Alphabet's Class C shares give you the same economic ownership of Google Search, YouTube, and Cloud as the voting GOOGL — for an effectively identical price. At $383 against a base IV of $513, you are buying a 16.8% ROIC compounder at roughly 75 cents on the dollar.

Plain English

Google sells ads next to answers. When you search, ask YouTube, or use Gmail, Google shows ads. Advertisers bid in an auction and Google takes a cut of every winning bid. Because nearly everyone uses Google to find things, advertisers must pay Google to reach them. The same loop runs on YouTube videos, Android phones, and Maps. Google also rents computers to businesses (Cloud) and is building AI assistants. The owner of Class C shares like GOOG owns the same slice of profits as voting shares, just without a vote.

Thesis

Alphabet Class C (GOOG) is the same business as GOOGL — Google Search, YouTube, Android, Chrome, Workspace, Cloud, and the Other Bets — without the founders' voting rights. Economically the two share classes are identical: same dividend, same buyback eligibility, same claim on the $122.7B in TTM owner earnings. The non-voting structure historically traded at a small discount to GOOGL but the discount has effectively vanished as buybacks consume Class C shares disproportionately and index inclusion drives demand.

The investment case is unchanged from GOOGL. A 10-year average ROIC of 16.8% on a capital base now measured in hundreds of billions is rare. ROIIC of 26.9% over the last five years is more interesting still — Alphabet's incremental capital, much of it AI infrastructure, is earning above its trailing average. FCF conversion of 89% confirms the earnings are real cash, not accounting fiction. Net debt to EBITDA of 0.31x is a rounding error.

The scorecard pegs intrinsic value at $345 (low) / $513 (base) / $555 (high). At $383.22 the shares trade at 0.75x base IV — meaningful margin of safety despite a 42x TTM P/E that looks expensive in isolation. The reverse-DCF only requires 10.3% growth to justify today's price; consensus and the 5-year ROIIC suggest more.

The choice between GOOGL and GOOG is largely a coin flip — pick whichever trades cheaper on your fill. Today: same thesis, same target, same conviction. Buy with margin of safety; trim only above bull-case IV.

Moat

Alphabet possesses one of the widest moats in public markets, anchored by network effects, intangibles, and cost advantages that compound on each other.

Pricing power (intangibles + network effects). Google Search is the global default for intent-driven discovery. The advertiser auction is a two-sided network: more queries attract more advertisers, which funds better infrastructure, which attracts more queries. The brand itself is a verb. Damodaran observes that brand value is the consequence of relentless investment, not the cause [1] — Alphabet has spent two decades making 'google it' synonymous with 'find out.' Pricing power shows up in cost-per-click resilience: even as click volumes fragment across TikTok, Amazon, and AI assistants, blended CPMs continue to rise because the auction matches commercial intent better than alternatives.

Switching costs. Damodaran's Microsoft example [2] applies directly. Workspace (Gmail, Docs, Drive, Calendar) embeds itself in personal and corporate workflow. Android ties three billion device users to Google Play, Google Pay, and Google account services. YouTube creators have built channels with subscriber lists and monetization histories that are non-portable. Chrome user profiles, sync, and saved passwords are sticky in the same way Excel files were sticky in 1995. None of these are absolute lock-ins, but the cumulative friction of leaving the Google ecosystem is enormous.

Network effects. Beyond Search's two-sided auction: YouTube has the largest video creator base, the largest video viewer base, and the best recommendation engine — a three-sided flywheel that has resisted assault from Facebook Watch, Quibi, and TikTok (the latter being a real but partial threat). Android is the dominant non-Apple mobile OS globally, with developer tools, app distribution, and OEM relationships that no challenger can replicate from scratch. Waymo, while small, has the largest real-world autonomous miles dataset.

Intangibles. Twenty-plus years of search query logs, click data, and result quality signals constitute a proprietary asset competitors cannot replicate. Google's research org has produced the Transformer paper, AlphaFold, and Gemini — the talent density itself is a moat. Patents around PageRank, ad auction mechanics, TPU design, and many areas of ML provide secondary protection.

Cost advantages. Custom TPUs deliver inference and training at lower cost per FLOP than commodity GPU buyers. Owned global fiber, undersea cable investments, and hyperscale data center footprint give Alphabet a structural cost edge over any new entrant in either search or AI. The capex base is itself a moat — a competitor would need to spend hundreds of billions to match the infrastructure.

Competitor stress test ($10B + 5 years). Microsoft has spent more than $10B and five years (Bing, Cortana, now Copilot/OpenAI partnership) and remains a single-digit Search share player outside enterprise. Meta has tried multiple times to enter Search and abandoned each attempt. Amazon dominates product search on its own platform but cannot dent general web search. The most credible threat — generative AI assistants — is one Alphabet itself is leading with Gemini, defending share with AI Overviews while monetizing via the same auction.

Erosion risks. Three are real. (1) AI-mediated search reduces clicks per query, compressing ad inventory even as engagement rises. (2) DOJ remedies in the Search monopoly case could force divestiture of Chrome or end the Apple default-search payment, both of which would dent Search share at the margin. (3) Regulatory action in the EU and elsewhere keeps chipping at Android, app-store economics, and ad-tech.

Verdict. The moat is not a single fortress but a layered ecosystem where each layer reinforces the others. Even credible 20% impairments to Search would leave a business with structural advantages most companies will never possess.

Moat verdict: WIDE

Management

Alphabet's capital allocation has shifted meaningfully over the last five years, and this matters for the GOOG/GOOGL holder because both share classes participate identically in those decisions.

Reinvestment. This is the dominant use of capital, and the most consequential. Capex has stepped up dramatically — well over $50B annually — funding TPUs, GPUs, data centers, undersea cables, and AI training infrastructure. The 5-year ROIIC of 26.9% suggests the incremental capital is earning well above the cost of capital so far. The honest caveat: AI capex is front-loaded and the back-end revenue depends on whether Gemini, Cloud AI, and AI Overviews monetize as expected. The scorecard correctly flags that maintenance capex is uncertain (>50% spread) — distinguishing 'keep the lights on' capex from growth capex inside Alphabet's reported number is genuinely hard, which is why the IV range is wide.

Acquisitions. Alphabet's M&A track record is mixed-to-good. The home runs (YouTube $1.65B in 2006, Android $50M in 2005, DoubleClick $3.1B in 2007) are some of the best acquisitions in tech history. The misses (Motorola Mobility, Nest at the prices paid, various smaller bets) are real but small relative to Alphabet's size. The Mandiant ($5.4B) and Wiz acquisitions are sensible cybersecurity tuck-ins for Cloud. There is no sign of empire-building or transformational M&A done at top-of-cycle prices.

Debt. Net debt to EBITDA of 0.31x is the balance sheet of a company that simply does not need leverage. The senior notes outstanding (2.375%-4.375% coupons across 2028-2064 maturities, per the 10-K cover) are opportunistic tax-arbitrage borrowings, not financing necessities. The interest coverage is unmeasurably high.

Buybacks. This is where capital allocation has visibly improved. Alphabet now repurchases tens of billions of dollars of stock annually, predominantly Class C (GOOG) because Class B founder shares are not in the float and Class A is only partially. The mechanical effect is that GOOG share count shrinks faster than GOOGL, structurally narrowing any historical discount. The harder question is the average price/IV at which buybacks are executed. Through 2022-2024 buybacks were happening at P/IV ratios well below 1.0 — value-accretive. At today's 0.75x P/IV they remain accretive. The risk is that Alphabet keeps the buyback running into a regime where shares trade above IV, which has happened in the past at peak multiples.

Dividends. Alphabet initiated a modest cash dividend in 2024 — a meaningful cultural signal that management acknowledges some of the cash flow is genuinely surplus. The yield is small but the message matters: leadership is not pretending every dollar can earn 25% inside Alphabet.

Communication. Sundar Pichai's letters and earnings commentary are clear and substantive but lack the candor of, say, Buffett or Bezos. Strategic shifts (the AI 'Code Red' of 2023, the workforce restructuring, the 'efficiency' theme) have been communicated reactively rather than proactively. Segment disclosure improved with the YouTube, Cloud, and Other Bets breakouts but Cloud unit economics remain harder to dissect than they should be. The dual-class structure with Page/Brin retaining control via Class B is a governance demerit — GOOG holders specifically have zero voting input — but this is a known and priced feature, not a surprise.

Net assessment. Reinvestment quality is high (ROIIC 26.9% confirms it). Buybacks are sized appropriately and have historically been done at sensible prices. The dividend initiation reflects mature judgment. The capex pace requires monitoring but is not yet imprudent. The dual-class structure is the principal blemish.

Capital allocator: B+

Industry

Porter's Five Forces applied to Alphabet's primary business (digital advertising, with secondary exposure to Cloud and devices):

Threat of new entrants — LOW. The capex required to build a competitive search index, ad auction system, and global infrastructure is on the order of $100B+ over many years, and that is before considering the data assets and ML talent. The two-sided network effect (advertisers, users) means a new entrant cannot acquire either side without already having the other. Microsoft, with effectively unlimited resources, has tried for two decades and remains a small-share niche player. The only meaningful new entrants in the last decade have been adjacent (TikTok via short-form video, Amazon via product search) rather than head-on. AI-native search assistants (Perplexity, ChatGPT) are the most credible new threat in twenty years, and even they monetize at a fraction of search rates.

Bargaining power of suppliers — LOW to MODERATE. Key 'suppliers' include device OEMs (Samsung, OEM Android partners), telecom infrastructure providers, semiconductor foundries (TSMC for TPU fabrication), content creators (YouTube), and Apple (the default-search agreement). The Apple TAC payment is the most consequential single supplier relationship — roughly $20B+ annually — and is currently under regulatory threat. Otherwise, Alphabet's scale gives it strong purchasing leverage.

Bargaining power of buyers — LOW (advertisers) / LOW (users). Advertisers as a class are highly fragmented and have nowhere with comparable ROAS to migrate billions of dollars of spend. Individual large advertisers can negotiate, but they cannot leave. End users do not pay, so their bargaining power manifests through attention rather than price — and attention has remained sticky despite TikTok's rise.

Threat of substitutes — RISING. This is the force that has changed most. Search query intent is migrating in three directions: (1) directly to AI assistants (ChatGPT, Claude, Perplexity), (2) to vertical platforms (Amazon for products, TikTok for entertainment/discovery, Reddit for community knowledge), (3) to AI-mediated experiences embedded in OS and apps. Alphabet's defense is to be the AI provider itself (Gemini), embed AI in its own search results (AI Overviews), and continue monetizing the underlying intent. The substitution risk is real but the monetization advantage of intent-matched search remains durable.

Industry rivalry — MODERATE. Within search, rivalry is muted (Google has 90%+ share globally). Within digital advertising more broadly, rivalry is intense — Meta, Amazon, TikTok, and the long tail of programmatic compete for every marginal dollar. But Alphabet's piece of the pie has remained roughly stable while the pie has grown. In Cloud, rivalry is fierce (AWS, Azure) and Alphabet is the #3 player, though it is gaining share.

Value pool location and trajectory. The global digital ad market continues to grow mid-single-digits to low-double-digits annually, with AI shifting some intent away from traditional search but creating new monetization surfaces. Cloud is a structural growth pool with long runway. The AI infrastructure layer (where Alphabet's TPUs and models live) is the most valuable new pool of the last decade. Alphabet sits at or near the center of all three.

Verdict. The industry structure is exceptional for the incumbent — high barriers, weak buyer/supplier power, low rivalry within Search — but the substitute threat from AI is the highest it has been since Alphabet was founded.

Industry Verdict: Excellent

Inversion

I am now the short-seller. The bull case is wrong, and here is why.

The single event that kills this. A DOJ remedy that forces Alphabet to either (a) divest Chrome, (b) end the Apple default-search payment, or (c) license Google's search index to competitors. Any one of these is a regulatory event that has been actively litigated, not a hypothetical. Chrome divestiture removes the funnel that delivers the highest-intent queries to Google. Ending the Apple TAC arrangement opens the door for OpenAI/Microsoft to bid for the iPhone default — and a Bing-with-ChatGPT default on iOS is the closest thing to a Search-share extinction event imaginable. Index licensing commoditizes the asset that took twenty years to build. Bulls assume remedies will be cosmetic. The judge has already ruled Google is a monopolist. History (Microsoft 2001, AT&T 1984) suggests structural remedies are very much on the table.

Why the moat is narrower than bulls think. Search has one job: route intent to the answer with the highest probability of satisfying it. For two decades that meant ten blue links and an ad auction. Generative AI changes the job description. The new job is to synthesize the answer directly — and at that job, Google has no structural advantage over OpenAI, Anthropic, or any well-capitalized AI lab. Twenty years of click data are valuable for ranking links but only modestly valuable for training a foundation model trained on the open web. The 'data moat' bulls cite is mostly a moat for the old job, not the new one. Worse, the new job has worse unit economics: synthesizing an answer with a frontier model costs orders of magnitude more compute per query than ranking links, while displaying fewer ads per query because the answer leaves less room for sponsored placements. Even if Google retains share, it is retaining share in a less profitable business.

Why management is worse than it appears. Alphabet was caught flat-footed by ChatGPT in late 2022 — the famous 'Code Red' was an admission that the company had let OpenAI ship a product Google had the technology to ship years earlier. The Bard launch was rushed and embarrassing. The Gemini image-generation launch in early 2024 was a self-inflicted reputational wound. Cost discipline only arrived in 2023 after activist pressure and a stock that had underperformed. The dual-class structure means GOOG shareholders cannot vote out a leadership team that was, until recently, fat and slow. The capex ramp into AI infrastructure is being communicated as 'investment' but bears should ask: what is the ROI hurdle? What revenue must Cloud AI generate to justify $50B+ of incremental annual capex? Management has not answered that publicly with the rigor a thoughtful capital allocator would.

What bulls are extrapolating that won't hold. Bulls extrapolate (a) Search ad revenue growing low-teens forever, (b) YouTube continuing to take TV ad share, (c) Cloud reaching AWS-like margins, (d) Other Bets eventually monetizing (Waymo). Each of these is a guess. Search ad growth is decelerating as click volume per query falls under AI Overviews. YouTube faces TikTok and ad-load saturation. Cloud's path to AWS-margins is not obvious — Alphabet is third in a market with two stronger competitors, and the AI-workload mix is lower-margin than traditional Cloud. Other Bets has consumed tens of billions over a decade with one near-commercial product (Waymo) whose unit economics remain unproven. The reverse-DCF requires 10.3% growth — modest in isolation but ambitious if Search is structurally challenged and Cloud growth disappoints.

Valuation trap. P/E of 42x TTM is more than 35% above the 10-year average of 31x. EV/FCF of 63x is the kind of multiple investors pay for unimpaired 20%+ growth, not 10% growth. The IV/base of $513 assumes mid-teens earnings growth and stable margins. Compress the multiple to a more historical 25x on TTM owner-earnings, hold earnings flat for two years while AI capex digests, and you arrive at a stock with a $260-$300 price tag. That is not a black-swan scenario; that is what 'multiple normalization' looks like when growth disappoints by even modest amounts.

If I am right, the stock could be worth $260-$300 within 2-3 years.

Lollapalooza Bias Check

Several biases are actively pulling at me as I write this analysis.

Social proof and authority. Alphabet is owned by virtually every long-only manager I respect. The names in the 13F filings — including Berkshire-adjacent value investors who would normally avoid 42x P/E names — provide a comforting chorus. The bias whispers: 'if everyone smart owns it, you don't need to think too hard.' I should disregard the chorus and focus on the cash flows.

Confirmation bias. I went into this analysis already believing Alphabet is one of the great businesses of the modern era. Every piece of evidence I encountered — the 16.8% ROIC, the 26.9% ROIIC, the moat layering — confirmed that prior. I did not work nearly as hard to find disconfirming evidence as I did to articulate the bull case. The inversion section above is a partial corrective, but I should be honest that I find it less viscerally compelling than the bull case despite trying to make it strong.

Anchoring. The IV base of $513 from the deterministic scorer is anchoring my sense of 'fair value.' That number is itself a function of inputs — owner-earnings, growth assumptions, discount rate — that contain real uncertainty. The 'uncertain maintenance capex' note in the scorer output is a flag I should weight more heavily than I am instinctively weighting it. If maintenance capex is meaningfully higher than the model assumes, the IV could be 20-30% lower.

Recency bias. The narrative that 'Alphabet won the AI race after Code Red' is recent and hopeful. The actual evidence is that Gemini is competitive but not clearly superior to GPT or Claude, that AI Overviews have not yet been proven to be a net revenue positive, and that the competitive landscape is the most fluid it has been in fifteen years. I am giving Alphabet credit for an outcome that is not yet in the financials.

Commitment and consistency. I just wrote a Buy thesis on GOOGL. Writing anything other than essentially the same thesis on GOOG would feel inconsistent — but the share-class distinction is real and worth examining on its own merits, not anchored to the GOOGL conclusion. (In this specific case the underlying business and economics are identical, so the consistency is appropriate. But I noticed the bias.)

Incentive bias (low). I have no economic exposure to GOOG and no professional incentive to reach a particular conclusion. This is the cleanest force-vector in the analysis.

Net effect: my biases collectively push toward a more bullish stance than the evidence strictly warrants. The Hold/Buy boundary should be approached with that tilt in mind.

10-Year Outlook

Will Alphabet have the same fundamental business model in 2036? Probably yes for the underlying job — connecting human intent with relevant information and monetizing that connection. Probably no for the surface form. Search-as-ten-blue-links will be a smaller share of total query volume than today; AI-mediated answers, agentic workflows, and conversational interfaces will be larger. Alphabet has positioned itself to be the provider of the new form (Gemini, AI Overviews) while still owning the back-end auction. Confidence in 'still in business and still big': very high. Confidence in 'same product surface': low.

Will the customer base be larger? Almost certainly yes for users (global internet usage continues to grow, AI lowers usage friction further). Probably yes for advertisers (digital ad market continues to grow, more SMBs come online). The interesting question is queries per user — AI may reduce search volume by satisfying questions in one shot rather than ten, but may increase total queries by enabling new use cases.

Will profit per customer be higher? This is the core uncertainty. Bulls assume yes — better targeting, more ad surfaces, premium AI offerings. Bears assume no — AI compresses ad inventory, regulators force concessions, competition intensifies. My honest read is roughly flat to modestly higher in real terms. That is not a disaster for a 16.8% ROIC business, but it is well below what the current 42x P/E implies.

Will the moat be wider in 2036? Probably narrower in Search (AI substitutes are real), probably wider in AI infrastructure (TPU advantage compounds), probably wider in YouTube (creator network effects continue), probably wider in Cloud (third place becomes second place). On balance, roughly the same. The ecosystem has more layers but each layer faces more credible competition than it did in 2016.

Single biggest threat. A regulatory remedy that breaks the Apple-default-search arrangement, combined with an AI-native search competitor that captures the iPhone default. That specific combination is the lowest-probability/highest-impact scenario in the matrix.

Summary judgment. Alphabet will be a large, profitable, structurally important business in 2036. Whether it will be a 16.8% ROIC business at materially larger scale is genuinely uncertain. The base case is yes; the bear case is plausible.

CONFIDENCE: medium

Position Guidance

• Recommendation: Buy
• Conviction: medium-high
• Target buy price: $385 or below (P/IV at or below 0.75x base)
• Aggressive add price: $345 (at the low IV — full margin of safety)
• Target trim price: $555 (above bull-case IV; reduce position into strength)
• Position sizing: 4-6% of portfolio at current price; up to 7-8% on a pullback to the $345 zone. Treat GOOG and GOOGL as fungible — own whichever trades cheaper at fill, do not double up across both share classes for a single thesis.
• Note on share class: GOOG is non-voting Class C; economic claim is identical to GOOGL Class A. The historical discount has compressed to near-zero due to disproportionate Class C buybacks. If GOOG trades more than 0.5% below GOOGL, prefer GOOG; otherwise either is fine.