GOVERNANCE WATCH
By InsidEntity Editorial Desk · Jul 25, 2026 · 9 min read
Read the headline and Alphabet’s second quarter looks like a clean beat. Revenue of $119.8 billion, up 24% year on year, a twelfth straight quarter of double-digit growth. Google Cloud, the number every AI-era investor watches, accelerated to 82% growth and turned a genuine profit, $8.8 billion of operating income against $2.8 billion a year earlier. Net income reached $112.1 billion against $28.2 billion in the same quarter last year, and diluted EPS came in at $9.11, up from $2.31, a 294% jump. CEO Sundar Pichai called it an “amazing quarter,” and the operating numbers underneath, a 34% margin and consolidated operating income up 30% to $40.8 billion, back him up.
Now look at where most of that profit actually came from. Not Search, not Cloud, not YouTube. A single “other income” line carried a net gain of roughly $99 billion, primarily unrealized marks on Alphabet’s equity holdings in private companies it does not operate, led by its stake in Anthropic. Of the $112.1 billion in net income, about $99 billion was gains on equity securities; the business the company actually runs produced $40.8 billion in operating income. Both numbers are real. Only one of them is the business. This is the same structural story this series covered in Tesla’s second quarter, where a $1.005 billion mark on Tesla’s SpaceX stake was the difference between a breakeven quarter and a reported profit. Alphabet just ran the identical play two orders of magnitude larger, and shares still fell after hours as capital-expenditure guidance rose to $195 to $205 billion, a reminder that markets read through the equity marks even when headlines do not.
That is exactly the kind of number the InsidEntity Company Risk Rating framework exists to place in context: not to say whether the quarter was good, but to show what governance structure a shareholder is actually relying on while that quarter gets reported. Alphabet carries an InsidEntity Company Risk Rating of 3.40, Excellent. Full breakdown, director-by-director, on Alphabet’s InsidEntity report.
The quarter, side by side
| Metric | Q2 2025 | Q2 2026 | Change |
|---|---|---|---|
| Revenue | $96.4bn | $119.8bn | +24% |
| Google Cloud revenue | $13.6bn | $24.8bn | +82% |
| Cloud operating income | $2.8bn | $8.8bn | +214% |
| Operating income | $31.3bn | $40.8bn | +30% |
| Operating margin | 32% | 34% | +2 pts |
| Net income | $28.2bn | $112.1bn | +298% |
| Of which, equity-securities gain | $1.3bn | $99.0bn | Substantially the entire increase |
| Diluted EPS | $2.31 | $9.11 | +294% |
| Capital expenditure | $44.9bn | 2026 guidance raised to $195–205bn | |
| Free cash flow | -$5.9bn | Turned negative on capex |
Source: Alphabet Inc. Q2 2026 earnings release (SEC Form 8-K exhibit) and Form 10-Q, filed July 22, 2026, with prior-year comparatives from the same filing.
Read the equity-gain row against the operating-income row. Alphabet is spending heavily, capex of $44.9 billion in the quarter, free cash flow now negative, guidance raised toward $205 billion for the year, to keep Cloud’s acceleration going, and that spending is a real and encouraging operating story. But the number that made the quarter look spectacular on a headline basis was a mark-to-market gain on minority stakes in AI companies whose accounts Alphabet does not control. Both stories are true. They are not the same story.
What the structure looks like underneath it
Board independence, disclosed rather than defaulted. Seventy percent of Alphabet’s ten-director board, seven of ten, is independent under Nasdaq and SEC standards, per the 2026 proxy statement. InsidEntity does not take a director’s declared independence at face value: Director Independence carries the heaviest weight in the Company Risk Rating, 40%, precisely because the platform tests how that independence is calculated and measured rather than accepting a self-reported label. Chair and CEO roles are split: John Hennessy, a former Stanford president and Alphabet director since 2004, chairs the board; Sundar Pichai runs the company. Alphabet also caps outside public-company board service at four boards including its own, and every director is currently in compliance.
Auditor tenure: dating to before the company was public at all. Ernst & Young’s consent letter for Google’s IPO registration statement is dated April 2004, meaning EY was already Google’s auditor before the company had public shareholders. Every set of Alphabet accounts ever filed with the SEC has been signed off by the same firm, a relationship now past twenty-two years as public-company auditor and longer still counting the pre-IPO years. On the tenure ladder this series has applied at AMD (Ernst & Young since 1969, nil) and Tesla (PwC, twenty-one years, nil), Alphabet’s relationship with EY sits in the same territory: no fresh set of eyes has ever examined Alphabet’s books.
Shareholder influence: concentrated in votes, not in economic ownership. This is where Alphabet’s structure earns its own paragraph rather than reusing GM’s or Tesla’s. Larry Page and Sergey Brin each hold a Class A/Class B stake below 6% of shares outstanding individually, under the 20% threshold the framework treats as a material shareholder, the same test Tesla’s Musk clears in the mid-teens. But shares are not votes at Alphabet. Class B stock carries ten votes per share against Class A’s one, and Page and Brin together control roughly 52.7% of total voting power on a combined economic interest of a little over a tenth of the company. BlackRock, the largest disclosed institutional holder, carries about 2.5% of total voting power against a considerably larger economic stake. Measured the way the framework measures Tesla and GM, no single holder over 20%, Alphabet’s register reads clean. Measured by what actually decides a contested vote, two people who together own a minority of the economics hold a majority of the votes. Both readings are correct. Only one of them is what “shareholder influence” is trying to capture.
Director capacity and the related-party map worth watching. L. John Doerr, chair of the venture firm Kleiner Perkins and an Alphabet director since 1999, sits on a board whose venture connections overlap with companies Alphabet itself invests in and acquires from, a network Alphabet discloses and its Audit Committee reviews. Alphabet’s proxy also discloses related-party arrangements connected to its founders, including airfield and hangar leases with founder-affiliated entities, each reviewed against market terms. None of this is a concealment story; all of it is in the filings. It is, like AMD’s OpenAI warrant, a governance object the earnings coverage never mentions: a board with real financial and personal ties running alongside the company it oversees, at the same moment that Alphabet’s own reported profit depends materially on the market value of its stakes in the AI companies it invests in and supplies. Where the specific transactions and amounts matter to a reader, the platform’s director records and Alphabet’s proxy are the place to read them in full.
The pillar scorecard
| Component | Score | Read |
|---|---|---|
| Director Independence | 5.0 | Benchmark, consistent with the 70% independent board disclosed in the 2026 proxy |
| Shareholder Influence | 5.0 | Benchmark on the framework’s 20%-economic test, and the clearest illustration in this series of what that test does not capture |
| Director Capacity | 3.6 | Good, the highest capacity score this series has recorded, on a board within its own four-seat service cap |
| Auditor Independence | 0.0 | Nil, the tenure ladder applied: EY since before the 2004 IPO |
| Overall Rating | 3.40 | Excellent |
(1 Risk, 2 Caution, 3 Good, 4 Excellent, 5 Benchmark, per InsidEntity’s Company Risk Rating methodology, weighted 40/20/20/20 with Director Independence carrying the heaviest weight. Check Alphabet’s live rating for the current view.)
The scorecard is worth reading against the section above it. Shareholder Influence scores a clean 5.0, Benchmark, because no holder crosses the 20% economic threshold. Yet Page and Brin control a majority of the votes on roughly a tenth of the economics. The 5.0 is not an error; it is the framework measuring exactly what it measures, ownership concentration, and by that measure Alphabet’s register is dispersed. The voting concentration is real and sits in the record beside the score, which is why a director-and-structure record exists alongside a single number. Auditor Independence at nil is the pillar doing the most damage to what would otherwise be a near-Benchmark profile: three pillars at 5.0, 5.0 and 3.6, one at zero, is the profile of a well-governed company with one structural relationship the framework has stopped crediting.
Reading the quarter and the structure together
None of this says Alphabet had a bad quarter, or a dishonestly reported one, everything above is disclosed, audited, and sitting in the filings for anyone who reads past the press release. What it says is narrower and more useful: the headline number that made this Alphabet’s best quarter on paper was a mark on AI companies it does not run, reviewed by a board that is majority-independent but auditor-unchallenged for over two decades, elected by a shareholder base where the two people with the most say own a fraction of the economics they vote. Cloud’s acceleration to 82% growth, and its swing to real operating profit, is the story in this release. The $99 billion gain is the one that will show up in the headline anyway.
The detail worth sitting with: Alphabet’s stake in Anthropic sits inside a company Alphabet also supplies with Google Cloud infrastructure and TPU capacity, a commercial relationship and an equity mark moving together, on a scale that dwarfs Tesla’s SpaceX position. The same customer whose deployments help drive Cloud’s reported growth is also the holding whose valuation drives the equity gain. Same structure as the vendor-financing circularity this series has flagged across the AI trade, at a different order of magnitude. Markets have started pricing the difference: the after-hours reaction to a beat-and-raise quarter was a decline, not a rally.
Pull up Alphabet’s full Company Risk Rating on InsidEntity, cross-check the director tenure clocks against Doerr’s KPCB co-investment pattern, and watch the Auditor Independence pillar the way this series has watched it at every stop so far. An earnings release tells you what the business did. The governance layer tells you what the number is actually resting on.
About the InsidEntity Company Risk Rating
The InsidEntity Company Risk Rating (CRR) scores companies on a 1 to 5 scale across four governance pillars, weighted 40/20/20/20: Director Independence, Director Capacity, Auditor Independence, and Shareholder Influence. A 5 is Benchmark, a 1 is Risk, and Director Independence carries the heaviest weight, so a weakness there shows up fastest in the headline number. For more on how the score is built, see our Company Risk Rating methodology. Ratings update regularly as corporate changes occur, so the platform is always the current view. Explore ratings for companies across 145 stock exchanges at InsidEntity.
This article is part of InsidEntity’s Company Risk Rating series. Issue 3 of The World Inside, InsidEntity’s newsletter, is available at the end of July 2026.