for.

FCF Bridge & CapEx Intensity

The free-cash-flow trajectory, $69.5B (FY23) → $72.8B (FY24) → ~$65B (FY25) → ~$15B (FY26E): shows the sharp compression that the capex step-up forces. The mechanical bridge for FY26E: operating cash flow tracks toward ~$215B on $478B of revenue (depreciation is climbing fast as the capex layers land), but ~$200B of gross capex leaves only ~$15B of reported FCF. This is the clearest single number behind the bear case, Alphabet is now spending nearly all of its operating cash flow on the AI build. The crossover comes in FY27-FY28: gross capex begins to plateau in our model while the FY24-FY26 layers amortise into higher depreciation, and FCF re-accelerates back above $80B by FY28 in the base case. The base case tolerates the FY26E compression because it is a reinvestment choice against a $514B backlog, not a demand or margin problem; a bear who assigns a low return to the incremental capex reads the same compression as value destruction.

Operating Leverage at ~33% Margin

Operating margin has expanded meaningfully from the FY23 trough, ~27% (FY23) → ~32% (FY24) → ~33% (FY25): and Q2 CY26 indicates the margin holding at ~34%. Three dynamics layered together to produce this: (i) the early-2023 cost-discipline programme rationalised headcount by ~12% and put hard caps on OPEX growth, which directly added several hundred basis points of margin without revenue loss; (ii) Cloud crossed into structural profitability in FY24 and is now an accretive contributor to the consolidated margin rather than a dilutive drag; and (iii) the AI mix shift has so far been net-neutral on consolidated gross margin, with TPU economics offsetting the would-be GM drag of higher AI inference workloads. The path forward depends on Cloud margin durability: Cloud operating margin has already reached ~36%, and the base case underwrites it holding in the 30s as capacity scales, which sustains consolidated margin around ~34-35%. Bears argue Cloud margin gives back gains as low-utilisation GPU capacity ramps ahead of revenue; bulls argue TPU v6 + Vertex AI scale push it higher still.

Balance Sheet

Alphabet ends Q2 CY26 with approximately ~$95B cash & marketable securities against ~$28B total debt: a net-cash position of ~$67B, among the strongest balance sheets in the hyperscaler cohort and a structural feature of the equity story that lets it fund the $195-205B capex without leverage stress. Credit ratings remain AA+ at S&P. The capex cycle does not stress the balance sheet at any plausible scenario in our horizon; the binding constraint on capital deployment is opportunity cost on the cash, not access to it. The historical posture has been to under-lever, Alphabet operates with materially less debt than MSFT (~$50B) or AMZN (~$60B) at comparable revenue scale, which we read as a management preference for optionality on transformative M&A (HubSpot rumours surfaced in 2024 and resurfaced periodically since), on settlement / remedy outcomes from the DOJ Search case, and on prospective Other Bets investment as Waymo scales.

4 · Capital Allocation

Alphabet's capital allocation in the current cycle reflects a notable shift: the introduction of a regular quarterly dividend in 2024: Alphabet's first ever, alongside a buyback that ran at approximately ~$75B in FY25. The combination is unusual at this scale: most megacaps in Alphabet's growth profile have historically run buyback-only programmes, and the dividend introduction was widely read as a signal that management views the FCF base as durable enough through the AI-investment cycle to commit to a recurring distribution. The yield is modest (~0.4% at spot) and the absolute pay-out is small relative to the buyback (~$10B annualised dividend vs ~$75B buyback), but the signalling weight is meaningful: dividends are sticky in a way buybacks are not, and the existence of one anchors a different class of long-duration ownership.

M&A has been more measured than Alphabet's balance sheet would permit. The cleanest historical reference is DeepMind (2014, ~$500M), strategically transformative at a price that was small even at the time. The 2024 HubSpot rumours did not materialise into a transaction, in part because the regulatory friction (DOJ Search case still open, EU DMA enforcement live) would have made a transformative deal impractical. We do not model a mega-deal in our base case; capability tuck-ins (AI talent, vertical SaaS, autonomous-driving adjacencies) are the steady-state cadence.

Lifecycle & the CapEx-vs-Buyback Trade-off

The CapEx-first stance reflects where Alphabet sits in its lifecycle: a mature compounder with a generationally good reinvestment opportunity inside Cloud and Search-AI. At $362 and a ~27x NTM multiple, the buyback is less mechanically accretive than at the ~22x Alphabet traded at during the DOJ overhang, but the share-count reduction still runs ~3% annually, above MSFT's ~1% and consistent with META's pace. Every incremental $1B of capex directed at TPU capacity that gets absorbed inside Vertex AI at the segment's incremental gross margin compounds inside the business at a high-teens to low-20s unlevered return, meaningfully above Alphabet's ~9% cost of capital. The trade-off is rational: the capex is the highest-return use of cash while Cloud growth is above 25%, and the buyback re-accelerates further if the cycle plateaus.

Dividend Introduction as Signalling Mechanism

The dividend deserves more weight than the ~0.4% yield suggests. Initiating a recurring distribution in 2024 was the cleanest signal management could send about the durability of free cash flow through the capex cycle: the boards of buy-and-hold megacap companies rarely introduce dividends opportunistically, and almost never reverse them. The signal is that management views ~$80B+ of FCF as the durable forward base, even with capex stepping toward $195-205B. The follow-up signal, whether the dividend gets raised in 2026 / 2027, and at what pace, will be a high-conviction read on what the board expects FCF to do once the FY26 capex peak passes through. We model a modest ~10-15% annual raise consistent with peer practice; an acceleration above 20% would signal real confidence in the FY27-FY28 FCF re-acceleration.

Stock-Based Comp & Dilution Offset

Stock-based compensation at Alphabet runs at roughly ~6-7% of revenue: similar to MSFT in percentage terms and notably above the historical Alphabet base, reflecting the post-2022 SBC normalisation across the megacap software cohort. The gross dilution from RSU vests and ESPP averages 1.2-1.5% of basic shares outstanding per year; the FY25 buyback at ~$75B is offsetting that issuance by a multiple of three on a net basis, producing a meaningful per-share tailwind. On a strict valuation framework where SBC is treated as a real economic cost, Alphabet's genuinely incremental shareholder return is the dividend plus the residual buyback after offsetting issuance, closer to ~2.5% of market cap annually at current pace. That number compares favourably to MSFT's ~1.5-2% on the same basis and is one of the underappreciated structural features of the equity story.

M&A Discipline & the HubSpot Question

Alphabet has the balance-sheet capacity to absorb a $50-100B+ acquisition, the question is whether the regulatory environment permits one. The DOJ Search case and the EU DMA enforcement track combine to make a horizontal acquisition in adjacent advertising / consumer-software markets effectively impossible until the remedies framework is resolved. Vertical or capability-adjacent transactions remain available (the HubSpot CRM rumour fits this category, outside Search, outside the DOJ frame), but the bar for management discipline is high and we model nothing transformative in the base case. The optionality is real, however, and a clean DOJ remedy resolution combined with a strategically logical HubSpot-scale transaction would meaningfully change the FY27-FY28 revenue and segment-mix path.

5 · Valuation & Comps

Peer Comparison

Where the Multiple Sits in the MAG7 Cohort

The headline number, ~27x NTM P/E vs MSFT ~25x, META ~25x, AMZN ~33x, AAPL ~29x: puts Alphabet roughly mid-cohort after a re-rating from the ~22x it traded at during the DOJ overhang. The discount that defined the story a year ago has closed, and for a defensible reason: Google Cloud is now growing +82% with a $514B backlog and ~36% operating margin, the fastest-growing large-cap cloud franchise, and the DOJ Search remedy landed as conduct-based rather than structural. The question is no longer "why the discount" but "is the re-rating fully paid for." We think the premium to MSFT is earned by Cloud's faster growth and the shorter AI research-to-product distance, but we do not assume further multiple expansion in the base case, the upside from here is earnings-driven, with the capex-to-AI-revenue conversion the swing factor.

AI Mix Shift & Margin Math

The single biggest concern beneath the multiple is no longer DOJ but whether the $195-205B capex out-earns its cost. On margins, the AI mix shift has not compressed consolidated gross margin (held ~58%), because the TPU v6 / Trillium economics with AVGO close most of the gap that pure-GPU workloads would open, and Cloud operating margin has actually expanded to ~36%. The arithmetic that supports the multiple: revenue growth of ~19-24% with operating margin holding ~34% produces high-teens EPS growth, and at ~27x that is a reasonable, not stretched, multiple. The risk is on free cash flow, not earnings: the capex ramp compresses FCF to a low level in FY26E, so a bear who assigns a low return to the incremental spend can argue the earnings are lower-quality than the multiple implies.

Contrast vs MSFT & AMZN

The cross-megacap comparison now frames GOOGL as fairly valued rather than cheap. MSFT trades at ~25x NTM on +15% revenue growth with Azure reaccelerating to +43%⁸; AMZN at ~33x pays for the AWS franchise plus retail-leverage upside⁹. Alphabet at ~27x on ~19% revenue growth sits mid-cohort, and the case for it is that its Cloud franchise is growing faster than AWS in percentage terms (+82% vs ~20%) at a ~36% operating margin; that it retains a strong net-cash balance sheet even through the capex ramp; and that its AI research-to-product distance is the shortest in the cohort. What it no longer has is a visible discount to close, so the return from here is earnings-driven: high-teens EPS growth at a stable ~27x supports the $420 PT, with further multiple expansion the bull-case optionality rather than the base.

SOTP Framing

An SOTP view supports the consolidated multiple. Working from approximate segment economics (segment detail per Alphabet's SEC filings)⁷: Google Services at ~$150B of steady-state EBITDA on a ~19x multiple (mature high-margin ad-franchise economics) contributes ~$2.9T of enterprise value; Google Cloud, now at a ~$100B run-rate growing +82% at a ~36% operating margin, supports ~$2.0T on a growth-adjusted basis (a mid-single-digit EV/Sales, well below where a pure-play at this growth would trade); Other Bets carries a placeholder ~$50B option value mostly attributable to Waymo. The summed enterprise value of ~$5.0T plus ~$67B net cash equates to ~$5.1T equity value, or roughly ~$415 per share, consistent with our $420 base PT and the ~$430 blended FV. The PT is built up from this SOTP and cross-checked against the segment DCF in the next section (base case ~$405).

Consensus

5b · DCF Model

The DCF projects each segment independently across a 5-year explicit horizon, then converges to a consolidated terminal value. Google Services grows high-single-digits reflecting Search plus YouTube CTV plus Subscriptions / Devices. Google Cloud grows ~44% in FY26E tapering toward the low-20s by FY30 as the base scales off the $514B backlog. Other Bets grows from a small base with measured Waymo monetisation. The model works in EBITDA terms (well above operating margin because depreciation on the ~$200B/yr capex is large): Services ~47% EBITDA, Cloud ramping from ~45% toward ~57% as scale leverage builds. Capex peaks near $195-205B in FY26E and tapers as a share of revenue across the horizon, so consolidated free cash flow compresses in FY26E and re-accelerates as capex intensity normalises.

Sensitivity: WACC × Terminal Growth

6 · Catalysts (Next 12 Months)

What the Q2 CY26 Print Changed

The Q2 CY26 print in late July was the pivot for the stock. Three things reset the framing. First, Google Cloud accelerated to +82% ($24.8B) with operating margin expanding to ~36%, and the $514B backlog gave the market confidence the growth is demand-backed, that is what drove the re-rating. Second, management raised the FY26 capex guide to $195-205B, which shifted the debate from antitrust to whether the spend out-earns its cost and compressed near-term free cash flow. Third, the consolidated print (+24% revenue, ~34% operating margin) confirmed Search and YouTube remain resilient alongside the Cloud inflection. The next checkpoint is the Q3 CY26 print (~late October), where the market watches Cloud growth and margin together, any early FY27 capex framing, and the pace of backlog conversion.

DOJ Search Remedy: Now an Appeals Tail, Not an Overhang

The DOJ Search remedy, which dominated the stock a year ago, has largely resolved along the conduct-based path rather than the structural one⁵: restrictions on default-search payments and data-sharing requirements rather than a Chrome or Android divestiture. What remains is a multi-year appeals tail that delays and could modestly reshape implementation through 2027-2028. That is a diminished but non-zero overhang: it is largely priced out of the multiple now, and the base case treats it as a manageable conduct framework. The tail risk, a structural remedy resurfacing on appeal, is low-probability and sits in the bear case; it is no longer the swing factor it was, which is one reason the multiple re-rated.

Product Cadence & Regulatory Schedule

The product cadence for the rest of 2026, Cloud Next in September 2026 (enterprise-facing, Vertex + Workspace AI SKUs) and the Gemini 3 rollout through H2 2026, defines the AI-franchise narrative into year-end. Gemini 3 reception will set the framing for the consumer-share narrative against ChatGPT and Claude, and the Vertex AI updates will define the enterprise Cloud narrative and the pace of backlog conversion. Layered on top is the regulatory schedule: the EU DMA enforcement milestones are still resolving (self-preferencing remedies, Play-store sideloading economics, browser-choice screens)⁶, and the EU AI Act GPAI compliance window opens in August 2026 with Gemini directly in scope. Management's posture will be to demonstrate compliance early to remove enterprise procurement friction, but the implementation cost is real and the EU revenue base (~20% of consolidated) makes the schedule worth tracking.

7 · Risk Factors

8 · Technicals

Key Levels

The AI Re-Rating and What It Tells You

The 12-month price path traces the AI re-rating. The stock based near its $194 52-week low in late 2025 as the DOJ overhang and a heavy capex guide weighed, then climbed steadily through early 2026 as Google Cloud accelerated and Gemini enterprise adoption compounded, reaching an all-time high near $402 in May 2026. It has since pulled back to ~$362 on the raised $195-205B capex guide and an AI-leadership shakeup, a healthy consolidation after a near-doubling rather than a trend break. The implication: the market has re-rated GOOGL from a legally-overhung compounder to a front-line AI winner, and the debate has shifted from "is the discount justified" to "does the capex out-earn its cost." The base-case PT of $420 implies ~16% upside from spot, a normal-for-megacap gap that reflects a fairly-valued name with earnings-driven upside rather than a large embedded discount.

Level Confluence & Position Sizing

The chart levels map cleanly to the scenario PTs. The $409 all-time high is the first milestone on the path to the base-case $420 PT; a clean breakout above $409 on volume would confirm the market is paying up for the capex-to-AI-revenue conversion and put the bull-case $500 in play. On the downside, the ~$318 200-day EMA region is the tactical support, a break would imply a Cloud-growth or capex disappointment the full ~27x multiple cannot absorb, and would open the path toward the bear-case $300. The ~$312 YTD base is the pivot in between. Sizing-wise, the implied tactical R/R from spot (~$58 reward to the PT vs ~$44 risk to the 200-day EMA) is roughly balanced, consistent with a fairly-valued name rather than the deep-discount setup of a year ago. Investors building exposure should treat pullbacks toward the $330-345 zone as the accumulation window and keep the position sized for a full multiple.

★ Interactive Tools

9 · Glossary