Alphabet at 29x: What the AI Discount Is Missing

Three businesses. Alphabet runs the world's dominant search engine, the world's second-largest cloud platform, and the world's most-watched video platform under one corporate umbrella. Understanding what you own when you buy GOOGL requires understanding how those three businesses interact, and increasingly, how AI is reshaping the economics of each.

Search is still the engine. Google Search generated roughly 57% of total 2025 revenue, approximately $230 billion from a single product at margins well above the company average. No other business on earth runs anything of comparable scale and profitability.

Cloud is the growth story. Google Cloud crossed $40 billion in annualized revenue and is now profitable, with operating margins expanding as scale compounds against a largely fixed infrastructure base. The business trails AWS and Azure in market share but is gaining ground, particularly in AI workloads where Google's in-house TPU chip advantage creates a differentiated cost structure.

YouTube is the sleeper. YouTube advertising revenue exceeded $36 billion in 2025, making it larger than Netflix's total subscription revenue. The market consistently undervalues this asset when evaluating Alphabet. The platform's shift toward Shorts has compressed near-term monetization rates but positions it well against TikTok and Instagram Reels in the longer format war for attention.

The thread connecting all three is Gemini, Alphabet's AI infrastructure stack. The dominant themes in April 2026 news coverage are AI and technology investment, which reflects the market's recognition that the next chapter for search, cloud, and video runs through the same technology platform.

Revenue grew 15.1% in 2025 to $402.8 billion. Net income grew 32.1% to $132.2 billion. These are not the numbers of a business under structural threat. They are the numbers of a business accelerating.

The gross margin trajectory tells a more interesting story than the top-line growth. In 2022, when Alphabet was broadly dismissed as a legacy advertising business facing AI disruption, gross margins sat at 55.4%. By 2024 they were 58.2%. By 2025 they hit 59.7%. Three consecutive years of margin expansion, not compression, through the very period when AI disruption of search was supposed to be reaching its inflection point.

Operating margins held at 32.1% in 2025 despite a significant step-up in AI infrastructure investment. Capital expenditures reached $91.4 billion, up from $52.5 billion in 2024 and $32.3 billion in 2023. That $91 billion went substantially into data centers, TPU capacity, and network infrastructure. Holding operating margins flat while absorbing that investment wave implies significant underlying leverage in the core advertising business.

The earnings surprise record is also instructive. Across the four reported quarters through 2025, Alphabet beat consensus EPS estimates by 39.8%, 5.5%, 23.7%, and 7.2%. Four consecutive beats. Consensus has systematically underestimated this business, which typically signals either a persistent modeling error or a structural misunderstanding of how the earnings compound.

Operating cash flow reached $164.7 billion in 2025. Free cash flow of $73.3 billion understates the underlying earnings power because the capex surge is front-loaded investment with a multi-year payback horizon. The operating cash flow figure tells you what the business generates before that investment is subtracted.

The consensus fear is that AI destroys the search advertising model by replacing ten blue links with a direct answer that carries no ad placement. The consensus data suggests something different is happening.

AI Overviews, Google's AI-generated answer feature, rolled out broadly across 2025. In the quarters following that rollout, search revenue grew. This either means the monetization of AI Overviews is working better than feared, or that the user base expanded enough to offset per-query revenue compression, or both. The operating margin picture does not suggest a business experiencing pricing pressure from its own product.

The more plausible scenario is that AI makes search more useful for complex queries, which increases total query volume, which increases the total monetizable surface area. Google does not need to monetize every AI answer at the same rate it monetizes traditional search queries. It needs total advertising revenue to grow, and through 2025, it did, at an accelerating rate.

The OpenAI competitive threat is real but constrained by distribution. AI assistants are genuinely excellent at certain tasks, particularly long-form content generation and coding assistance. They are not, in practice, replacing the reflexive daily search behavior of billions of users who type queries into a Chrome address bar or an Android search widget. Changing distribution requires changing the hardware or changing the browser, both of which are structurally defended positions for Alphabet.

Recent coverage from April 12 reports OpenAI building toward a $100 billion advertising empire by 2030. This is the right framing for the competitive threat: a 2030 problem, not a 2026 problem. At current Google Cloud growth rates and current search margins, Alphabet will have generated substantially more cash before OpenAI's advertising business exists at scale.

Alphabet has returned roughly $115 billion to shareholders over the past two years through a combination of buybacks and a newly initiated dividend program. In 2024, buybacks totaled $62.2 billion against $7.4 billion in dividends. In 2025, buybacks moderated to $45.7 billion as capital was redirected toward the $91 billion capex program, while dividends grew to $10 billion.

The share count reduction tells a clean story. From 2020 through 2024, Alphabet reduced its share count from approximately 13.66 billion shares to 12.35 billion shares, a reduction of roughly 9.6% over four years. Systematic, consistent, and value-accretive at every price point at which it was executed.

The buyback moderation in 2025 deserves correct framing. Operating cash flow rose from $125.3 billion in 2024 to $164.7 billion in 2025. Alphabet is not reducing buybacks because cash generation is weakening. It is redirecting capital toward AI infrastructure that management believes will compound value at a higher rate than share repurchases at current prices. That is different from a business cutting buybacks under financial stress.

The dividend initiation in 2024, growing from $7.4 billion to $10 billion in 2025, is a behavioral signal as much as a capital allocation decision. Companies with genuine uncertainty about their earnings trajectory preserve optionality. They do not initiate and grow dividends.

The most persistent misunderstanding about Google's competitive position is that it can be disrupted by a better technology. This misunderstands what the moat actually is.

The moat is distribution: the Android operating system on approximately two billion active devices, the Chrome browser with roughly two-thirds of global desktop market share, and the default search agreements that put Google at the entry point for every query on Apple hardware. These distribution channels did not exist because Google built the best algorithm. They exist because Google spent two decades and hundreds of billions of dollars building the infrastructure that connects users to its services at the hardware and software layer.

Matching Google's technology is difficult. Matching Google's distribution is structurally impractical without regulatory intervention. This is precisely why the antitrust case matters and why it is the legitimate bear case risk. The government's proposed remedies attack the distribution moat, not the algorithmic one. A court order ending default search agreements would remove a structural advantage that competitors cannot replicate organically.

Google Cloud's competitive position is separate from Search and consistently underappreciated. The integration of Google's proprietary AI chips into Cloud infrastructure provides a cost and latency advantage for AI workloads that neither AWS nor Azure can easily replicate. This advantage took over a decade of TPU research to build, which means the lead has a meaningful time horizon before it is eroded.

At 29.3x trailing earnings, Alphabet trades at a discount to the broader technology sector, a discount to Microsoft, and roughly in line with the overall market index. For a business that grew net income 32% last year and operates at the intersection of the three dominant technology themes of the decade, this multiple is the central debate.

The simple math: $132 billion in net income growing at low-to-mid teens over the next three years reaches $170-180 billion by 2028. At a 25x exit multiple, conservatively, that implies a $4.25-4.5 trillion enterprise value against a current valuation around $3.8 trillion. The return does not depend on multiple expansion. It depends on the earnings continuing to compound, which they have done without interruption through macro cycles, pandemic disruptions, and now AI disruption fears.

The EV/EBITDA of 20.8x is perhaps more useful given the capex surge. EBITDA of $180 billion in 2025 reflects the true earnings power before the AI infrastructure investment is deducted. Paying 20x EBITDA for a business with 60% gross margins, structural pricing power in its core market, and a technology position that every major competitor is working to replicate is not obviously expensive.

The sentiment data adds context. Through late March 2026, normalized sentiment for GOOGL fell to a trough range of 0.51-0.59, likely reflecting the broader tariff-driven market sell-off and AI disruption headlines amplified by macro volatility. By April 12, that sentiment had recovered to 0.84, even as the fundamental trajectory remained unchanged throughout. The financial picture does not explain the March sentiment trough. The noise level does. Entries made into that noise have historically been the ones worth making.