Google Stock After Q2 2026: The $6.26 That Isn't Earnings

Alphabet's headline earnings tripled, its trailing P/E collapsed to roughly 18, and neither number describes the operating business — because $6.26 of the $9.11 in diluted EPS was an unrealized paper gain on stakes in SpaceX and Anthropic. Strip it out and the quarter earned about $2.85 per share against a $2.89 consensus. The stock's behaviour since — down through late July on capex, then down 3.9% to $360.71 on 5 August after an AI leadership shake-up, erasing close to $190 billion of a $4.61 trillion market value — makes considerably more sense once you know that. 

Which ticker is "Google stock"?

Google itself does not trade; the listed parent is Alphabet Inc., under two tickers with identical economics and different voting rights. GOOGL is the Class A share and carries one vote; GOOG is the Class C share and carries none. They typically trade within a small percentage of each other, and for an investor with no interest in voting, the practical rule is to buy whichever is cheaper on the day. (Class B shares, with ten votes each, are held by founders and insiders and are not publicly traded.) Both classes carry the same dividend — most recently $0.22 per share with an ex-date of 4 September 2026. 

What did Alphabet report in Q2 2026?

Alphabet grew revenue 24% to $119.8 billion with Google Cloud accelerating to 82% growth, then reported a net income figure inflated almost threefold by investment marks. Consolidated operating income rose 30% and operating margin expanded two percentage points to 34%, while Google Cloud revenue increased 82% to $24.8 billion, led by enterprise AI infrastructure and solutions. Google Services generated $94.5 billion, up 15%, with Search and other advertising up 17% to $63.3 billion and YouTube advertising up 13% to $11.1 billion.

Cloud operating income reached $8.8 billion versus $2.8 billion a year earlier. The operating business is genuinely strong. The bottom line is something else. 

Why is the headline EPS misleading?

Because Alphabet itself quantified the distortion: the equity-securities gain added $6.26 to diluted EPS. The company disclosed that other income reflected a net gain of $98.0 billion, primarily the result of net unrealized gains on its equity securities, and that net income available to common stockholders increased 298% while EPS increased 294% to $9.11. The filing further specifies that the $99.0 billion equity gain increased the provision for income taxes by $21.9 billion, net income by $77.1 billion and diluted EPS by $6.26. 

CNBC reported adjusted EPS of $2.85 against $2.89 expected by LSEG, alongside revenue of $119.80 billion versus $116.93 billion expected — a clear revenue beat and a narrow earnings miss, which is a very different quarter from "profit up 298%." 

Where did the gain come from? Alphabet reclassified its SpaceX holding from a cost-method private stake to marketable equity securities after SpaceX's 12 June 2026 IPO; as of 30 June 2026 the position was worth $94.0 billion, representing effective ownership of about 4.9%. Alphabet described the gains as primarily related to unrealized gains from SpaceX and a private company, the latter widely understood to be Anthropic. 

The critical implication is symmetry. Mark-to-market cuts both ways: the same line that added $77 billion to net income this quarter will subtract from it if those positions fall. Alphabet's P/E ratio now reads around 18, which makes it look much cheaper than it really is precisely because that multiple rests on unadjusted earnings skewed by how SpaceX and other investments perform. Any valuation screen ranking Google stock on trailing P/E is currently measuring the wrong thing.

The real story is capital expenditure

Alphabet is now spending more on infrastructure than it generates in cash, and it has started funding the gap in the equity market. Capital expenditures totalled $44.9 billion in the quarter, producing negative free cash flow, and full-year 2026 guidance was raised to $195–205 billion from $180–190 billion. Free cash flow was negative $5.9 billion in the quarter as capital spending exceeded operating cash generation, with roughly 60% of planned spending directed to servers and 40% to data centres and networking. Analysts polled by Visible Alpha had been modelling about $188 billion for 2026.

The financing detail deserves more attention than it has received: in June 2026 Alphabet issued a combination of Class A stock, Class C stock and mandatory convertible preferred stock for aggregate net proceeds of $49.6 billion, to be used for general corporate purposes including capital expenditures to scale AI infrastructure and global compute.

For two decades Alphabet was a company that retired shares. It has now issued equity — including a convertible instrument that dilutes on conversion — to fund compute. The balance sheet still held $242.5 billion in cash, equivalents and marketable securities, including $87.1 billion of marketable equity securities. With that cash pile, raising $49.6 billion externally is a signal about the scale and duration of the spending plan, not about liquidity. That is the single most important fact in this quarter for a long-horizon holder. 

The bull case for the spend is the backlog: Cloud backlog reached $514 billion while management signalled continued supply constraints. Contracted future revenue of that size is a real, if lagging, answer to "where is the return?" 

Is the AI talent exodus a fundamental problem?

The 5 August departures cost more market value in one day than a quarter of Alphabet's annual capex budget, which tells you the market prices AI leadership as a hard asset. Demis Hassabis became chair of Google DeepMind and chief scientist of Alphabet while Koray Kavukcuoglu stepped up as senior vice president owning Gemini model development, frontier research and the Gemini app; chief scientist Jeff Dean left after 27 years to start a public benefit corporation. Google veterans Oriol Vinyals, Quoc Le and Sanjay Ghemawat were reported to be joining Dean as co-founders, deepening a recent pattern of AI talent departures. Alphabet Class A fell 3.8% intraday, with early estimates putting the equity value loss near $175 billion. 

Two readings, and the honest answer is that this quarter's data cannot distinguish them. The benign reading: Hassabis is being elevated, not sidelined; Kavukcuoglu is a 13-year internal successor; Dean's new venture is Alphabet-backed. The bearish reading: investors have run this play before — Alphabet closed 5.1% lower on 22 June, its worst session in a year, after two researchers left days apart. A repeated pattern of senior exits during a capex supercycle is exactly the sequence that would precede execution slippage. 

Where does the antitrust overhang stand?

The search case produced behavioural remedies rather than a breakup, and both sides are appealing. The district court rejected the plaintiffs' proposals for structural relief — an immediate Chrome divestiture and a contingent Android divestiture — imposing several behavioural remedies instead. Google filed its notice of appeal on 16 January 2026 challenging the data-sharing requirements and technical committee oversight, while the DOJ filed a cross-appeal by its 3 February 2026 deadline seeking stronger remedies including forced divestitures, with appellate arguments expected in late 2026 or early 2027. The remedies order itself took effect on 3 February 2026 while both appeals proceed at the D.C. Circuit. 

Practically: the worst-case structural outcome was avoided at the district level but is back in play on appeal, and the timeline runs past this year. A separate ad-tech remedies proceeding before a different court has been pending; readers should verify its current status before acting, as the outcome would bear directly on Alphabet's advertising stack.

What are you paying at roughly $361, and what breaks the thesis?

You are paying about 4.6 trillion dollars for a business whose reported earnings are currently unreadable and whose cash generation is temporarily negative by design. Analyst targets over the past month averaged $421.45 with a high of $515 and a low of $240 — a spread of more than 2x between the most and least optimistic, which is unusual for a mega-cap and reflects genuine disagreement about capex returns rather than about the ad business.

Four checkpoints, in order of information value:

1. Q3 other income. If SpaceX or Anthropic marks fall, the same line reverses. Watch whether GAAP net income declines sharply on an unchanged operating business — and do not read that as deterioration if it happens.

2. Free cash flow. Negative $5.9 billion this quarter against a rising capex guide. Two or three more negative quarters would test the market's tolerance and the dividend/buyback capacity. 

3. Cloud backlog conversion. $514 billion of backlog is the justification for the spend; the revenue recognition pace is the proof.

4. Search revenue growth against AI substitution. Management noted that from Q3 the company begins lapping an acceleration in Search performance that started in the third quarter of last year — the comparison gets harder just as AI-native competitors scale.

FAQ

Q: Should I buy GOOGL or GOOG?
They represent the same economic ownership of Alphabet and pay the same dividend; GOOGL (Class A) carries one vote per share and GOOG (Class C) carries none. Unless a vote matters to you, buy whichever trades lower at the time of purchase. The price gap between them is usually small.

Q: Why did Alphabet's profit jump 298% in Q2 2026?
Almost entirely because of investment marks, not operations. Other income reflected a net gain of $98.0 billion, primarily unrealized gains on equity securities. That gain added $6.26 to diluted EPS of $9.11. The underlying figure of about $2.85 was slightly below the $2.89 consensus. 

Q: Is Google stock cheap at a P/E of 18?
Not in the way the number suggests. The trailing multiple is based on unadjusted earnings that are skewed by how SpaceX and other investments perform, which means the same accounting can make the stock look expensive next quarter without anything changing in Search, YouTube or Cloud. Valuation work here needs an ex-equity-gains earnings base.

Q: How much is Alphabet spending on AI in 2026?
Full-year 2026 capital expenditure guidance is $195–205 billion, raised from $180–190 billion, with $44.9 billion spent in the second quarter alone. Alphabet also raised $49.6 billion in net proceeds from a June 2026 equity and mandatory convertible preferred issuance, partly to fund AI infrastructure. 

Q: What is the current status of the Google antitrust case?
The court rejected structural relief such as a Chrome divestiture and imposed behavioural remedies instead. Google appealed in January 2026 and the DOJ cross-appealed in February 2026 seeking divestitures, with appellate arguments expected in late 2026 or early 2027. A separate ad-technology remedies decision has been pending before a different court.